0

ASME Conference Presenter Attendance Policy and Archival Proceedings

2016;():V002T00A001. doi:10.1115/DSCC2016-NS2.
FREE TO VIEW

This online compilation of papers from the ASME 2016 Dynamic Systems and Control Conference (DSCC2016) represents the archival version of the Conference Proceedings. According to ASME’s conference presenter attendance policy, if a paper is not presented at the Conference by an author of the paper, the paper will not be published in the official archival Proceedings, which are registered with the Library of Congress and are submitted for abstracting and indexing. The paper also will not be published in The ASME Digital Collection and may not be cited as a published paper.

Commentary by Dr. Valentin Fuster

Mechatronics

2016;():V002T17A001. doi:10.1115/DSCC2016-9601.

In this effort, we focus on determining the safe operational domain of a coupled actuator-valve configuration. The so-called “Smart Valves” system has increasingly been used in critical applications and missions including municipal piping networks, oil and gas fields, petrochemical plants, and more importantly, the US Navy ships. A comprehensive dynamic analysis is hence needed to be carried out for capturing dangerous behaviors observed repeatedly in practice. Using some powerful tools of nonlinear dynamic analysis including Lyapunov exponents and Poincaré map, a comprehensive stability map is provided in order to determine the safe operational domain of the network in addition to characterizing the responses obtained. Coupled chaotic and hyperchaotic dynamics of two coupled solenoid actuated butterfly valves are captured by running the network for some critical values through interconnected flow loads affected by the coupled actuators’ variables. The significant effect of an unstable configuration of the valve-actuator on another set is thoroughly investigated to discuss the expected stability issues of a remote set due to others and vice versa.

Commentary by Dr. Valentin Fuster
2016;():V002T17A002. doi:10.1115/DSCC2016-9658.

This paper presents a model to explain complex non-minimum phase (CNMP) zeros seen in the non-collocated frequency response of a large displacement XY flexure mechanism, which employs multiple double parallelogram flexure modules (DPFM) as building-blocks. Geometric non-linearities associated with large displacement along with the kinematic under-constraint in the DPFM, lead to a coupling between the X and Y direction displacements. Via a lumped-parameter model that captures the most relevant geometric non-linearity, it is shown that specific combinations of the operating point (i.e. flexure displacement) and mass asymmetry (due to manufacturing tolerances) give rise to CNMP zeros. This model demonstrates the merit of an intentionally asymmetric design over an intuitively symmetric design in avoiding CNMP zeros. Furthermore, a study of how the eigenvalues and eigenvectors of the flexure mechanism vary with the operating point and mass asymmetry indicates the presence of curve veering when the system transitions from minimum phase to CNMP. Based on this, the hypothesis of an inherent correlation between CNMP zeros and curve veering is proposed.

Commentary by Dr. Valentin Fuster
2016;():V002T17A003. doi:10.1115/DSCC2016-9683.

This paper presents a mechatronic modeling analysis of a 4×4 hybrid-electric vehicle (HEV) which uses an active driveline system connected to the hybrid powertrain. This active driveline uses a power-splitting device, named a Hybrid-Electric Power-Transmitting Unit (HE-PTU) to control the power split between the front and rear axles. The mathematical model of the active driveline along with two passive drivelines demonstrates the coupling of driveline-steering-system influenced lateral dynamics of the vehicle. Thus, a more flexible, active driveline is able to effectively decouple the driveline and steering systems by producing a compensating torque that influences the tire lateral forces and thus vehicle lateral dynamics. Because of the multiple domains involved in modeling the HE-PTU, a mechatronics-based modeling solution is required to demonstrate the advantage of the active driveline. The mechatronics-based simulation results show how use of the active driveline with a hybrid-electric power transmitting unit can improve the vehicle’s turnability and stability characteristics.

Commentary by Dr. Valentin Fuster
2016;():V002T17A004. doi:10.1115/DSCC2016-9724.

As a novel alternative of internal combustion engine (ICE), the free piston engine (FPE) eliminates the mechanical crankshaft and the associated constraints on its piston motion. Due to this extra degree of freedom and reduced inertia, the FPE is able to generate variable output power with higher efficiency and less emissions, while possessing a short response time. Hence, a hydraulic FPE (HFPE), which combines the FPE with a linear hydraulic pump, is a promising candidate as a fluid power source, especially for mobile applications. In this paper, such potential is investigated. The working principle of a prototype HFPE as a fluid power source is described and a comprehensive HFPE model is developed. Two novel control methods are proposed to regulate the output flow rate at any given load pressure so as to realize throttle-less fluid power control. Effectiveness of these two methods are demonstrated through simulation, where results clearly show the effectiveness of both methods in providing different output flow rates at given load pressure, thus demonstrating the HFPE’s capability as an efficient and flexible mobile fluid power source.

Commentary by Dr. Valentin Fuster
2016;():V002T17A005. doi:10.1115/DSCC2016-9764.

In the recent past the design of many aquatic robots has been inspired by the motion of fish. In some recent work the authors described an underactuated planar swimming robot, that is propelled via the motion of an internal rotor. This robot is inspired by a simplified model of the fluid-body interaction mediated by singular distributions of vorticity. Such a model is a significant simplification of the fluid-structure interaction that can be understood using resource intensive numerical computations of the Navier Stokes equation that are unwieldy from a controls perspective. At the same time the simplified model incorporates the creation of vorticity and interaction of the body with the vorticity which many control theoretical models ignore. In this paper we show that despite the complexity of the interaction between the aquatic robot and the ambient vorticity in a fluid, the response of the robot is a nearly linear function of the control input. This surprisingly simple feature emerges in our theoretical model and is validated by our experimental data of the motion of the robot. This simplifying observation is an important step towards developing control algorithms for aquatic robots.

Commentary by Dr. Valentin Fuster
2016;():V002T17A006. doi:10.1115/DSCC2016-9766.

This paper proposes a discrete-time, multi-time-scale estimation and control design for quadrotors in the presence of external disturbances and model uncertainties. Assuming that not all state measurements are available, they will need to be estimated. The sample-data Extended High-Gain Observers are used to estimate unmeasured states, system uncertainties, and external disturbances. Discretized dynamic inversion utilizes those estimates and deals with an uncertain principal inertia matrix. In the plant dynamics, the proposed control forces the rotational dynamics to be faster than the translational dynamics. Numerical simulations and experimental results verify the proposed estimation and control algorithm. All sensing and computation is done on-board the vehicle.

Commentary by Dr. Valentin Fuster
2016;():V002T17A007. doi:10.1115/DSCC2016-9779.

The operation of autonomous underwater vehicles is often hindered by their battery capacity, limiting the duration of its use. Here, we propose an integrated solution for autonomous charging of a robotic fish via direct contact through a novel claw mechanism for docking guidance. To assist the robotic fish in the docking process, the system incorporates a charging station designed with form-fit claws. A controller is designed to monitor the battery level of the robotic fish during free swimming and coordinate the docking process with respect to the maneuvers of both the robot and form-fit claws. Upon recognizing a low battery level, the controller commands the robotic fish to begin the docking process, and video feedback from an overhead camera is used to inform the autonomous navigation toward the charging station. After reaching a battery level threshold, the robotic fish is then released back in the water and returns to free swimming until the battery is discharged again. Through a series of experiments, we demonstrate the possibility of prolonged operation, consisting of repeated cycles of autonomous charging. Our proposed charging method enables prolonged autonomous swimming with minimal human supervision, opening the door for new, transformative applications of robotic fish in laboratory research and field deployment.

Topics: Robotics
Commentary by Dr. Valentin Fuster
2016;():V002T17A008. doi:10.1115/DSCC2016-9841.

The objective of this paper is to present an adaptive multi-level fuzzy controller to stabilize the deflection of an electrostatically actuated microplate beyond its pull-in range. Using a single mode approximation along with utilizing the Lagrange equations, the dynamic behavior of the microplate is described in modal space by an ordinary differential equation. By different static and dynamic simulations, the system and the dependence of the deflection to the input applied voltage is identified linguistically. Then, based on the linguistic description of the system, a fuzzy controller is designed to stabilize the microplate at the desired deflections. To improve the performance specifications of the closed-loop system, another fuzzy controller at a higher level is designed to adjust the parameters of the main controller in real time. The simulation results reveal that by using the proposed single level and adaptive two level controllers, the control objective is met effectively with good performance specifications. It is also observed that adding a supervisory level to the main controller can reduce the overshoot and the settling time in beyond pull-in stabilization of electrostatically actuated microplates. The qualitative knowledge resulting from this research can be generalized and used for development of efficient controllers for N/MEMS actuators and electrostatically actuated nano/micro positioning systems.

Commentary by Dr. Valentin Fuster
2016;():V002T17A009. doi:10.1115/DSCC2016-9857.

Research in animal behavior has benefited from the availability of robots able to elicit controllable, customizable, and versatile stimuli in behavioral studies. For example, biologically-inspired robotic fish can be designed to mimic the morphophysiology of predators and conspecifics to study fear response and sociality. However, size is a critical limitation of the existing arrays of robotic fish. Here, we present the design of a miniature robotic fish for future animal-robot interaction studies featuring a novel application of multi-material three-dimensional (3D) printing and utilizing a solenoid for actuation. The use of multi-material printing enables a skeletal design of only two parts, while retaining the complete functionality of larger prototypes enclosing requisite electronics and incorporating an active joint for propulsion. Parametric tests are conducted to test the swimming speed of the robotic fish and a compact dynamic model with two degrees of freedom to elucidate swimming of the robotic fish is presented.

Topics: Robots , Robotics
Commentary by Dr. Valentin Fuster
2016;():V002T17A010. doi:10.1115/DSCC2016-9904.

A hybrid control system for multi degrees of freedom robotic manipulator is designed by integrating a proportional-integral-derivative controller (PID) and a model reference adaptive controller (MRAC) in order to further improve the accuracy and joint convergence speed performance. For the 1-DOF link, because the inertia matrices and nonlinear term of the dynamic equation are constant, we can directly combine the PID and MRAC controller to design the PID+MRAC controller. However, for the more than 1-DOF link case, it is no longer applicable because the inertia matrices and nonlinear term of the dynamic equation are not constant. By using an improved adaptive algorithm and structure, and by combining the PID and improved MRAC controllers, a controller is designed for the more than 1-DOF link case. The convergence performance of the PID controller, MRAC and the PID+MRAC hybrid controller for 1-DOF, 2-DOF and subsequently 3-DOF manipulators are compared.

Commentary by Dr. Valentin Fuster

Mechatronics and Controls in Advanced Manufacturing

2016;():V002T18A001. doi:10.1115/DSCC2016-9640.

This paper discusses the compensation of tool paths for machining flexible parts. Despite various research published on the topic, machining in practice nowadays remains limited to tool path planning based on only the geometric models of the parts and tools. This is mainly because that tool path compensation methods usually require accurate physical information of the systems and rely on analytical or finite element simulations, which are often not available to the end-users. In regards to this problem, this paper presents data-oriented nonparametric learning methods that require solely the geometric measurements of the trial machined contour(s). The physical parameters of the parts and tools as well as simulations of the machining process are not required. Two algorithms are developed based on Gaussian Process Regression and Artificial Neural Network respectively. Experimental tests are conducted. A plan of further improving the results using an auxiliary real-time vision sensor is also discussed.

Topics: Machining
Commentary by Dr. Valentin Fuster
2016;():V002T18A002. doi:10.1115/DSCC2016-9649.

This article considers the problem of achieving a desired supersaturated gas concentration profile inside a polymer part by controlling the gas concentration on the part boundaries at elevated pressure. Such controlled gas concentration can be used to achieve a desired cell-structure inside the polymer when foaming the gas inside as the part is rapidly heat-treated. The goal is to achieve the interior gas concentration at the end of the controllable pressurization, i.e., right before the heat treatment. When the gas diffusion dynamics is Fickian, the main contribution of this article is to show that any specified concentration profile can be achieved provided a sufficiently large relatively-flat offset of the concentration profile is acceptable. This is demonstrated by showing that the interior gas concentration is controllable by changing the gas concentration at the boundary, and the use of an offset ensures that the control input (the outside concentration) remains positive. The controllability, in turn, allows the use of standard optimal control to achieve the state transition to the desired interior gas concentration before the heat treatment. A model-based simulation is used to illustrate the proposed approach.

Commentary by Dr. Valentin Fuster
2016;():V002T18A003. doi:10.1115/DSCC2016-9698.

We consider the problem of dynamic coupling between the rapid thermal solidification and mechanical compression of steel in twin-roll steel strip casting. In traditional steel casting, molten steel is first solidified into thick slabs and then compressed via a series of rollers to create thin sheets of steel. In twin-roll casting, these two processes are combined, thereby making control of the overall system significantly more challenging. Therefore, a simple and accurate model that characterizes these coupled dynamics is needed for model-based control of the system. We model the solidification process with explicit consideration for the mushy (semi-solid) region of steel by using a lumped parameter moving boundary approach. The moving boundaries are also used to estimate the size and composition of the region of steel that must be compressed to maintain a uniform strip thickness. A novelty of the proposed model is the use of a stiffening spring to characterize the stiffness of the resultant strip as a function of the relative amount of mushy and solid steel inside the compression region. In turn this model is used to determine the force required to carry out the compression. Simulation results demonstrate key features of the overall model.

Commentary by Dr. Valentin Fuster
2016;():V002T18A004. doi:10.1115/DSCC2016-9770.

This paper studies possible robust control design methods in triple-stage actuation settings for achieving minimum position error signal (PES) while maintaining enough stability margins. Firstly, the sensitivity-decoupling design technique, is utilized to estimate the resulting increase in low frequency disturbance attenuation and servo bandwidth. A systematic tuning methodology based on μ-synthesis is then proposed for track-following servo design of triple-stage actuation systems. In this approach, the objective is to minimize the PES, by considering all constraints and uncertainties explicitly in the design. We describe a step by step Multi-Input Single-Output (MISO) controller design methodology which includes system modeling, noise characterization, control objective determination and controller synthesis and verification. In this methodology, servo bandwidth is not the only performance metric. Rather, the control objective will be to minimize the closed-loop system H norm directly, while all stroke and control constraints are satisfied and enough stability margin is ensured. The proposed method is applied to design track-following feedback controllers for single-, dual- and triple-stage actuation systems. Simulation results show that compared to dual-stage actuation, triple-stage actuation enhances low frequency disturbance rejection by 6 dB at around 100Hz and increases servo bandwidth from ∼3kHz to ∼5kHz.

Commentary by Dr. Valentin Fuster
2016;():V002T18A005. doi:10.1115/DSCC2016-9777.

Additive Manufacturing (AM) is a growing class of manufacturing processes where parts are fabricated by repeated addition of material. Many of these processes show great promise for the production of complex, functional parts for use in critical applications. One such process, Laser Metal Deposition (LMD), uses a laser and a coaxial blown metal powder source to produce functional metal parts. However, it has been demonstrated that the LMD process possesses complex two-dimensional dynamics which, when not appropriately accounted for in the modeling and control stages, can lead to build failures. Additionally, even when the two-dimensionality of the process is accounted for, modeling and process uncertainties can lead to degraded performance or instability. Here, in the context of a control oriented model of the LMD process developed previously, process and modeling uncertainties are modeled and quantified in the frequency domain.

Commentary by Dr. Valentin Fuster
2016;():V002T18A006. doi:10.1115/DSCC2016-9831.

Track settling control for a hard disk drive with three actuators has been considered. The objective is to settle the read/write head on a specific track by following the minimum jerk trajectory. Robust output feedback model predictive control methodology has been utilized for the control design which can satisfy actuator constraints in the presence of noises and disturbances in the system. The controller is designed based on a low order model of the system and has been applied to a higher order plant in order to consider the model mismatch at high frequencies. Since the settling control generally requires a relatively low frequency control input, simulation result shows that the head can be settled on the desired track with 10 percent of track pitch accuracy while satisfying actuator constraints.

Commentary by Dr. Valentin Fuster

Modeling and Control of Automotive Systems

2016;():V002T19A001. doi:10.1115/DSCC2016-9780.

Constrained optimization control techniques with preview are designed in this paper to derive optimal velocity trajectories in longitudinal vehicle following mode, while ensuring that the gap from the lead vehicle is both safe and short enough to prevent cut-ins from other lanes. The lead vehicle associated with the Federal Test Procedures (FTP) [1] is used as an example of the achieved benefits with such controlled velocity trajectories of the following vehicle. Fuel Consumption (FC) is indirectly minimized by minimizing the accelerations and decelerations as the autonomous vehicle follows the hypothetical lead. Implementing the cost function in offline Dynamic Programming (DP) with full drive cycle preview showed up to a 17% increase in Fuel Economy (FE). Real time implementation with Model Predictive Control (MPC) showed improvements in FE, proportional to the prediction horizon. Specifically, 20s preview MPC was able to match the DP results. A minimum of 1.5s preview of the lead vehicle velocity with velocity tracking of the lead was required to obtain an increase in FE.

The optimal velocity trajectory found from these algorithms exceeded the presently allowable error from standard drive cycles for FC testing. However, the trajectory was still safe and acceptable from the perspective of traffic flow. Based on our results, regulators need to consider relaxing the constant velocity error margins around the standard velocity trajectories dictated by the FTP to encourage FE increase in autonomous driving.

Commentary by Dr. Valentin Fuster
2016;():V002T19A002. doi:10.1115/DSCC2016-9792.

Control-oriented models of automotive turbocharger compressors typically describe the compressor power assumming an isentropic thermodynamic process with a fixed isentropic efficiency and a fixed mechanical efficiency for power transmission between the turbine and compressor. Although these simplifications make the control model tractable, they also introduce additional errors due to unmodeled dynamics, especially when the turbocharger is operated outside its normal operational region. This is also true for map-based approaches, since these supplier-provided maps tend to be sparse or incomplete at the boundary operational regions and often ad-hoc extrapolation is required, leading to large modeling error. Furthermore, these compressor maps are obtained from the steady flow bench tests, which introduce additional errors under pulsating flow conditions in the context of internal combustion engines. In this paper, a physics-based model of compressor power is developed using Euler equations for turbomachinery, where the mass flow rate and compressor rotational speed are used as model inputs. Two new coefficients, speed and power coefficients are defined. This makes it possible to directly estimate the compressor power over the entire compressor operating range based on a single analytic relationship. The proposed modeling approach is validated against test data from standard turbocharger flow bench, steady state engine dynamometer as well as transient simulation tests. The validation results show that the proposed model has adequate accuracy for model-based control design and also reduces the dimension of the parameter space typically needed to model the compressor dynamics.

Commentary by Dr. Valentin Fuster
2016;():V002T19A003. doi:10.1115/DSCC2016-9801.

This paper discusses the control challenges of a parallel evaporator organic Rankine cycle (ORC) waste heat recovery (WHR) system for a diesel engine. A nonlinear model predictive control (NMPC) is proposed to regulate the mixed working fluid outlet temperature of both evaporators, ensuring efficient and safe ORC system operation. The NMPC is designed using a reduced order control model of the moving boundary heat exchanger system. In the NMPC formulation, the temperature difference between evaporator outlets is penalized so that the mixed temperature can be controlled smoothly without exceeding maximum or minimum working fluid temperature limits in either evaporator. The NMPC performance is demonstrated in simulation over an experimentally validated, high fidelity, physics based ORC plant model. NMPC performance is further validated through comparison with a classical PID control for selected high load and low load engine operating conditions. Compared to PID control, NMPC provides significantly improved performance in terms of control response time, overshoot, and temperature regulation.

Commentary by Dr. Valentin Fuster
2016;():V002T19A004. doi:10.1115/DSCC2016-9819.

This paper studies the use of an electric turbogenerator (ETG) for waste energy recovery from the exhaust gas of a 13 L Heavy Duty Diesel (HDD) engine. Up to 1% brake specific fuel consumption (BSFC) reduction is predicted for this system at high engine loads using a validated mean value engine model. However, the addition of the ETG reduces the air-fuel equivalence ratio (λ) and increases exhaust gas recirculation (EGR) rate by 10%, deteriorating the engine-out smoke emissions. This challenge is addressed by decreasing the EGR valve position and the asymmetry in the twin scroll turbine. With these modifications, the predicted high load BSFC reduction is 2% and the EGR and λ approach their original values.

The HDD engine is then tested experimentally with the ETG emulated by a valve downstream of the main turbocharger. The experimental results confirm the simulation predictions with the stock engine calibration and geometry, where EGR valve sweeps show the potential of this actuator for remedying the detrimental ETG backpressure effects, which ultimately improves the combined engine and ETG BSFC by 0.6% at high loads. Combining the simulated turbo sizing and the experimental EGR valve results indicates that up to 1.6% BSFC reductions are possible for the HDD engine with an integrated ETG, without deteriorating emission levels. Finally, simulations show that during a torque step the ETG should be bypassed to avoid deterioration in the dynamic response of the engine.

Commentary by Dr. Valentin Fuster
2016;():V002T19A005. doi:10.1115/DSCC2016-9856.

In this paper, we discuss the development of a control-oriented model for the power developed by a Variable Geometry Turbine (VGT). The turbine exit flow velocity, Cex, is obtained based on a polytropic process assumption for the full turbine stage. The rotor inlet velocity, Cin, is estimated, through an empirical relationship between Cex and Cin as a function of a dimensionless parameter ψ. The turbine power is developed based on Euler’s equations of Turbomachinery under the assumptions of zero exit swirl and alignment between the nozzle orientation and the Cin velocity vector. A power loss sub-model is also designed to account for the transmission loss associated with the power transfer between the turbine and compressor. The loss model is an empirical model and accounts for bearing friction and windage losses. Model validation results, for both steady state and transient operation, are shown.

Commentary by Dr. Valentin Fuster

Modeling and Control of Combustion Engines

2016;():V002T20A001. doi:10.1115/DSCC2016-9676.

One potential method to reduce fuel consumption in diesel engines with variable geometry turbines (VGT) and exhaust gas recirculation (EGR) is to reduce the transient engine pumping work through improved EGR-VGT control. Numerical dynamic programming is applied to investigate optimal EGR-VGT control policies for reduced pumping work on a three-state model of a 6.7-liter medium-duty diesel engine. Optimality is defined by a multi-objective cost function that penalizes pumping work, EGR rate control error, and boost pressure control error. Multiple dynamic programs, each with a different set of cost function weights, are performed over an acceleration in the Heavy-Duty Federal Test Procedure cycle to generate the optimal trade-off between the stated objectives. Additionally, a production-representative EGR-VGT controller is simulated, and the resulting suboptimal performance is compared to the optimal frontier to establish the potential fuel consumption benefit of improved EGR-VGT control.

Commentary by Dr. Valentin Fuster
2016;():V002T20A002. doi:10.1115/DSCC2016-9690.

Turbocharging and downsizing (TRBDS) a gasoline direct injection (GDI) engine can reduce fuel consumption but with increased drivability challenges compared to larger displacement engines. This tradeoff between efficiency and drivability is influenced by the throttle-wastegate control strategy. A more severe tradeoff between efficiency and drivability is shown with the introduction of Low-Pressure Exhaust Gas Recirculation (LP-EGR). This paper investigates and quantifies these trade-offs by designing and implementing in a one-dimensional (1D) engine simulation two prototypical throttle-wastegate strategies that bound the achievable engine performance with respect to efficiency and torque response. Specifically, a closed-wastegate (WGC) strategy for the fastest achievable response and a throttle-wastegate strategy that minimizes engine backp-pressure (MBWG) for the best fuel efficiency, are evaluated and compared based on closed loop response. The simulation of an aggressive tip-in (the driver’s request for torque increase) shows that the wastegate strategy can negotiate a 0.8% efficiency gain at the expense of 160 ms slower torque response both with and without LP-EGR. The LP-EGR strategy, however offers a substantial 5% efficiency improvement followed by an undesirable 1 second increase in torque time response, clarifying the opportunities and challenges associated with LP-EGR.

Commentary by Dr. Valentin Fuster
2016;():V002T20A003. doi:10.1115/DSCC2016-9691.

This paper proposes a novel master-slave control strategy for coordination of throttle, wastegate and supercharger actuators in an electrically twincharged engine in order to guarantee efficient boost control during transients, while at steady state a throttle-wastegate coordination provides minimum engine backpressure hence engine efficiency elevation. The benefits and challenges associated with Low Pressure Exhaust Gas Recirculation (LP-EGR) in a baseline turbocharged engine, including improved engine efficiency, mainly due to better combustion phasing, and sluggish engine response to a torque demand due to slowed down air path dynamics were studied and quantified in [1]. Hence in this paper an electrical Eaton TVS roots type supercharger at high pressure side of the turbocharger compressor (TC compressor) is added to the baseline turbocharged engine and the performance of the proposed controller in the presence of LP-EGR, which is a more demanding condition, is evaluated and compared to the turbocharged engine. One dimensional (1D) crankangle resolved engine simulations show that the proposed master-slave control strategy can effectively improve the transient response of the twincharged engine, making it comparable to naturally aspirated engines, while the consumed electrical energy during transients can be recovered from the decreased fuel consumption due to LP-EGR conditions at steady state in approximately 1 second. Finally, a simple controller is developed to bypass the TC compressor and maximize the engine feeding charge during the transients in order to avoid TC compressor choking and achieve faster response.

Commentary by Dr. Valentin Fuster
2016;():V002T20A004. doi:10.1115/DSCC2016-9696.

Reactivity controlled compression ignition (RCCI) is an advanced low temperature combustion strategy introduced to achieve near-zero NOx and soot emissions while maintaining diesel-like efficiencies. Precise control of RCCI combustion phasing is necessary in realizing high fuel conversion efficiency as well as meeting stringent emission standards. Model-based control of combustion phasing provides a powerful tool for real-time control during transient operation of the RCCI engine, which requires a computationally efficient combustion model that encompasses factors such as, injection timings, fuel blend composition and reactivity. In this work, physics-based models are developed to predict the combustion phasing of a 1.9-liter RCCI engine. A mean value control-oriented model (COM) of RCCI is developed by combining the auto-ignition model, the burn duration model, and a Wiebe function to predict combustion phasing. Development of a model-based controller requires a dynamic model which can predict engine operation, i.e., estimation of combustion phasing, on a cycle-to-cycle basis. Hence, the mean-value model is extended to encompass the full-cycle engine operation by including the expansion and exhaust strokes. In addition, the dynamics stemming from the thermal coupling between cycles are accounted for, that results in a dynamic RCCI control-oriented model capable of predicting the transient operation of the engine. This model is then simplified and linearized in order to develop a linear observer-based feedback controller to control the combustion phasing using the premixed ratio (the ratio of the port injected gasoline fuel to the total gasoline/diesel fuel injected). The designed controller depicts an accurate tracking performance of the desired combustion phasing and successfully rejects external disturbances in engine operating conditions.

Commentary by Dr. Valentin Fuster
2016;():V002T20A005. doi:10.1115/DSCC2016-9726.

Previously, the authors have proposed a novel combustion control enabled by the free piston engine (FPE), e.g. the piston trajectory-based HCCI combustion control. Extensive simulation results show that, by employing specific piston trajectories, the FPE is able to increase the engine thermal efficiency significantly, and reduces the emissions production simultaneously. However, a systematic approach to designing the optimal piston trajectory, according to variable working conditions and versatile fuel properties, still remains elusive. In this paper, the study of this optimization is presented. First, a control-oriented model, which includes thermodynamics of the in-cylinder gas and chemical kinetics of the utilized fuel, is adapted for the optimization study. The unique phase separation method was also implemented into the presented model to sustain sufficient chemical kinetics information and reduce the computational burden at the same time. Two optimization methods are then proposed in this paper: one is converting the original problem to parameters optimization; the other is transforming it to a constrained nonlinear programming and solving it via the sequential quadratic programming (SQP) method. The corresponding optimization results and detailed discussions are followed, which clearly demonstrate the advantage of the trajectory-based HCCI combustion with regard to FPE output work and NOx emission.

Commentary by Dr. Valentin Fuster
2016;():V002T20A006. doi:10.1115/DSCC2016-9727.

Connected Vehicle (CV) technology, which allows traffic information sharing, and Hybrid Electrical Vehicles (HEV) can be combined to improve vehicle fuel efficiency. However, transient traffic information in CV environment necessitates a fast HEV powertrain optimization for real-time implementation. Model Predictive Control (MPC) with Linearization is proposed, but the computational effort is still prohibitive. The Equivalent Consumption Minimization Strategy (ECMS) and Adaptive-ECMS are proposed to minimize computation time, but unable to guarantee charge-sustaining-operation (CS). Fast analytical result from Pontryagin’s Minimum Principles (PMP) is possible but the input has to be unconstrained. Numerical solutions with Linear Programming (LP) are proposed, but over-simplifications of the cost and constraint functions limit the performance of such methods. In this paper, a nonlinear CS constraint is transformed into linear form with input variable change. With linear input and CS constraints, the problem is solved with Separable Programming by approximating the nonlinear cost with accurate linear piecewise functions which are convex. The piecewise-linear functions introduce new dimensionless variables which are solved as a large-dimension constrained linear problem with efficient LP solvers. Comparable fuel economy with Dynamic Programming (DP) is shown, with maximum fuel savings of 7% and 21.4% over PMP and Rule-Based (RB) optimizations. Simulations with different levels of vehicle speed prediction uncertainties to emulate CV settings are presented.

Commentary by Dr. Valentin Fuster

Modeling and Validation

2016;():V002T21A001. doi:10.1115/DSCC2016-9635.

This paper presents a methodology for automatically generating a scene to be used in high fidelity robotic simulators. Modeling and simulation play an important role in the development and testing of robotic motion planning algorithms. Virtual Robotic Experimentation Platform (V-REP) is a robotic simulator that can be used to test state of the art robotics algorithms in environments called scenes. V-REP contains a remote application programming interface (API) for Matlab that allows for control of the simulation from the external application. Using this functionality, an algorithm was developed to automatically create simulation environments. Given the dimensions of the space, the desired total number of rooms, and a room configuration type, the algorithm organizes the layout of the space into a set of rooms and hallways. Using the remote capabilities provided by the Matlab V-REP API, the scene is opened, each of the models is loaded, and the models are put into the appropriate location. The result is a saved V-REP scene file that can be used for testing of any relevant mobile robotic applications. Ultimately this tool can play an important role in running parametric studies and Monte Carlo simulations to test the performance of various motion planning and coordination algorithms.

Topics: Simulation , Robotics
Commentary by Dr. Valentin Fuster
2016;():V002T21A002. doi:10.1115/DSCC2016-9651.

This paper proposes an energy-based approach for modeling a screw extruder used for 3D printing. This approach was used due to the difficulty in measuring the salient variables associated with regulation of the process state. The control-oriented steady-state model for the screw extruder is based on the reliably available process variables of heater current and screw speed, which constitute the manipulated variables. The controlled variable for this extrusion process is the extrusion rate. This model is based on balancing the energy between the work done by the screw, the heat delivered by the heater at the nozzle, and the enthalpy of the extruded product stream. The fine measurement available is the current commanded by the heater control system to maintain a fixed temperature at the nozzle. An array of thermistors are used as feedback for the temperature profile along the extruder. The screw speed is calibrated for a stepping motor used for conveying the material. This steady state model can then be helpful for developing a dynamic model for a controller capable of accurate flow control based on preview of the extrusion rate but with a simple yet robust hardware requirement.

Commentary by Dr. Valentin Fuster
2016;():V002T21A003. doi:10.1115/DSCC2016-9717.

Hydraulic circuits have been employed by automotive, aerospace, oil & gas and mining industries for an extended amount of time. During the design stage of the hydraulic circuits, the system level dynamic interactions between components need to be analyzed to enhance system performance and reliability. Physics based, low dimensional modeling and simulation is proposed to help system level design and to verify proper operation of hydraulic components. Presented in this paper is physics based modeling process of a hydraulic circuit. Low dimensional model of the hydraulic circuit is used to investigate dynamic interactions between components on MATLAB-Simscape® platform. The existence of hydraulic oscillations due to aging regulator seals is simulated and a method to reduce these oscillations is presented. It is also shown that the inadequate source to supply pilot pressure to hydraulically piloted valves can create similar hydraulic oscillations and reduce system reliability and remaining operational life.

Commentary by Dr. Valentin Fuster
2016;():V002T21A004. doi:10.1115/DSCC2016-9722.

Loop detectors are installed along many of the freeways across the state of California to provide real-time and historical traffic data. These data are used by Caltrans for traffic management operations, such as freeway ramp metering, and to evaluate the performance of freeway corridor traffic management systems. These data are also being used to calibrate traffic flow models and to perform model-based predictions of freeway corridor congestion and traffic throughput performance. However, such detection is prone to contain errors and inconsistencies, which can pose problems in further use of the data, and is also of such large quantities that identification of errors can be tedious. This paper proposes a fault detection algorithm associating loop detector data to the cell transmission model to identify significant errors among such detectors. It discusses how such an algorithm would apply to loop detection along the mainline freeway, as well as extends the algorithm to determine errors along on and off ramp detectors. It also gives a real-life example with appropriate identification of detectors in error.

Commentary by Dr. Valentin Fuster
2016;():V002T21A005. doi:10.1115/DSCC2016-9734.

This paper presents an experimental methodology to test and validate two Maximum Power Point Tracking (MPPT) strategies on variable speed wind turbines. The first technique of this study is an Extremum Seeking (ES) control strategy which does not require any wind turbine model or wind speed measurements. The analysis shows that its convergence can be quite slow in some cases. For this reason, we improve the ES control with a specific inner-loop that speeds up the convergence of the strategy. Additionally a conventional Perturb and Observe (P&O) algorithm is also implemented for comparison purposes. The proposed ES strategy with an additional inner loop controller shows fast tracking capability and high stability under both constant and variable wind speed in simulations and experiments. Both approaches are verified in Matlab simulations and experiments with a lab-scale wind turbine and a fully instrumented wind tunnel at CWRU-CESC.

Commentary by Dr. Valentin Fuster
2016;():V002T21A006. doi:10.1115/DSCC2016-9739.

Cerebral palsy can cause gait impairments in children that require the prescription of passive ankle-foot orthoses. This project aims to develop a pediatric-sized hydraulic active ankle-foot orthosis with computer-controlled stiffness. The orthosis will allow a clinician to investigate a range of AFO stiffnesses while collecting gait performance metrics to determine the optimal stiffness value for the AFO prescription. The ankle-foot orthosis uses hydraulic technology to generate the large required torques in a light, compact package. The preliminary design uses additive manufacturing to further reduce the weight of the manifolds on the medial and lateral sides of the ankle. The simulation prototype of the design illustrated that the orthosis should be capable of generating 91 Nm of ankle torque and a maximum angular velocity of 483 °/sec. The device will be a valuable resource in the clinic, saving time and resources in the AFO prescription process while improving the healthcare of the patient.

Commentary by Dr. Valentin Fuster
2016;():V002T21A007. doi:10.1115/DSCC2016-9773.

Zebrafish is becoming an important animal model in pre-clinical studies for its genetic similarity to humans and ease of use in the laboratory. In recent years, animal experimentation has faced several ethical issues, calling for alternative methods that capitalize on dynamical systems theory. Here, we propose a computational modeling framework to simulate zebrafish swimming in three dimensions (3D) in the form of a coupled system of stochastic differential equations. The model is capable of reproducing the burst-and-coast swimming style of zebrafish, speed modulation, and avoidance of tank boundaries. Model parameters are calibrated on an experimental dataset of zebrafish swimming in 3D and validated by comparing established behavioral measures obtained from both synthetic and experimental data. We show that the model is capable of accurately predicting fish locomotion in terms of the swimming speed and number of entries in different sections of the tank. The proposed model lays the foundations for in-silico experiments of zebrafish neurobehavioral research.

Commentary by Dr. Valentin Fuster
2016;():V002T21A008. doi:10.1115/DSCC2016-9782.

Model-based control design has the ability to meet the strict closed-loop control requirements imposed by the rising performance and efficiency demands on modern engineering systems. While many modeling frameworks develop control-oriented models based on the underlying physics of the system, most are energy domain specific and do not facilitate the integration of models across energy domains or dynamic time-scales. This paper presents a graph-based modeling framework, derived from the conservation of mass and energy, which captures the structure and interconnections in the system. Subsequently, these models can be used in model-based control frameworks for thermal management. This framework is energy-domain independent and inherently captures the exchange of power among different energy domains. A thermal fluid experimental system demonstrates the formulation of the graph-based models and the ability to capture the hydrodynamic and thermodynamic behaviors of a physical system.

Commentary by Dr. Valentin Fuster
2016;():V002T21A009. doi:10.1115/DSCC2016-9795.

In this paper, we investigate the two dimensional fluid-structure interaction problem of the oscillation of a shape-morphing plate in a quiescent, Newtonian, viscous fluid. The plate is considered as a moving wall for the fluid undergoing two concurrent periodic motions: a rigid oscillation along its transverse direction coupled to a shape-morphing deformation to an arc of a circle with prescribed maximum curvature. Differently from studies concerned with passive flexible structures, here, we introduce the prescribed deformation to specifically manipulate vortex-shedding and modulate hydrodynamic forces and energy losses during underwater oscillations. Computational fluid dynamics simulations are performed to evaluate the effect of the prescribed deformation strategy on the added mass and damping effect along with the hydrodynamic power dissipation. We observe that a minimum in the hydrodynamic power dissipation exists for an optimum curvature of the plate. This finding may allow significant power expenditure reduction in underwater vibrating systems where minimization of energy losses or maximization of quality factor are desirable.

Commentary by Dr. Valentin Fuster
2016;():V002T21A010. doi:10.1115/DSCC2016-9855.

In this study, we present a methodology for the assessment of overall performance for vertical axis wind turbines (VAWT) with straight blades. Salient features of our approach include a validated computational fluid dynamics (CFD) model and a hardware-in-the loop (HIL) test-bed. The two-dimensional, time-dependent CFD model uses the k-ε turbulence model and is coupled with the dynamics of the rotor involving friction and generator torques. The power coefficient curve for the rotor is obtained from the CFD simulations by varying the generator torque over time, and then used in the HIL test-bed that consists of an electrical motor, a gearbox, a permanent magnet synchronous generator, and an electronic load. In this setup, the VAWT rotor is mimicked by the electrical motor based on a power coefficient curve obtained from CFD simulations. Effects of the electrical conversion and control design on the overall performance of the VAWT are studied in the HIL setup. Additionally, a simple non-linear control (SNC) algorithm that mimics a model predictive controller and two different adaptations of the maximum power point tracking (MPPT) algorithm with fixed and variable step-sizes are designed and implemented in HIL simulations. According to results, the generator has a profound effect on the overall power output and the efficiency of the turbine; and the SNC and MPPT algorithms perform satisfactorily under step wind conditions.

Commentary by Dr. Valentin Fuster
2016;():V002T21A011. doi:10.1115/DSCC2016-9901.

This paper proposes a novel computationally efficient dynamics modeling approach for downhole well drilling system. The existing drilling modeling methods are either computationally intensive such as those using finite element method (FEM), or weak in fidelity for complex geometry such as those using transfer matrix method (TMM). To take advantage of the benefits of FEM and TMM and avoid their drawbacks, this paper presents a new hybrid method integrating both of the aforementioned modeling approaches, enabled by the unique structural geometry of the drilling system. Numerical simulation results are presented to demonstrate the effectiveness of the proposed hybrid modeling approach.

Commentary by Dr. Valentin Fuster
2016;():V002T21A012. doi:10.1115/DSCC2016-9908.

While advances in technology have greatly improved the process of mass production, producing small batches or one-offs in an efficient manner has remained challenging for the manufacturing industry. Additionally, in both large and small companies, there are often available manufacturing resources that sit idle between projects. In this paper we present a Production as a Service framework for providing manufacturing options to designers of new products based on available manufacturing resources. The designed framework aims to bridge the gap between the theoretical work that has been done on Service Oriented Architectures in manufacturing, and what is required for implementation. An industrial use case is provided as an example of the framework.

Commentary by Dr. Valentin Fuster

Motion and Vibration Control Applications

2016;():V002T22A001. doi:10.1115/DSCC2016-9607.

Disturbances during engine starts and stops can propagate to the drive wheels with minimal damping in the power split hybrid transaxle due to the system architecture. One such disturbance is the engine cranking torque from pumping & compression during the motoring phase of engine starts and stops before fueling. This paper proposes a control system to compensate for the engine disturbance and reduce the magnitude of the disturbance on the driveline. A high fidelity model of the power split transaxle is developed and validated with test data for NVH analysis. A simple model of a four cylinder engine is developed that can run in real-time and accurately estimate the cranking torque disturbance from the engine. It is show that the engine speed in the power split transaxle is controlled via a generator speed feedback loop. Performing a linear system analysis reveals that the addition of a feedforward term in the loop which is based on the engine cranking torque can reduce the driveline disturbance. It is shown through simulation that the proposed engine model with the feedforward controller can reduce the engine disturbance on the driveline. The engine model and feedforward controller are finally implemented on the production control module of a test vehicle and it is shown that the proposed control strategy can reduce the driveline disturbance by as much a 65% during engine starts. Moreover, it shown that the engine model can run in real-time and correlates well with in-cylinder pressure data measured from the engine.

Commentary by Dr. Valentin Fuster
2016;():V002T22A002. doi:10.1115/DSCC2016-9678.

Cable Driven Parallel Manipulators (CDPMs) utilize flexible wire to actuate an end-effector, allowing rapid accelerations across large workspaces. CDPMs are predominantly modeled with rigid cables, greatly simplifying the analysis. This model is satisfactory for small, fixed masses traveling short distances. However, as cable length increases, the flexibility of the cables, including the variation in stiffness and damping as length changes, cannot be ignored. In addition, the end-effector, which may be modeled as a pendulum, will rotate and contribute to the motion. This paper presents the modeling and control of a large-scale, cable-driven parallel manipulator, with application to inspection of large workspaces. The multi-degree-of-freedom model developed takes into account flexibility of cables and the oscillatory dynamics of the end effector. The dominant dynamics are identified and used to design a control system to limit vibration.

Commentary by Dr. Valentin Fuster
2016;():V002T22A003. doi:10.1115/DSCC2016-9687.

Vibration control of large structures has been the focus of a lot of research in recent years. Some of these structures include high rise buildings, offshore platforms, and bridges. In this article, we present the results of an experimental investigation of the usage of linear impact dampers in the control of the elasto-dynamic vibrations of 3D structures. Linear Particle Chain Impact Dampers (LPCIDs) are the off-spring of the commonly used conventional (single unit) impact damper. The free vibration response of a 3D symmetric frame subjected to a bidirectional initial condition is measured and analyzed. The objective is to examine the efficacy of the LPCID in attenuating the free vibrations of 3D frame structures. The settling times and amplitudes of vibration of the structure, with and without the LPCIDs, under free vibration conditions are measured and analyzed to study the efficacy of the dampers. The experimental study showed that the LPCID can be more effective in reducing the structure’s vibration when placed in specific orientations on the structure. Therefore, it can by concluded from the experiments’ outcomes that LPCIDs may be used to effectively attenuate the free vibrations of 3D structures.

Commentary by Dr. Valentin Fuster
2016;():V002T22A004. doi:10.1115/DSCC2016-9697.

In this work a robust current control strategy is developed in order to counteract the imbalance forces of a rotor which is levitated by an active magnetic bearing (AMB) system. A model-based and physics based design procedure for active imbalance rejection is presented, which solely derives from the configuration of a given AMB system such as physical parameters, rotational speed and controller capacity. An output feedback controller is developed which eliminates the synchronous vibrations by applying synchronous forces at the AMB force locations. The controller is developed by making use of the sliding mode control (SMC) formalism. Within the boundary layer the controller is equivalent to a MIMO proportional-derivative (PD) controller with an output reference trajectory. The active vibration suppression forces are obtained as a control rule, which implement the tracking of a reference trajectory in the output space. The time-varying reference trajectory is obtained as the particular solution of a first-order linear ODE with periodic forcing function, where the periodic forcing function is the generalized synchronous imbalance excitation forces. For a given AMB system and known unbalance, the achievable rotational speeds such that the AMB system can be operated, are derived. Within the boundary layer, which is the main operation range, the MIMO PD controller is parameterized by the rotational speed.

Topics: Bearings , Rotors
Commentary by Dr. Valentin Fuster
2016;():V002T22A005. doi:10.1115/DSCC2016-9832.

This paper proposes a novel adaptive feedforward control method for rejecting unknown disturbances acting on linear systems with uncertain dynamics. The proposed algorithm does not require a model of the plant dynamics and does not use batches of measurements in the adaptation process. Moreover, it is applicable to both minimum and non-minimum phase plants. The algorithm is a “direct” adaptive method, in the sense that the identification of system parameters and the control design are performed simultaneously. In order to verify the effectiveness of the proposed method, an adaptive feedforward controller is designed as an add-on compensator to the existing baseline controller of a hard disk drive. An accelerometer mounted on the disk drive casing provides the input signal for the controller. The control objective is to minimize the standard deviation of the position error signal in the presence of external random vibrations. Simulation results show that reduction of 46% in the standard deviation of the position error signal can be obtained.

Commentary by Dr. Valentin Fuster
2016;():V002T22A006. doi:10.1115/DSCC2016-9837.

Vibration with multiple large peaks at high frequencies may cause significant performance degradation and have become a major concern in modern high precision control systems. To deal with such high-frequency peaks, it is proposed to design a frequency-shaped sliding mode controller based on H synthesis. It obtains an ‘optimal’ filter to shape the sliding surface, and thus provides frequency-dependent control allocation. The proposed frequency-shaping method assures the stability in the presence of multiple-peak vibration sources, and minimizes the weighted H norm of the sliding surface dynamics. The evaluation is performed on a simulated hard disk drive with actual vibration sources from experiments, and the effectiveness of large vibration peak suppression is demonstrated.

Commentary by Dr. Valentin Fuster
2016;():V002T22A007. doi:10.1115/DSCC2016-9839.

The secondary actuator in dual-stage hard disk drives (HDDs) is subject to both amplitude and rate saturations. This paper proposes an anti-windup scheme based on robust control methodologies by modeling the saturations as bounded nonlinear uncertainties. The scheme is constructed as an add-on linear conditioning structure which consists of a dual-input-dual-output filter. With the proposed scheme, the induced l2 norm from the disturbance to the position error is minimized with guaranteed stability. Loop transformations to reduce redundant uncertainties and weighting functions to accommodate different performances are discussed. Simulation validation on a dual-stage HDD system is performed and the effectiveness of the proposed anti-windup scheme is demonstrated.

Commentary by Dr. Valentin Fuster
2016;():V002T22A008. doi:10.1115/DSCC2016-9873.

Disturbance observer based (DOB) control has been implemented in motion control to reject unknown or time-varying disturbances. In this research, an internal model-based disturbance observer (DOB) design combined with a PID type feedback controller is formulated for wind turbine speed and power regulation. The DOB controller facilitates model-based estimation and cancellation of disturbance using an inner feedback control loop. The disturbance observer combined with a compensator is further designed to deal with the model mismatch. The proposed method is applied to National Renewable Energy laboratory (NREL) offshore 5-MW wind turbine. Our case studies show that the DOB controller can achieve improved speed and power regulation compared to the baseline PID controller, and exhibit excellent robustness under different turbulent wind fields.

Topics: Design , Wind turbines
Commentary by Dr. Valentin Fuster
2016;():V002T22A009. doi:10.1115/DSCC2016-9898.

Single-point metal turning processes can create chip nests that are hazards to both parts and machine tools. This is mitigated by a process called Modulated Tool Path (MTP) machining, which superimposes an oscillation in the tool tip feed direction in order to break these chips and provide an adequate surface finish. MTP machining is highly sensitive to the amplitude and frequency of this oscillation, both of which can often be diminished by standard machine tool controllers. These controllers are also unresponsive to iteration-varying disturbances such as temperature fluctuations, which can cause positional and velocity-related inaccuracies. This paper presents a library-based variant of Iterative Learning Control (ILC) called Disturbance and Performance-Weighted ILC (DPW-ILC), which is designed to improve the accuracy of machine tool trajectories that are highly oscillatory in nature, as well as provide robustness to varying, but measurable disturbances. DPW-ILC has been shown in simulation to provide a tremendous accuracy benefit over standard ILC techniques, specifically in the presence of two separate types of temperature-based disturbances.

Commentary by Dr. Valentin Fuster
2016;():V002T22A010. doi:10.1115/DSCC2016-9902.

Iterative learning control (ILC) is an effective technique to improve the tracking performance of systems through adjusting the feedforward control signal based on the memory data. It is critically important to design the learning filters in the ILC algorithm that assures the robust stability of the convergence of tracking errors from one iteration to next. The design procedure usually involves lots of tuning work especially in high-order ILC. To facilitate this procedure, this paper proposes an approach to design learning filters for an arbitrary-order ILC with guaranteed convergence and ease of tuning. The filter design problem is formulated into an H optimal control problem. This approach is based on an infinite impulse response (IIR) system and conducted directly in iteration-frequency domain. Important characteristics of the proposed approach are explored and demonstrated on a simulated wafer scanning system.

Commentary by Dr. Valentin Fuster
2016;():V002T22A011. doi:10.1115/DSCC2016-9923.

Through modeling and experimentation, we analyze common gaits on a waveboard, an underactuated mechanical system whose motion is governed by both nonholonomic constraints and momentum conservation. We take advantage of the system’s symmetries to derive a reduced system model that differentiates between kinematic and dynamic components of motion. We evaluate this model using marker trajectory data gathered through an optical tracking system for various types of gaits. By extracting relevant trajectory parameters via state reconstruction and fitting our joint variables to an ellipse, we determine the kinematic components of gaits commonly used by human riders. In particular, we demonstrate that traditional forward motion is purely dynamic, while sustained turning motion contains kinematic components. In order to validate our model, we compare experimentally obtained trajectories with reconstructed displacements based on the model. Finally, we suggest an approach for further analysis of the dynamic components of these gaits.

Commentary by Dr. Valentin Fuster

Multi-Agent and Networked Systems

2016;():V002T23A001. doi:10.1115/DSCC2016-9616.

In this paper, we present a leader-follower type solution to the formation maneuvering problem for multiple, nonholonomic wheeled mobile robots. Our solution is based on the graph that models the coordination among the robots being a spanning tree. Our kinematic control law ensures, in the least squares sense, that the robots globally exponentially acquire a given planar formation while the formation globally exponentially tracks a desired trajectory. The proposed control is demonstrated by numerical simulations of five unicycle vehicles.

Commentary by Dr. Valentin Fuster
2016;():V002T23A002. doi:10.1115/DSCC2016-9693.

This paper presents implementation details and performance metrics for software developed to connect the Robot Operating System (ROS) with Simulink Real-Time (SLRT). The communication takes place through the User Datagram Protocol (UDP) which allows for fast transmission of large amounts of data between the two systems. We use SLRT’s built-in UDP communication and binary packing blocks to send and receive the data over a network. We use implementation metrics from several examples to illustrate the effectiveness and drawbacks of this bridge in a real-time environment. The time latency of the bridge is analyzed by performing loop-back tests and obtaining the statistics of the time delay. A proof of concept experiment is presented that utilizes two laboratories that ran a driver-in-the-loop system despite a large physical separation. This work provides recommendations for implementing data integrity measures as well as the potential to use the system with other applications that demand high speed real-time communication.

Commentary by Dr. Valentin Fuster
2016;():V002T23A003. doi:10.1115/DSCC2016-9733.

We consider the problem of output feedback exponentially stabilizing the 1-D bilayer Saint-Venant model, which is a coupled system of two rightward and two leftward convecting transport partial differential equations (PDEs). The PDE backstepping control method is employed. Our designed output feedback controller is based on the observer built in this paper and the state feedback controller designed in [1], where the backstepping control design idea can also be referred to [2] and can be treated as a generalization of the result for the system with constant system coefficients [2] to the one with spatially-varying coefficients. Numerical simulations of the bilayer Saint-Venant problem are also provided to verify the result.

Topics: Feedback
Commentary by Dr. Valentin Fuster
2016;():V002T23A004. doi:10.1115/DSCC2016-9752.

This paper considers the problem of deploying an arbitrary multi-agent system in a desired formation over an n-dimensional motion space. Each agent is considered to be a ball and collision avoidance is addressed. System evolution in ℝn is decomposed into n one dimensional motion problems, where evolution of the agents qth (q = 1, 2, 3) components are independently guided by two q-leaders. The remaining agents are considered q-followers, updating the qth component of their positions by local interactions with two neighboring q-agents. Communications among the q-agents are weighted by values consistent with the qth position components of agents in the desired configuration. This paper shows how specifying certain constraints on q-leader motion can address the problem of inter-agent collision avoidance when followers acquire their desired positions only by local communication.

Commentary by Dr. Valentin Fuster
2016;():V002T23A005. doi:10.1115/DSCC2016-9783.

This paper addresses a three-dimensional (3D) leader-follower formation control problem where an event-triggered transmission scheme is developed to schedule the information exchange between the leader and follower unmanned aerial vehicles (UAVs). A novel 3D model based on a local level spherical frame is developed to characterize the relative dynamics of the leader and follower UAVs. Based on this 3D model, a novel non-linear tracking control law is developed to achieve an exponential tracking performance of the system. An event-triggered communication scheduling scheme is adopted under the proposed 3D leader-follower framework in order to achieve an efficient use of communication bandwidth by adapting the transmission time for the changes on UAV states. The stability of the formation control law and the efficiency of the event-triggered method are verified and demonstrated in simulation.

Commentary by Dr. Valentin Fuster
2016;():V002T23A006. doi:10.1115/DSCC2016-9789.

This paper deals with the problem of designing a distributed fault diagnosis and estimation algorithm for multi-robot systems that are subject to faults in the form of abrupt velocity biases. To solve this problem, the multi-robot collective is converted to a network of interconnected diagnostic nodes (DNs) that is deployed to monitor the health of the system. Each node consists of a reduced-order estimator with relative state measurements and an online parameter learning filter. The local estimator executes a distributed variation of the particle filtering algorithm using the local sensor measurements and the fault progression model of the robots. The parameter learning filter is used to obtain an approximation of the severity of faults. Numerical simulations demonstrate the efficiency of the proposed approach.

Commentary by Dr. Valentin Fuster
2016;():V002T23A007. doi:10.1115/DSCC2016-9820.

This paper presents deployment strategies for a team of multiple mobile robots with line-of-sight visibility in 1.5D and 2.5D terrain environments. Our objective is to distributively achieve full visibility of a polyhedral environment. In the 1.5D polyhedral terrain, we achieve this by determining a set of locations that the robots can distributively occupy. In the 2.5D polyhedral terrain, we achieve full visibility by simultaneously exploring, coloring, and guarding the environment.

Topics: Robotics , Teams
Commentary by Dr. Valentin Fuster
2016;():V002T23A008. doi:10.1115/DSCC2016-9834.

This paper proposes a method for optimal selection of number of agents in a cooperative multi-agent system for dynamic aerial transportation of a cable-suspended load along a desired trajectory. We consider the system of n planar quadrotors with a cable suspended load, show that this system is differentially flat, and study the effect of varying the number of quadrotors, load and quadrotor masses, on the maximum velocity of the load trajectory and minimum thrust required. We determine the minimum (maximum) number of agents required for traversing a load trajectory for various values of the load mass.

Topics: Cables
Commentary by Dr. Valentin Fuster
2016;():V002T23A009. doi:10.1115/DSCC2016-9917.

Localization and communication are both essential functionalities of any practical mobile sensor network. Achieving both capabilities through a single Simultaneous Localization And Communication (SLAC) would greatly reduce the complexity of system implementation. In this paper a technique for localizing a mobile agent using the line of sight (LOS) detection of an LED-based optical communication system is proposed. Specifically, in a two-dimensional (2D) setting, the lines of sight between a mobile robot and two base nodes enable the latter to acquire bearing information of the robot and compute its location. However, due to the mobile nature of the robot, establishing its LOS with the base nodes would require extensive scan for all parties, severely limiting the temporal resolution and spatial precision of the localization. We propose the use of a Kalman filter to predict the position of the robot based on past localization results, which allows the nodes to significantly reduce the search range in establishing LOS. Simulation results and preliminary experimental results are presented to illustrate and support the proposed approach.

Commentary by Dr. Valentin Fuster

Path Planning and Motion Control

2016;():V002T24A001. doi:10.1115/DSCC2016-9644.

In this paper, a collision avoidance system is proposed to steer away from a leading target vehicle and other surrounding obstacles. A virtual target lane is generated based on an object map resulted from perception module. The virtual target lane is used by a path planning algorithm for an evasive steering maneuver. A geometric method which is computationally fast for real-time implementations is employed. The algorithm is tested in real-time and the simulation results suggest the effectiveness of the system in avoiding collision with not only the leading target vehicle but also other surrounding obstacles.

Commentary by Dr. Valentin Fuster
2016;():V002T24A002. doi:10.1115/DSCC2016-9645.

In this paper, we present a novel bi-directional cooperative obstacle avoidance system of heterogeneous unmanned vehicles, consisting of an unmanned ground vehicle (UGV) and a microaerial vehicle (MAV), equipped with cameras, operating in an indoor environment without Global Positioning System (GPS) signals. The system demonstrates the synergistic relationship between the two platforms by sharing different perspectives and information collected independently using their on-board sensors in performing a navigation task in an indoor environment, including avoiding obstacles and entering narrow pathways. The MAV uses an aerial view of the environment to develop an obstacle free path for the UGV using the A* algorithm. The UGV deploys the planned path in conjunction with information gathered from its own front facing camera to navigate through a cluttered environment using a Lyapunov stable sliding mode controller. The UGV is responsible for detecting low and narrow pathways and to guide the MAV to move through them. The bidirectional cooperation has been tested in hardware as well as in simulation, showing the system’s effectiveness.

Topics: Vehicles
Commentary by Dr. Valentin Fuster
2016;():V002T24A003. doi:10.1115/DSCC2016-9669.

Motivated by area coverage optimization problems with time varying risk densities, we propose a decentralized control law for a team of autonomous mobile agents in a two dimensional area such that their asymptotic configurations optimize a generalized non-autonomous coverage metric. The generalized non-autonomous coverage metric explicitly depends on a nonuniform time-varying measurable scalar field that is not directly controllable by agents. Several interesting scenarios emerge with time varying risk density. In this work, we consider the case of area surveillance against moving targets or external threats penetrating through the perimeter, and the case of environmental monitoring and intervention with deployment of mobile sensors in areas affected by penetration of substances governed by diffusion mechanisms, as for example oil in a marine environment. In the presence of time-varying risk density the coverage metric is non-autonomous as it includes a time varying component that does not depend on the evolution of the agents. Our non-autonomous feedback law accounts for the time-varying component through a term that vanishes when the risk eventually stops evolving. Optimality with respect to the induced non-autonomous coverage is proven in the framework of Barbalat’s lemma, and the performance is illustrated through simulation of the these two scenarios.

Topics: Feedback , Risk
Commentary by Dr. Valentin Fuster
2016;():V002T24A004. doi:10.1115/DSCC2016-9675.

This paper describes control design and stability analysis for a horizontal pendulum using translational control of the pivot. The system is a one-link mechanism subject to linear damping and moving in the horizontal plane. The goal is to drive the system to a desired configuration such that the system oscillates in an arbitrarily small neighborhood of that desired configuration. We consider two cases: prescribed displacement inputs and prescribed force inputs. The proposed control law has two parts, a proportional-derivative part for control of actuated coordinates, and a high-frequency, high-amplitude oscillatory forcing to control the motion of unactuated coordinate. The control system is a high-frequency, time-periodic system. Therefore we use averaging techniques to determine the necessary input amplitudes and control gains. We show that using a certain oscillatory input, the amplitudes of that input must follow a constraint equation. We discuss the geometric interpretation of constraint equation and stability conditions of the system. We also discuss the effects of damping and relative phase of the oscillatory inputs on the system and their physical and geometric interpretation.

Topics: Pendulums
Commentary by Dr. Valentin Fuster
2016;():V002T24A005. doi:10.1115/DSCC2016-9761.

Rapidly-exploring Random Tree (RRT) is a sampling-based algorithm which is designed for path planning problems. It is efficient to handle high-dimensional configuration space (C-space) and nonholonomic constraints. Under the nonholonomic constraints, the RRT can generate paths between an initial state and a goal state while avoiding obstacles. Since this framework assumes that a system is deterministic, more improvement should be added when the method is applied to a system with uncertainty. In robotic systems with motion uncertainty, probability for successful targeting and obstacle avoidance are more suitable measurement than the deterministic distance between the robot system and the target position. In this paper, the probabilistic targeting error is defined as a root-mean-square (RMS) distance between the system to the desired target. The proximity of the obstacle to the system is also defined as an averaged distance of obstacles to the robotic system. Then, we consider a cost function that is a sum of the targeting error and the obstacle proximity. By numerically minimizing the cost, we can obtain the optimal path. In this paper, a method for efficient evaluation and minimization of this cost function is proposed and the proposed method is applied to nonholonomic flexible medical needles for performance tests.

Topics: needles
Commentary by Dr. Valentin Fuster
2016;():V002T24A006. doi:10.1115/DSCC2016-9790.

In this paper, we apply a Markov Decision Process (MDP) to assure collision avoidance and specify optimal paths for evolution of an agent with nonlinear dynamics. We consider evolution of an autonomous wheeled robot in a motion field where a malicious agent has stochastic transitions over the field. By knowing the history of malicious agent transition as well as initial and goal destinations of the autonomous robot, the optimal path is obtained using the Bellman equation. We propose a novel finite time controller that assures reachability of desired way-points along the optimal path, where the optimal path is obtained without deriving the Hamiltonian.

Commentary by Dr. Valentin Fuster
2016;():V002T24A007. doi:10.1115/DSCC2016-9798.

The problem of finding the shortest path for a vehicle visiting a given sequence of target points subject to the motion constraints of the vehicle is an important problem that arises in several monitoring and surveillance applications involving unmanned aerial vehicles. There is currently no algorithm that can find an optimal solution to this problem. Therefore, heuristics that can find approximate solutions with guarantees on the quality of the solutions are useful. The best approximation algorithm currently available for the case when the distance between any two adjacent target points in the sequence is at least equal to twice the minimum radius of the vehicle has a guarantee of 3.04. This article provides a new approximation algorithm which improves this guarantee to Display Formula1+π32.04. The developed algorithm is also implemented for hundreds of typical instances involving at most 30 points to corroborate the performance of the proposed approach.

Commentary by Dr. Valentin Fuster
2016;():V002T24A008. doi:10.1115/DSCC2016-9816.

In this paper, we address the runtime verification problem of robot motion planning with human-in-the-loop. By bringing together approaches from runtime verification, trust model, and symbolic motion planning, we developed a framework which guarantees that a robot is able to safely satisfy task specifications while improving task efficiency by switches between human supervision and autonomous motion planning. A simple robot model in a domain path planning scenario is considered and the robot is assumed to have perfect localization capabilities. The task domain is partitioned into a finite number of identical cells. A trust model based on the robot and human performance is used to provide a switching logic between different modes. Model checking techniques are utilized to generate plans in autonomous motion planning and for this purpose, Linear Temporal Logic (LTL) as a task specification language is employed to formally express specifications in model checking. The whole system is implemented in a runtime verification framework to monitor and verifies the system execution at runtime using ROSRV. Finally, we illustrated the effectiveness of this framework as well as its feasibility through a simulated case study.

Topics: Robot motion
Commentary by Dr. Valentin Fuster
2016;():V002T24A009. doi:10.1115/DSCC2016-9835.

This paper presents an obstacle filtering algorithm that mimics human driver-like grouping of objects within a model predictive control scheme for an autonomous road vehicle. In the algorithm, a time to collision criteria is first used as risk assessment indicator to filter the potentially dangerous obstacle object vehicles in the proximity of the autonomously controlled vehicle. Then, the filtered object vehicles with overlapping elliptical collision areas put into groups. A hyper elliptical boundary is regenerated to define an extended collision area for the group. To minimize conservatism, the parameters for the tightest hyper ellipse are determined by solving an optimization problem. By excluding undesired local minimums for the planning problem, the grouping alleviates limitations that arise from the limited prediction horizons used in the model predictive control. The computational details of the proposed algorithm as well as its performance are illustrated using simulations of an autonomously controlled vehicle in public highway traffic scenarios involving multiple other vehicles.

Commentary by Dr. Valentin Fuster
2016;():V002T24A010. doi:10.1115/DSCC2016-9882.

This paper presents a new approach to simultaneously determine the optimal robot base placement and motion plan for a prescribed set of tasks using expanded Lagrangian homotopy. First, the optimal base placement is formulated as a constrained optimization problem based on manipulability and kinematics of the robot. Then, the constrained optimization problem is expressed into the expanded Lagrangian system and subsequently converted into a homotopy map. Finally, the Newton-Raphson method is used to solve the constrained optimization problem as a continuation problem. The complete formulation for the case of a 6-DOF manipulator is presented and a planar optimal mobile platform motion planning approach is proposed. Numerical simulations confirm that the proposed approach is able to achieve the desired results. The approach also shows the potential for incorporating factors such as joint limits or collision avoidance into the motion planning process as inequality constraints and will be part of future research.

Topics: Robots , Path planning
Commentary by Dr. Valentin Fuster
2016;():V002T24A011. doi:10.1115/DSCC2016-9892.

Through application of Pontryagin’s maximum principle we derive minimum-time optimal trajectories for a “steered particle” agent with constraints on speed and turning rate to reach a point on the plane with free terminal heading. We also present a formulation of the optimal trajectories in the form of a state-feedback control law that is applicable to real time motion planning on a robotic system with these motion constraints.

Commentary by Dr. Valentin Fuster
2016;():V002T24A012. doi:10.1115/DSCC2016-9912.

Diverse problems in robotic locomotion have previously been modeled in terms of connections on principal bundles. Ordinarily, one identifies points in the base manifold of such a bundle with different internal configurations of a robot, identifies points in the fiber over a given base point with different positions and orientations of the robot in its environment, and assumes control to be applied in the former but not the latter. We examine ways in which the ordinary application of this theory may be adapted to problems in aquatic and terrestrial locomotion that fail to accommodate the preceding description.

Topics: Robotics
Commentary by Dr. Valentin Fuster

Robot Manipulators

2016;():V002T25A001. doi:10.1115/DSCC2016-9815.

This paper proposes an efficient greedy algorithm to estimate the object shape within a limited number of sample data (i.e. the touch-down points). Specifically, we treat the object shape as an implicit surface which is defined as the zero level-set of an unknown function and apply Gaussian process to estimate the surface. The mutual information criteria is utilized to decide which point should be sampled in the next iteration. To expedite the estimation process, we implement sequential Gaussian process for computational efficiency and significantly reduce the computational cost by selecting search area based on the estimated level-set variance. We present some simulation results to compare the performance of the proposed algorithm with the random sampling algorithm and demonstrate the improvements in both speed and accuracy over the random sampling algorithm.

Topics: Algorithms
Commentary by Dr. Valentin Fuster
2016;():V002T25A002. doi:10.1115/DSCC2016-9878.

This paper describes a modular 2-DOF serial robot manipulator and accompanying experiments that have been developed to introduce students to the fundamentals of robot control. The robot is designed to be safe and simple to use, and to have just enough complexity (in terms of nonlinear dynamics) that it can be used to showcase and compare the performance of a variety of textbook robot control techniques including computed torque feedforward control, inverse dynamics control, robust sliding-mode control, and adaptive control. These various motion control schemes can be easily implemented in joint space or operational space using a MATLAB/Simulink real-time interface.

By adding a simple 2-DOF force sensor to the end-effector, the robot can also be used to showcase a variety of force control techniques including impedance control, admittance control, and hybrid force/position control. The 2-DOF robots can also be used in pairs to demonstrate control architectures for multi-arm coordination and master/slave teleoperation.

This paper will describe the 2-DOF robot and control hardware/software, illustrate the spectrum of robot control methods that can be implemented, and show sample results from these experiments.

Commentary by Dr. Valentin Fuster
2016;():V002T25A003. doi:10.1115/DSCC2016-9906.

In glass manufacturing industry, glass grinding process has significant involvement of human workers. Human workers need to load and unload glass pieces to/from the grinder. A 6 DOF industrial robot could be used to automate the process by the “pick and place” task. In this paper, a vision system is implemented to robustly detect glass piece location and the placing destination. Two distinct detection methods are used for different glass settings. A “pick and place” trajectory is automatically generated based on the detected locations. A simulation is first performed to visualize robot motion before operating on the robot.

Topics: Glass , Robots
Commentary by Dr. Valentin Fuster
2016;():V002T25A004. doi:10.1115/DSCC2016-9909.

Continuum manipulators are continuously bending robots with unlimited number of kinematic degrees of freedom. Most existing continuum manipulators have a central strut made of a single elastic material, and multiple cables placed around the strut are employed to actuate the manipulator. The kinematics for such robots has been extensively studied by assuming the manipulator has a circular shape. In this paper, we aim to investigate the kinematic and static modeling for novel soft continuum manipulators fabricated by multi-material 3D printing with heterogeneous soft materials. We model cylindrical shape manipulators consisting of three sections with three different materials, and they are actuated by three independent cables placed symmetrically. By pulling the cables with different displacements, the manipulators can be bent in three-dimensional space. As our initial study, we employ a single cable for all prototypes. Experimental results are compared with the simulation results to validate the proposed modeling method.

Commentary by Dr. Valentin Fuster
2016;():V002T25A005. doi:10.1115/DSCC2016-9914.

Soft robotic actuators may provide the means to develop a soft robotic catheter, enabling safer and more effective transcatheter procedures. In many clinical applications, device contact force affects the quality of diagnostic or the degree of therapy delivered. Therefore precise end effector force control will be a requirement for the soft robotic catheter. In this study a bending soft actuator system was fabricated, and the relationship between volume input and end effector contact force is examined. Static and dynamic system identification were conducted under two different loading conditions loosely related to actuation in a blood vessel. The experimental data from these tests led to the creation of a non-linear system model. A reduced term model was developed using a Root Mean Square Error (RMSE) method in order to observe the importance of system dynamics and nonlinearities. A different system model was designed for each loading condition. These two reduced models matched with experimental result, but differed in model terms and parameters, suggesting that either loading condition identification or end effector closed-loop sensing will be needed for accurate contact force control of a soft robotic actuator in an intravascular environment.

Commentary by Dr. Valentin Fuster

Sensors and Actuators

2016;():V002T26A001. doi:10.1115/DSCC2016-9663.

Temperature monitoring is essential in automation, mechatronics, robotics and other dynamic systems. Wireless methods which can sense multiple temperatures at the same time without the use of cables or slip-rings can enable many new applications. A novel method utilizing small permanent magnets is presented for wirelessly measuring the temperature of multiple points moving in repeatable motions. The technique utilizes linear least squares inversion to separate the magnetic field contributions of each magnet as it changes temperature. The experimental setup and calibration methods are discussed. Initial experiments show that temperatures from 5 to 50 °C can be accurately tracked for three neodymium iron boron magnets in a stationary configuration and while traversing in arbitrary, repeatable trajectories. This work presents a new sensing capability that can be extended to tracking multiple temperatures inside opaque vessels, on rotating bearings, within batteries, or at the tip of complex end-effectors.

Commentary by Dr. Valentin Fuster
2016;():V002T26A002. doi:10.1115/DSCC2016-9700.

This paper discusses the development of the attitude estimation algorithm for a MEMS based 9-axis motion tracking sensor, which includes a tri-axis accelerometer, a tri-axis gyroscope and a tri-axis magnetometer. The comparison between the Euler angles and the direction cosine matrix (DCM) based approach is presented to illustrate the advantage of DCM. It will be shown that the kinematic model for DCM can be transformed into a linear time-varying state space form, which greatly simplifies the development of the estimation algorithm. Different from the existing estimation algorithms, which incorporate a nonlinear kinematic model and the nonlinear Kalman filter, such as extended Kalman filter (EKF) or unscented Kalman filter (UKF), the non-linearity in the kinematic model is not the trouble maker anymore. Hence, global convergence can always be guaranteed. Finally, the estimation algorithm is demonstrated by using the real measurement data collected from InvenSenses MPU9250, which is one of the most popular 9-axis motion tracking sensors in the market.

Commentary by Dr. Valentin Fuster
2016;():V002T26A003. doi:10.1115/DSCC2016-9708.

To develop the next generation of high-performance robots capable of working in human environments, it is required that the joint actuators have variable stiffness to achieve both precision motion control and ability of reaction under unexpectedly huge impact caused by collision with obstacles or human. Variable stiffness actuators (VSA) partially realize such objectives by employing an auxiliary input to change the joint stiffness. However, it requires prior information of external load condition. Load sensors or online load estimation techniques need to be implemented to detect sudden unexpected load for stiffness adjustment, adding complexity to the system with bandwidth issues. In this paper, we propose a new design of compliant actuator in which the stiffness automatically varies depending on the unexpected external load. A novel doubly-clamped box structure is used to connect the load inertia to the motor inertia. Specifically, the load inertia is confined inside a box clamped by two stoppers on two opposite sides with two pre-compressed springs. A secondary motor connects to the load inertia through another spring, compensating for known unbalanced forces such as gravity, Coriolis force and inertia force. It is shown that if the unexpected external load force is below the pre-compression force of the springs, the load inertia will be confined exactly within the box and the system behaves like a rigid actuator, otherwise one of the springs will be further compressed and the system behaves like a compliant actuator. Such a mechanical structure has the ability of achieving both precision motion control and automatic reaction under unexpectedly huge external impact, without the need of additional load sensing/estimation. Control algorithms for accurate position tracking under potentially huge unexpected load is developed for this new type of actuator. Simulations are conducted to verify the effectiveness of the design concept and control.

Commentary by Dr. Valentin Fuster
2016;():V002T26A004. doi:10.1115/DSCC2016-9738.

Electric motors are a popular choice for mobile robots because they can provide high peak efficiencies, high speeds, and quiet operation. However, the continuous torque performance of these actuators is thermally limited due to joule heating, which can ultimately cause insulation breakdown. In this work we illustrate how motor housing design and active cooling can be used to significantly improve the ability of the motor to transfer heat to the environment. This can increase continuous torque density and reduce energy consumption. We present a novel housing design for brushless DC motors that provides improved heat transfer. This design achieves a 50% increase in heat transfer over a nominal design. Additionally, forced air or water cooling can be easily added to this configuration. Forced convection increases heat transfer over the nominal design by 79% with forced air and 107% with pumped water. Finally, we show how increased heat transfer reduces power consumption and we demonstrate that strategically spending energy on cooling can provide net energy savings of 4% – 6%.

Commentary by Dr. Valentin Fuster
2016;():V002T26A005. doi:10.1115/DSCC2016-9803.

Micro- and millimeter-scale resonant mass sensors have received widespread research attention due to their robust and highly-sensitive performance in a wide range of detection applications. A key performance metric associated with such systems is the sensitivity of the resonant frequency of a given device to changes in mass, which needs to be calibrated for different sensor designs. This calibration is complicated by the fact that the position of any added mass on a sensor can have an effect on the measured sensitivity, and thus a spatial sensitivity mapping is needed. To date, most approaches for experimental sensitivity characterization are based upon the controlled addition of small masses. These approaches include the direct attachment of microbeads via atomic force microscopy or the selective microelectrodeposition of material, both of which are time consuming and require specialized equipment. This work proposes a method of experimental spatial sensitivity measurement that uses an inkjet system and standard sensor readout methodology to map the spatially-dependent sensitivity of a resonant mass sensor — a significantly easier experimental approach. The methodology is described and demonstrated on a quartz resonator and used to inform practical sensor development.

Topics: Resonance , Sensors
Commentary by Dr. Valentin Fuster
2016;():V002T26A006. doi:10.1115/DSCC2016-9916.

Soft robots made from soft materials can closely emulate biological system using simple soft mechanical structures. Compared with traditional rigid-link robots, they are safe to work with humans and can adapt to confined environments. As a result, they are widely used for various robotic locomotions and manipulations. Nevertheless, for soft robots, being able to sense its state to enable closed-loop control using soft sensors remains a challenge. Existing sensors include external sensors such as camera systems, electromagnetic tracking systems, and internal sensors such as optical fibers, conductive liquid, and carbon black filled strips.

In this paper, we investigate a new soft sensor made from low-cost conductive nylon sewing threads. By continuously inserting twists into a thread under some weight, coils can be formed to enable a coiled soft sensor. The resistance of the sensor varies with the change of length. The fabrication and experiments for this new coiled sensor is described in this paper. Embedding this sensor to a 3D printed soft manipulator demonstrates the sensing capability. Compared to existing soft sensors, the coiled sensor is low-cost, easy to fabricate, and can also be used as an actuator. It can be embedded to any soft robot to measure the deformation for closed-loop feedback control.

Topics: Sensors , Robots
Commentary by Dr. Valentin Fuster

Tracking Control Systems

2016;():V002T27A001. doi:10.1115/DSCC2016-9721.

This paper presents a new design method of a nonlinear feedforward controller for electrohydraulic actuators with asymmetric piston areas. While the use of flatness based inversion of the plant model to design a feedforward controller has been reported for electrohydraulic actuators with symmetric piston area, the extension of this method to actuators with asymmetric piston areas is non-trivial. In asymmetric electrohydraulic actuators, the areas of the hydraulic piston are different in the two chambers, and hence, the amount of fluid going into one chamber of the actuator is not equal to the amount of fluid coming out of the other. This asymmetry leads to loss of flatness, and hence, flatness based inversion of the plant is no longer possible. In this paper, we present a method for calculation of the feedforward control signal for a given trajectory by numerically solving the inverse problem for the system. We demonstrate the effectiveness of the proposed feedforward controller by simulation of trajectory tracking in an asymmetric electrohydraulic actuator. For benchmarking, the tracking performance has been compared with three other feedforward schemes: a linearized model based Zero Phase Error Tracking (ZPET) feedforward controller, a nonlinear feedforward controller implementing an approximate plant inversion based on differential flatness, and a pressure feedback based feedforward controller.

Commentary by Dr. Valentin Fuster
2016;():V002T27A002. doi:10.1115/DSCC2016-9759.

We present an iterative learning control algorithm for accurate task space tracking of kinematically redundant robots with stringent joint position limits and kinematic modeling errors. The iterative learning control update rule is in the task space and consists of adding a correction to the desired end-effector pose based on the tracking error. The new desired end-effector pose is then fed to an inverse kinematics solver that uses the redundancy of the robot to compute feasible joint positions. We discuss the stability, the rate of convergence and the sensitivity to learning gain for our algorithm using quasi-static motion examples. The efficacy of the algorithm is demonstrated on a simulated four link manipulator with joint position limits that learns the modeling error to draw the figure eight in 4 trials.

Commentary by Dr. Valentin Fuster
2016;():V002T27A003. doi:10.1115/DSCC2016-9769.

During the development of next generation tactical aircraft, thermal management is given significant consideration due to higher transient cooling demands, with stricter temperature limits along with the smaller size and weight in the cooling system hardware. Traditional control approaches, such as proportional-integral-derivative (PID), are sufficient to achieve the desired steady-state error performance for a thermal system with no constraints on the control inputs. The traditional control techniques may not be well-suited for thermal systems with constrained inputs. In this paper, we apply the Model Predictive Control (MPC) technique on an input-constrained thermal system and examine the system performance under a large transient thermal load and control input limits through the anticipation of the known thermal load. The results include design and implementation of an MPC controller for a high-fidelity, nonlinear vapor compression cycle model, as well as comparison of the MPC results to those of a finely tuned PID controller.

Commentary by Dr. Valentin Fuster
2016;():V002T27A004. doi:10.1115/DSCC2016-9876.

Exact output tracking requires preview information of the desired output for nonminimum-phase systems. For situations when preview information is not available, this article proposes an output-boundary regulation (OBR) approach that maintains the output-tracking error within prescribed bounds for nonlinear nonminimum-phase systems. OBR transitions the output-tracking error to zero whenever the output error reaches a set magnitude using polynomial output trajectories for each transition. The main contribution is to show that an output-transition-based OBR (O-OBR, which uses post-actuation input to transition the system state after the output-error transition is completed) can enable OBR of more aggressive output trajectories when compared to a state-transition-based OBR (S-OBR) that transitions the full system state and therefore achieves the output transition as well. Results from an example simulation system is used to illustrate the proposed OBR approach and comparatively evaluate the S-OBR and O-OBR approaches, which show that, for the example system, the O-OBR can track 3 times faster desired output trajectory than the S-OBR approach.

Commentary by Dr. Valentin Fuster
2016;():V002T27A005. doi:10.1115/DSCC2016-9890.

During crane operation, the task of retrieval and deployment of payloads can be partitioned into two components: the initial move towards the target or deployment location and the retrieval or deployment of the payload. If the payload is not stationary, as is the case in the retrieval of a sea-going vessel, a third component, tracking, must be included. The target payload in this research is an Autonomous Surface Vehicle (ASV) primarily used for surveying. This paper studies the transition between the initial move towards the payload and the initialization of tracking. Input Shaping is used to limit residual vibration caused by the initial move to the ASV. A set of Fuzzy Logic membership functions are then used to transition from the initial move to the tracking portion of the retrieval process. These membership functions map position and velocity error to a gain that is applied to the tracking controller. As the gain increases, the contribution of the tracking controller input is increased. Zero Phase Error Tracking Control is utilized for accurate tracking of the target payload. Through a combination of these control methods, the tracking accuracy is improved.

Commentary by Dr. Valentin Fuster

Uncertain Systems and Robustness

2016;():V002T28A001. doi:10.1115/DSCC2016-9732.

Analog-to-digital conversion (ADC) and uncertainties in modeling the plant dynamics are the main sources of imprecisions in the design cycle of model-based controllers. These implementation and model uncertainties should be addressed in the early stages of the controller design, otherwise they could lead to failure in the controller performance and consequently increase the time and cost required for completing the controller verification and validation (V&V) with more iterative loops. In this paper, a new control approach is developed based on a nonlinear discrete sliding mode controller (DSMC) formulation to mitigate the ADC imprecisions and model uncertainties. To this end, a DSMC design is developed against implementation imprecisions by incorporating the knowledge of ADC uncertainties on control inputs via an online uncertainty prediction and propagation mechanism. Next, a generic online adaptive law will be derived to compensate for the impact of an unknown parameter in the controller equations that is assumed to represent the model uncertainty. The final proposed controller is an integrated adaptive DSMC with robustness to implementation and model uncertainties that includes (i) an online ADC uncertainty mechanism, and (ii) an online adaptation law. The proposed adaptive control approach is evaluated on a nonlinear automotive engine control problem in real-time using a processor-in-the-loop (PIL) setup with an actual electronic control unit (ECU). The results reveal that the proposed adaptive control technique removes the uncertainty in the model fast, and significantly improves the robustness of the controllers to ADC imprecisions. This provides up to 60% improvement in the performance of the controller under implementation and model uncertainties compared to a baseline DSMC, in which there are no incorporated ADC imprecisions.

Commentary by Dr. Valentin Fuster
2016;():V002T28A002. doi:10.1115/DSCC2016-9758.

This paper focuses on Norm-Optimal Iterative Learning Control (NO-ILC) framework for Single-Input-Single-Output (SISO) Linear Time Invariant (LTI) systems and considers the weighting matrices design problem. The ideal design of weighting matrices should ensure Robust Monotonic Convergence (RMC) against modeling uncertainties while maximizing the convergence speed and minimizing the steady state error. The state-of-art RMC design methodologies either lead to conservative performance or require manual tunings. This paper provides a methodology to systematically achieve an optimal balance between robustness, convergence speed and steady state error. To this end, optimization problems are formulated at each frequency to maximize the convergence speed and minimize the steady state error. Two optimization formulations are proposed: one for an optimal nominal performance and one for an optimal performance against uncertainties. Both formulations offer a systematic approach for designing the weighting matrices for NO-ILC, thereby eliminating the manual tuning process and avoiding an unnecessarily conservative design. A simulation example is given to confirm the analysis and demonstrate the utility of the developed methodologies to design the weighting matrices.

Commentary by Dr. Valentin Fuster
2016;():V002T28A003. doi:10.1115/DSCC2016-9797.

In this work, a procedure is presented for performance analysis of the resilience property of discrete-time systems with perturbed controller and observer gains. The resilience property is defined in terms of both multiplicative and additive perturbations on the gains so that the closed loop eigenvalues do not leave a specified region in the complex plane. In this work, this region is chosen as a disk in the unit circle. Maximum gain perturbation bounds can be obtained based on the designer’s choices of controller eigenvalue region. The linear matrix inequality technique is used throughout the analysis process. Illustrative examples are included to demonstrate the effectiveness of the proposed methodology. The observer counterpart of the results is also provided in an appendix.

Commentary by Dr. Valentin Fuster
2016;():V002T28A004. doi:10.1115/DSCC2016-9885.

This paper proposes a regularized filtered basis functions (RFBF) approach for robust tracking of discrete-time linear time invariant systems with bounded random (unstructured) uncertainties. Identical to the filtered basis functions (FBF) approach, studied in prior work by the authors, the RFBF approach expresses the control trajectory as a linear combination of user-defined basis functions with unknown coefficients. The basis functions are forward filtered using a model of the system and their coefficients are selected to fulfill the tracking control objective. The two approaches differ in the coefficient selection process. The FBF approach selects the coefficients such that the tracking error is minimized in the absence of uncertainties, whereas, the proposed RFBF approach formulates the coefficient selection problem as a constrained game-type problem where the coefficients are selected to minimize the worst case tracking error in the presence of model uncertainty. Illustrative examples are used to demonstrate significantly more accurate tracking of uncertain systems using RFBF compared with FBF.

Commentary by Dr. Valentin Fuster
2016;():V002T28A005. doi:10.1115/DSCC2016-9910.

This paper presents a novel methodology to handle time-varying safety-critical constraints under high level of model uncertainty with application to dynamic bipedal walking with strict foot-placement constraints. This paper builds off recent work on optimal robust control through quadratic programs that can handle stability, input / state dependent constraints, as well as safety-critical constraints, in the presence of high level of model uncertainty. Under the assumption of bounded uncertainty, the proposed controller strictly guarantees time-varying constraints without violating them. We evaluate our proposed control design for achieving dynamic walking of an underactuated bipedal robot subject to (a) torque saturation constraints (input constraints), (b) contact force constraints (state constraints), and (c) precise time-varying footstep placements (time-varying and safety-critical constraints). We present numerical results on RABBIT, a five-link planar bipedal robot, subject to a large unknown load on its torso. Our proposed controller is able to demonstrate walking while strictly enforcing the above constraints with an unknown load of up to 15 Kg (47% of the robot mass).

Topics: Safety
Commentary by Dr. Valentin Fuster

Unmanned, Ground and Surface Robotics

2016;():V002T29A001. doi:10.1115/DSCC2016-9654.

This paper studies the transient performance improvement problem for path following control of underactuated surface vessels (USVs) in the presence of oceanic disturbances. The traditional practice that chooses the tangent direction of the desired path as the desired heading may deteriorate the tracking performance in the curve-path following. That is because the sideslip angle is not zero in turnings, which unavoidably makes the lateral offset hard to converge to zero. Also, the disturbances in wave filed greatly affect the transient control of the path following errors. To this end, this paper makes two contributions: 1) An amendment on the choice of the desired heading is presented to consider the sideslip angle in turnings and then achieve a more accurate path-following maneuver; 2) A novel robust composite nonlinear feedback (CNF) technique is proposed based on a multiple-disturbances observer to improve the transient performance for path following control in seaway environment considering the input saturation. Comparative simulations verify the reasonability of the amendment on the desired heading direction and the effectiveness of the CNF approach in improving the transient performance for the path following control of USVs.

Commentary by Dr. Valentin Fuster
2016;():V002T29A002. doi:10.1115/DSCC2016-9688.

The paper addresses control of variable mass and configuration mechanical systems subjected to holonomic or nonholonomic constraints, which are imposed due to systems desired performance, tracking specified motions or other control needs. The control design is model-based and an analytical dynamics modeling framework underlying controller design is presented. The framework novelty is that constraints, including nonholonomic ones and these on variable mass, can be merged into variable mass system dynamics and final motion equations are free of the constraint reaction forces so they can be used directly to control design. Many mechanical systems change their mass or configuration when they move, e.g. inertia-based propelled underwater vehicles, mobile robots and manipulators transporting loads or space vehicles flying their space missions.

The dynamics modeling framework presented in the paper can be applied to all variable mass system examples mentioned above. An underwater inertia-based propelled vehicle model dynamics and control performance illustrate the theoretical development presented in the paper. The paper contribution is two folded. It presents a unified approach to constrained variable mass or configuration systems modeling and introduces analytical dynamics methods to the nonlinear control domain.

Commentary by Dr. Valentin Fuster
2016;():V002T29A003. doi:10.1115/DSCC2016-9718.

Autonomous and semi-autonomous aerial systems (AES) are often needed to perform tasks in complex and dynamic environments, especially in search and rescue applications. The safe navigation assurance as well as safety assurance of AES are open research issues. This paper investigates modeling of fall-back layer for AES assurance. To realize given advanced requirement the System Safety Surveillance and Control (SSSC) system concept is introduced. To fulfill safety requirements also for software developments formal requirements are formulated, to be realized with the formal modeling technique Strictly Formalized Situation-Operator-Modeling (sf-SOM). Fall-back system integration into AES can achieve system safety by separated safety consideration and emergency behavior integration and realization. Universally concept design permits the fall-back layer realization also for other applications. This in turn allows the first proof of concept of sf-SOM based SSSC system for fall-back layer realization using an experimental example. Here a Three-tank system is used to show the successful fall-back layer realization and the concept transferability to the introduced AES example.

Commentary by Dr. Valentin Fuster
2016;():V002T29A004. doi:10.1115/DSCC2016-9921.

In this paper, a hybrid low-pass and de-trending (HLPD) filtering technique is proposed to achieve robust position estimates using an optical flow based sensor which calculates velocity information at a rate of 400 Hz. In order to filter out the high-frequency oscillation in the velocity information, a standard low-pass filter is implemented. The low-pass filter successfully eliminates sudden jumps and missing data-points, which prevents unprecedented maneuvers and mid-air crashes. The integrated position estimate has the accumulated drift which occurs due to electrical signal and temperature fluctuations together with other environmental factors which affect the data acquisition from the optical flow sensor. A recursive linear least squares fit is performed for the drift model and de-trending is applied to the integrated position signal. The performance of the proposed estimator is validated by comparing with model-identification based weighted average (MI-WA) position estimator, which is commonly used in quadcopters for position estimation. Simulation and experimental flight tests are conducted and the results show that the flight performance of HLPD filter is better than the extensively used MI-WA position filter in hover and square pattern flight tests.

Topics: Filters
Commentary by Dr. Valentin Fuster
2016;():V002T29A005. doi:10.1115/DSCC2016-9929.

The aim of the present work is to design, fabricate, control and experimentally study quadcopters with tilting propellers. Conventional quadcopters, due to limitation in mobility, belong to a class of under-actuated robots which cannot achieve any arbitrary desired state or configuration. The tilting rotor quadcopter provides the added advantage in terms of additional stable configurations that are made possible by additional actuated controls that convert the conventional quadcopter to a fully actuated robot. The tilting rotor quadcopter design is accomplished by using an additional motor for each rotor that enables the rotor to rotate along the axis of the quadcopter arm.

The dynamic model and the controller that have been proposed and reported by the authors of this paper in the previous publications were verified with the help of numerical studies for different flight scenarios. Subsequently, two different models of the vehicle were designed, fabricated and experimentally tested for different flight plans.

Commentary by Dr. Valentin Fuster

Vehicle Dynamic Controls

2016;():V002T30A001. doi:10.1115/DSCC2016-9626.

This paper presents a generalized, multi-body dynamics model for a tracked vehicle equipped with a winch for towing operations. The modeling approach couples existing formulations in the literature for the powertrain components and the vehicle-terrain interaction to provide a comprehensive model that captures the salient features of terrain trafficability. This coupling is essential for making realistic predictions of the vehicle’s mobility capabilities due to the power-load relationship at the engine output. Simulation results are presented jointly with experimental data to validate these dynamics under conditions where no action is taken by the winch. Extended modeling includes dynamics of the hydraulic system that powers the winch so that the limitation of the winch as an actuator and the load it puts on the engine are realized. A second set of simulation results show that for a set of open loop control actions by the winch, the vehicle is able to maintain its mobility in low traction terrain by paying the towed load in and out.

Commentary by Dr. Valentin Fuster
2016;():V002T30A002. doi:10.1115/DSCC2016-9629.

This paper introduces a Hybrid Electric Vehicle (HEV) with eAWD capabilities via the use of a traditional Series-Parallel hybrid transaxle at the front axle and an electric Rear Axle Drive (eRAD) unit at the rear axle. Such a vehicle requires proper wheel torque allocation to the front and rear axles in order to meet the driver demands. A model of the drivetrain is developed using Bond Graphs and is used in co-simulation with a vehicle model from the CarSim software suite for validation purposes. A longitudinal slip ratio control architecture is proposed which allocates slip ratio to the front and real axles via a simple optimization algorithm. The Youla parametrization technique is used to develop robust controllers to track the optimal slip targets generated by the slip ratio optimization algorithm. The proposed control system offers a unified approach to longitudinal vehicle control under both traction and braking events under any road surface condition. It is shown in simulation that the proposed control system can properly allocate slip ratio to the front and rear axles such that tires remain below their force saturation limits while vehicle acceleration/braking is maximized while on a low friction road surface.

Commentary by Dr. Valentin Fuster
2016;():V002T30A003. doi:10.1115/DSCC2016-9646.

In order to achieve higher fuel efficiency and better ride comfort, this paper introduces a shock absorber system including Mechanical-Motion-Rectifier (MMR), power converter and its current/force tracking (ICFT) controller.

MMR based shock absorbers has the benefit of higher efficiency and better mechanical reliability than conventional regenerative shock absorbers. However, the one-way clutches and inertia in MMR induce disengagement between input shaft and generator. This nonlinear behavior makes the input current/force of MMR uncontrollable with conventional feedback controller design. To solve this problem, this paper presents an input current/force tracking (ICFT) controller for MMR based suspension system. By adding additional control laws to the conventional controller, ICFT controller successfully solves the nonlinearity problem during MMR control. This ICFT controller is tested by tracking the reference force from skyhook control to improve ride comfort. The vehicle body displacement is simulated under specified speedbump.

By using this ICFT controller, the simulation result show displacement error between skyhook and ICFT-MMR is within 5% and its total harvested energy is 56 joules, as 56 W of average input power. Equivalent circuits used for circuit simulation are proved to have identical performances as mechanical models.

Commentary by Dr. Valentin Fuster
2016;():V002T30A004. doi:10.1115/DSCC2016-9791.

This paper presents a socially acceptable collision avoidance system for an automated vehicle based on the elastic band method. Both stationary and moving Vulnerable Road Users (VRUs: pedestrians or bicyclists) are considered in the proposed system. A collision free path is first determined and then Model Predictive Control (MPC) based vehicle front wheel steering is applied to track this collision free path. For the purposes of benchmarking and comparison, the results of a conventional PID steering controller are also presented. The designed system is tested with simulations on a path chosen from the west campus of the Ohio State University, whose waypoints are extracted automatically from OpenStreetMap (OSM). Simulation results show that the MPC based steering control system successfully achieves the required collision avoidance and path following and has comparable or better performance when compared with the conventional PID solution.

Commentary by Dr. Valentin Fuster
2016;():V002T30A005. doi:10.1115/DSCC2016-9793.

The cooperation mode between the engagement and disengagement clutches for vehicles equipped with Dual Clutch Transmission (DCT) is of vital importance to achieve a smooth gearshift, in particular for the downshift process as its unavoidable power interruption during the inertia phase. Hence, to elevate the performance of DCT downshifting process, an analytical model and experimental validation for the analysis, simulation and control strategy are presented. Optimized pressure profiles applied on two clutches are obtained based on the detailed analysis of downshifting process. Then, according to the analysis results, a novel control strategy that can achieve downshift task with only one clutch slippage is proposed. The system model is established on Matlab/Simulink platform and used to study the variation of output torque and speed in response to different charging pressure profiles and various external loads during downshifting process. Simulation results show that, compared with conventional control strategies, the proposed one can not only avoid the torque hole and power circulation, but shorten the shift time and reduce the friction work. Furthermore, to validate the effectiveness of the control strategy, the bench test equipped with DCT is conducted and the experiment results show a good agreement with the simulation results.

Commentary by Dr. Valentin Fuster

Vehicle Dynamics and Traffic Control

2016;():V002T31A001. doi:10.1115/DSCC2016-9674.

The determination of vehicle’s center of gravity position is an important but challenging task for control of advanced vehicles such as automated vehicles, especially under daily usage condition where the system configurations and payload condition may change. To address this problem, a new method is proposed in this paper to estimate the vehicle’s 3-dimensional center of gravity position parameters without relying on detailed suspension configuration parameters or lateral tire force models. In the estimation problem, the vehicle’s planar dynamic equations are synthesized together to reduce the number of unknown lateral tire forces, then the condition of Ackermann’s Steering Geometry can be found to eliminate the influence of the remaining unknown front wheel lateral tire forces. When the unknown tire forces are cancelled, the recursive least squares (RLS) regression technique is used to identify the 3-dimensional center of gravity position parameters. The vehicle model with the sprung mass modeled as an inverted pendulum is developed to assist the analysis and conversion of sensor measured signals. Simulations conducted in a high-fidelity CarSim® vehicle model have demonstrated the capability of this proposed method in estimating the vehicle’s center of gravity position parameters.

Commentary by Dr. Valentin Fuster
2016;():V002T31A002. doi:10.1115/DSCC2016-9701.

Currently, there is a lack of low-cost, real-time solutions for accurate autonomous vehicle localization. The fusion of a precise a priori map and a forward-facing camera can provide an alternative low-cost method for achieving centimeter-level localization. This paper analyzes the position and orientation bounds, or region of attraction, with which a real-time vehicle pose estimator can localize using monocular vision and a lane marker map. A pose estimation algorithm minimizes the residual pixel-level error between the estimated and detected lane marker features via Gauss-Newton nonlinear least-squares. Simulations of typical road scenes were used as ground truth to ensure the pose estimator will converge to the true vehicle pose. A successful convergence was defined as a pose estimate that fell within 5 cm and 0.25 degrees of the true vehicle pose. The results show that the longitudinal vehicle state is weakly observable with the smallest region of attraction. Estimating the remaining five vehicle states gives repeatable convergence within the prescribed convergence bounds over a relatively large region of attraction, even for the simple lane detection methods used herein. A main contribution of this paper is to demonstrate a repeatable and verifiable method to assess and compare lane-based vehicle localization strategies.

Topics: Vehicles
Commentary by Dr. Valentin Fuster
2016;():V002T31A003. doi:10.1115/DSCC2016-9741.

This paper presents an approach to the lane change safety system for collision avoidance. The solution is presented in two distinct steps. We first propose a decision strategy based on a discrete time Markov process to determine the safe lane utilizing a set of transition probabilities. These probabilities are calculated according to the distance of the subject vehicle from the surrounding vehicles. The output of decision process is fed to a controller formulated using an scheme to move the vehicle to the desired lane. The overall strategy can be viewed as a combination of continuous control with a discrete decision process. The performance of the proposed scheme is compared with the so-called human-driver model (HDM) based control, which has been broadly discussed in the literature. The simulation study shows the superiority of the proposed controller in terms of trajectory tracking of the reference path, disturbance rejection of the wind load, and effective control input.

Commentary by Dr. Valentin Fuster
2016;():V002T31A004. doi:10.1115/DSCC2016-9772.

This paper focuses on development of an active sensing system for a bicycle to accurately track rear vehicles. Cost, size and power constraints highly limit the type of sensor that can be used on a bicycle for measurement of distances to vehicles. A single beam laser sensor mounted on a rotationally controlled platform is proposed for this sensing mission. The rotational orientation of the laser sensor needs to be controlled in real-time in order to focus on a target point on the vehicle, as the vehicle’s lateral and longitudinal distances change. This tracking problem involves two challenges: Controlling the real-time angular position of the laser sensor based on very limited information and tracking the vehicle’s position for different types of maneuvers. The first challenge is addressed by developing an algorithm to detect whether a reflection is from the front or side of the target vehicle and then controlling sensor orientation to alternately obtain both lateral and longitudinal distance measurements. The second challenge is addressed by using an interacting multiple model observer that incorporates straight and turning vehicle motion models. Simulation results are presented to show the advantages of the developed tracking control system compared to simpler alternatives.

Topics: Vehicles , Bicycles
Commentary by Dr. Valentin Fuster
2016;():V002T31A005. doi:10.1115/DSCC2016-9818.

An application of electric power assist steering (EPAS) system has rapidly grown and overtaken hydraulic power assist steering (HPAS) system in recent automotive industry. The EPAS system has better fuel efficiency and potential application on vehicle dynamic control compared to HPAS. However, it is widely believed that the steering feel of EPAS system is inferior to HPAS system due to its mechanical construction.

This paper first presents a comprehensive model of column electric power assist steering (CEPAS) system consisting of steering wheel, worm gear, assist motor, intermediate shaft, and rack and pinion. In this model, the friction in steering system is modeled by LuGre friction model and basic control strategies are also implemented. Using the proposed CEPAS model, the steering feel responses have been investigated with varying system parameters through simulation, and important factors affecting the steering feel response have been identified. This result gives insights on how the steering feel is affected by various factors and can be useful to improve the steering feel control algorithm design.

Commentary by Dr. Valentin Fuster
2016;():V002T31A006. doi:10.1115/DSCC2016-9827.

Ramp metering is proved to be an effective strategy for reducing or avoiding freeway traffic congestion. As a result, huge amount of research has been conducted on synthesizing effective ramp metering controls. In the previous works, freeway is assumed to be a deterministic system which is in contrast with the intrinsic stochastic nature and behavior of freeways. Our work focuses on bridging this gap, and we propose a framework for freeway ramp metering in a probabilistic setting. Our algorithm finds onramp flows in a freeway network while treating exogenous vehicular arrivals as random variables with known distributions, allowing for the network arrivals to conform with their stochastic nature. We use sampling techniques in a model predictive control setup to formulate a tractable approximation of our stochastic optimization. Furthermore, we demonstrate how to relax the non-linear constraints of our optimization to create a linear program with an augmented set of constraints. We prove that the solution of our linear program formulation is the same as the solution of the original mixed-integer formulation. We showcase the results of our algorithm on an exemplar freeway network and introduce multiple interesting future research directions that are important and can be pursued solely in a stochastic framework.

Topics: Highways
Commentary by Dr. Valentin Fuster

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In