ASME Conference Presenter Attendance Policy and Archival Proceedings

2016;():V001T00A001. doi:10.1115/DSCC2016-NS1.

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presenter attendance policy, if a paper is not presented at the Conference by an author of the paper, the paper will not be published in the official archival Proceedings, which are registered with the Library of Congress and are submitted for abstracting and indexing. The paper also will not be published in The ASME Digital Collection and may not be cited as a published paper.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

Advances in Control Design Methods

2016;():V001T01A001. doi:10.1115/DSCC2016-9656.

Traditionally, a dynamic system is first designed then its controller is designed in a sequential process, which offers many organizational and computational advantages. However, this sequential design approach may not lead to system optimality compared to the combined design, or co-design, of the dynamic system and its controller. It has been proposed that a control proxy function (CPF) be used in the first step of this sequential design process to achieve, or approach, system-optimality. This paper considers the use of a CPF based on controllability and observability Gramians and a balanced realization to ensure the design of the dynamic system for both ease of control and estimation. The proposed approach is illustrated using an example of linear quadratic design for combined passive and active vibration control.

Topics: Design
Commentary by Dr. Valentin Fuster
2016;():V001T01A002. doi:10.1115/DSCC2016-9684.

In this work an error-integral-driven sliding mode controller (EID-SMC) for multi-input multi-output (MIMO) minimum phase linear time-invariant (LTI) systems with feedthrough controls and output disturbance is analyzed. Though the sliding variable remains in the vicinity of the sliding surface without reaching it, it is shown that the steady-state error vanishes exponentially asymptotically within the boundary layer even if parameter uncertainty and unmatched input/output disturbances exist. The pole-placement is accomplished indirectly by an iterative optimization routine by considering limits on controls and state contrary to the existing practice in SMC where either direct pole placement or quadratic norm optimality of a performance is used. For the proposed controller framework the Luenberger observer is presented.

Topics: Errors
Commentary by Dr. Valentin Fuster
2016;():V001T01A003. doi:10.1115/DSCC2016-9702.

RF MEMS switches have many advantages over solid-state switches but their main disadvantage is poor reliability. The reliability problem stems from the impact forces generated at the time of closing of the switch and subsequent bouncing. This paper proposes a feedback control strategy to rapidly close the switch without bouncing. The stiffness of the system is switched between a positive value and a negative value to achieve small rise and setting times with no overshoot. Simulation results are presented to demonstrate the feasibility of the proposed control strategy.

Commentary by Dr. Valentin Fuster
2016;():V001T01A004. doi:10.1115/DSCC2016-9705.

The inherent compliance, high power-density, and musclelike properties of soft actuators are especially attractive and useful in many applications, including robotics. In comparison to classical/modern control approaches, model-based control techniques, e.g., sliding mode control (SMC), applied to flexible fluidic actuators (FFAs) offer significant performance advantages and are considered to be state-of-the-art. Improvements in position tracking are possible using nonlinear control approaches that offer enhanced performance for common applications such as tracking of sinusoidal trajectories at high frequencies.

This paper introduces a SMC approach that increases the tracking capabilities of prolate flexible pneumatic actuators (PF-PAs). A model-based proportional, integral, derivative sliding mode control (PIDSMC) approach designed for position control of PFPAs is proposed. SMC and PIDSMC systems are implemented on low-cost open-source controls hardware and tested for tracking sinusoidal trajectories at frequencies of 0.5 Hz and 1.0 Hz with an amplitude of 8.255 mm and an offset of 12.7 mm. The PIDSMC approach reduced the maximum tracking error by 20.0%, mean error by 18.6%, and root-mean-square error by 10.5% for a 1 Hz sinusoidal trajectory and by 8.7%, 14.7%, and 3.8%, respectively, for a 0.5 Hz sinusoidal trajectory. These reductions in tracking errors demonstrate performance advantages of the PIDSMC over conventional sliding mode position controllers.

Commentary by Dr. Valentin Fuster
2016;():V001T01A005. doi:10.1115/DSCC2016-9740.

This paper addresses the problem of command following for linear, multi-input multi-output, time delay systems using filtered dynamic inversion. The dynamic inversion control input is obtained for linear time-delay systems utilizing the ring framework. This control input is combined with a low-pass-filter to construct a high-gain-stabilizing controller. It is shown that the time-delay-filtered-dynamic-inversion (TD-FDI) uses output feedback, reference-model-input feed-forward, and only requires limited model information. Numerical simulations demonstrate that for sufficiently large choice of a single control parameter, TD-FDI makes the L of the command-following error arbitrarily small.

Commentary by Dr. Valentin Fuster
2016;():V001T01A006. doi:10.1115/DSCC2016-9762.

This paper uses Linear Matrix Inequality (LMI) techniques to apply regional eigenvalue assignment constraints to a dynamic state-feedback controller design for discrete-time systems with vanishing nonlinear perturbations. The controller design also incorporates the H∞ performance criterion. The regional eigenvalue assignment place the eigenvalues of the linear part of the system in two distinct regions, one region for the controller eigenvalues and one region for the observer eigenvalues, in such a way that the state estimation error goes to zero significantly faster than the state reaches steady state.

Commentary by Dr. Valentin Fuster
2016;():V001T01A007. doi:10.1115/DSCC2016-9804.

DC-DC converters are an efficient way to convert a source voltage from one to level to another and have found extensive applications in many areas such as portable electronics, solar and wind energy systems. This paper presents a comparison of first order and higher order sliding mode control of buck-boost converters. Sliding mode control is ideal for controlling non-linear systems like switched voltage regulators as a result of its robustness to internal parameter uncertainties and external disturbances. First order sliding mode control is subject to a phenomenon known as chattering, which causes an undesirable oscillation about the desired output. Computer simulation studies are presented and show that the higher order controller reduces this problem.

Commentary by Dr. Valentin Fuster
2016;():V001T01A008. doi:10.1115/DSCC2016-9806.

A new technique was presented that facilitates design of independent full-state feedback controllers at the subsystem levels. Different types of local controllers, for example, eigenvalue assignment, robust, optimal (in some sense L1, H2, H, ...) may be used to control different subsystems. This feature has not been available for any known linear feedback controller design. In the second part of the paper, we specialize the results obtained to the three time-scale linear systems (singularly perturbed control systems) that have natural decomposition into slow, fast, and very fast subsystems. The proposed technique eliminates numerical ill-condition of the original three-time scale problems.

Commentary by Dr. Valentin Fuster
2016;():V001T01A009. doi:10.1115/DSCC2016-9813.

Dynamic systems with time-varying delay in the control input are studied in the present paper. The delay is considered as a varying parameter and Padé approximation is applied to transfer the infinite-dimensional delay problem into a finite-dimensional paradigm represented in the form of a non-minimum phase system (NMP). Inherited delay characteristics are now represented through unstable internal dynamics for the NMP system, which poses restrictions on the achievable control bandwidth thereby resulting in an imperfect tracking performance and poor stability condition. Presented in this paper, is a methodical parameter-varying loop-shaping control design approach, which simultaneously satisfy a variety of control requirements and offer an insight into the limitations posed by the NMP representation. The suggested method is then applied to fueling control in lean-burn gasoline engines addressing the varying transport and combustion delay. The developed approach is validated with experimental data on a Ford F-150 truck SI lean-burn engine with large time-varying delay in the control loop and the closed-loop system responses are presented to demonstrate disturbance rejection, measurement noise attenuation, and robustness properties against delay estimation errors.

Commentary by Dr. Valentin Fuster
2016;():V001T01A010. doi:10.1115/DSCC2016-9838.

The objective of this paper is to develop a novel two-level supervised fuzzy controller to stabilize the response of electrostatically actuated microbeams beyond their pull-in range. To this end, Lagrange equations are utilized to derive the differential equations governing the dynamic behavior of the system. To investigate the possibility of using a passive control strategy, the static behavior of the system is studied in detail. Through some open loop simulations, the qualitative and quantitative dependence of the beam deflection to the applied voltage and system parameters are studied. Based on the understanding obtained from these studies, a single level fuzzy controller is designed to control the response of the microstructure. In order to enhance the performance of the closed-loop system, another higher level supervisory fuzzy controller is designed to tune the maximum allowable voltage the lower level controller can apply. Simulation results reveal that both single level and multi-level fuzzy controllers can extend the travel range of the microbeams beyond its pull-in range. However the rise time, overshoot and settling time in the multilevel controlled system is far better than that of a simple single level fuzzy controller. The novel controller presented in this paper can be applied in most intrinsically nonlinear nano/micro structures to help them to have more efficient regulations and command tracking maneuvers.

Commentary by Dr. Valentin Fuster
2016;():V001T01A011. doi:10.1115/DSCC2016-9905.

This paper proposes a method for near energy optimal allocation of control effort in dual-input over-actuated systems using a linear time-invariant (LTI) controller. The method assumes a quadratic energy cost functional, and the non-causal energy optimal control ratio within the redundant actuation space is defined. Near energy optimal control allocation is addressed by using a LTI controller to align the control inputs with a causal approximation of the energy optimal control ratio. The use of a LTI controller for control allocation leads to low computation burden compared to techniques in the literature which require optimization at each time step. Moreover, the proposed method achieves broadband, near optimal control allocation, as opposed to traditional allocation methods which make use of a static system model for control allocation. The proposed method is validated through simulations and experiments on an over-actuated hybrid feed drive system. Significant improvements in energy efficiency without sacrificing positioning performance are demonstrated.

Topics: Optimal control
Commentary by Dr. Valentin Fuster

Advances in Nonlinear and Optimal Control

2016;():V001T02A001. doi:10.1115/DSCC2016-9627.

In this paper, we focus on interconnected trajectory optimization of two sets of solenoid actuated butterfly valves dynamically coupled in series. The system undergoes different approach angles of a pipe contraction as a typical profile of the so-called “Smart Valves” network containing tens of actuated valves. A high fidelity interconnected mathematical modeling process is derived to reveal the expected complexity of such a multiphysics system dealing with electromagnetics, fluid mechanics, and nonlinear dynamic effects. A coupled operational optimization scheme is formulated in order to seek the most efficient trajectories of the interconnected valves minimizing the energy consumed enforcing stability and physical constraints. We examine various global optimization methods including Particle Swarm, Simulated Annealing, Genetic, and Gradient based algorithms to avoid being trapped in several possible local minima. The effect of the approach angles of the pipeline contraction on the amount of energy saved is discussed in detail. The results indicate that a substantial amount of energy can be saved by an intelligent operation that uses flow torques to augment the closing efforts.

Topics: Optimization , Pipes , Valves
Commentary by Dr. Valentin Fuster
2016;():V001T02A002. doi:10.1115/DSCC2016-9633.

The multi-objective optimal control design usually generates hundreds or thousands of Pareto optimal solutions. How to assist an user to select an appropriate controller to implement is a post-processing issue. In this paper, we develop a method of cluster analysis of the Pareto optimal designs to discover the similarity of the optimal controllers. After we identify the clusters of optimal controllers, we then develop a switching strategy to select controls from different clusters to improve the performance. Numerical results show that the switching control algorithm is quite promising.

Commentary by Dr. Valentin Fuster
2016;():V001T02A003. doi:10.1115/DSCC2016-9712.

This contribution considers a model-free control method based on an optimal iterative learning control framework to design a suitable controller. Using this framework, the controller requires neither the information about the systems dynamical structure nor the knowledge about system physical behaviors. The task is solved using only the system outputs and inputs, which are assumed as measurable. The structure of the proposed method consists of three parts. The first part is implemented through an intelligent PID controller on the system. In the second part, a robust second order differentiator via sliding mode is applied in order to estimate accurately the evolution of the state function. In the third part, an optimal iterative learning control is chosen to improve the performance. Numerical examples are shown to demonstrate the successful application and performance of the method.

Commentary by Dr. Valentin Fuster
2016;():V001T02A004. doi:10.1115/DSCC2016-9745.

Two approximate solutions for optimal control of switched systems with autonomous subsystems and continuous-time dynamics are developed. The proposed solutions consist of online training algorithms with recursive least squares training laws. The first solution is the classic policy iteration algorithm which imposes heavy computational burden (full back-up). In order to relax the computational burden in the policy iteration algorithm, the second algorithm is presented. The convergence of the proposed algorithms to the optimal solution in online training is investigated. Simulation results are presented to illustrate the effectiveness of the discussed algorithms.

Topics: Optimal control
Commentary by Dr. Valentin Fuster
2016;():V001T02A005. doi:10.1115/DSCC2016-9786.

The reference governor modifies set-point commands to a closed-loop system in order to enforce state and control constraints. In this paper, we describe an approach to reference governor implementation for nonlinear systems, which is based on bounding (covering) the response of a nonlinear system by the response of a linear model with a set-bounded disturbance input. Such a design strategy is of interest as it reduces the online optimization problem to a convex quadratic programming (QP) problem with linear inequality constraints, thereby permitting standard QP solvers to be used. A numerical example is reported.

Commentary by Dr. Valentin Fuster
2016;():V001T02A006. doi:10.1115/DSCC2016-9879.

In this paper, a novel fuzzy gain-scheduling output feedback control method is presented for T-S fuzzy systems subject to actuator saturation. To deal with saturation nonlinearity, the dead-zone function of control input is treated as an additional controller input. With the help of set conclusion condition, the controller can be synthesised based on fuzzy Lyapunov functions to guarantee the exponential stability of the closed-loop system in a larger region and a better L2 gain performance. Moreover, the full block S-procedure, which is widely used in robust control theory, is introduced to relax the synthesis conditions for T-S fuzzy systems to reduce the conservatism caused by quadratic terms in the conditions. Finally, a numerical example is provided to illustrates the effectiveness of the proposed control method.

Commentary by Dr. Valentin Fuster

Advances in Robotics

2016;():V001T03A001. doi:10.1115/DSCC2016-9653.

The division of labor amongst a heterogeneous swarm of robots increases the range and sophistication of the tasks the swarm can accomplish. To efficiently execute a task the swarm of robots must have some starting organization. Over the past decade segregation of robotic swarms has grown as a field of research drawing inspiration from natural phenomena such as cellular segregation. A variety of different approaches have been undertaken to devise control methods to organize a heterogeneous swarm of robots. In this work, we present a convex optimization approach to segregate a heterogeneous swarm into a set of homogeneous collectives. We present theoretical results that show our approach is guaranteed to achieve complete segregation and validate our strategy in simulation and experiments.

Commentary by Dr. Valentin Fuster
2016;():V001T03A002. doi:10.1115/DSCC2016-9677.

Micro robots have been recognized as a promising solution for exploring complex and dangerous environments during rescue and reconnaissance missions. The idea of micro centipede robots is drawing increasing attention because of their great potentials to get through rubble and reaching buried areas. This paper explores the possibility of using micro electromagnetic actuators on micro centipede robots, which could give better mobility than using piezoelectric actuators. This paper also discusses the whole-body dynamics of centipede robots. A series of simulation studies are conducted to draw conclusions on how the propulsion characteristics of individual legs and the ground friction affect the dynamics and stability of the whole centipede robot.

Commentary by Dr. Valentin Fuster
2016;():V001T03A003. doi:10.1115/DSCC2016-9751.

This paper provides a natural, yet low-cost way for human to interact with electronic devices indoor: the development of a human-following mobile robot capable of controlling other electrical devices for the user based on the user’s gesture commands. The overall experimental setup consists of a skid-steered mobile robot, Kinect sensor, laptop, wide-angle camera and two lamps. The OpenNI middleware is used to process data from the Kinect sensor, and the OpenCV is used to process data from the wide-angle camera. A new human-following algorithm is proposed based on human motion estimation. The human-following control system consists of two feedback control loop for linear and rotational motions, respectively. A lead-lag and lead controller are developed for the linear and rotational motion control loop, respectively. Experimental results show that the tracking algorithm is robust and reduced the distance and angular error by 40% and 50%, respectively. There are small delays (0.5 s for linear motion and 1.5 s for rotational motion) and steady-state errors (0.1 m for linear motion and 1.5° for rotational motion) of the system’s response. However, the delays and errors are acceptable since they do not cause the tracking distance or angle out of the desirable range (±0.05m and ±10° of the reference input). There are four gestures designed for the user to control the robot, two switch-mode gestures, lamp-creation, lamp-selection and color change gesture. Success rates of gesture recognition are more than 90% within the detectable range of the Kinect sensor.

Topics: Mobile robots
Commentary by Dr. Valentin Fuster
2016;():V001T03A004. doi:10.1115/DSCC2016-9903.

Advancement in solid state device technology has made it possible to replicate some learning behaviors on those devices as observed in biological neurons. A very widely used mechanism of learning is Spike Timing Dependent Plasticity (STDP) to realize an unsupervised learning process. In this work, we have developed a novel solution for learning using such networks in robots that can be implemented on novel resistive memory devices. This artificial brain mechanism is capable of learning by observing the environment and taking actions to achieve a desired goal. We have demonstrated this learning scheme using a mathematical model representing the reconfiguration of strengths in synapses. This model can be easily implemented on miniaturized microprocessors using resistive crossbar memories. The reconfigurable resistive memory devices in crossbar arrays are capable of mimicking the synapses in human brains by changing their resistances on application of appropriate voltage signal. In this work, we have demonstrated the potential of this learning scheme by applying it to navigate a two-wheeled differential drive robot in an environment cluttered with obstacles. It can be observed that the robot is able to navigate the environment autonomously, and can reach a given target while avoiding obstacles.

Commentary by Dr. Valentin Fuster
2016;():V001T03A005. doi:10.1115/DSCC2016-9915.

With advances in actuation and sensing, smart materials has drawn a growing attention from researchers in under water robotic fish. In this paper, a compact, noiseless, and untethered biomimetic robotic fish propelled by Ionic Polymer-Metal Composite (IPMC) actuators is developed. The robot fish employs two pectoral fins to generate steering and one caudal fin to generate main propulsion. A passive plastic fin is attached to the IPMC beam to enhance propulsion. With multiple IPMC fins, the fish is capable of 2D maneuvering. One small size programmable circuit board is designed for the 2D controllable fish. The Experimental results have shown that the forward-swimming speed can reach up to 1cm/sec and the both left-turning and right turning speed can reach up to 2 rad/sec.

Topics: Robotics , Biomimetics , Fins
Commentary by Dr. Valentin Fuster
2016;():V001T03A006. doi:10.1115/DSCC2016-9918.

The increase of potential threats to the integrity of our aquatic ecosystems has caused global concerns which have led to interest in the use autonomous aquatic robots to monitor such environments. In recent years, underwater robots that propel and maneuver themselves like real fish, often called robotic fish, have emerged as mobile sensing platforms for freshwater and marine environments. These robots achieve locomotion via actively controlled fins, and actuation is often achieved via oscillatory inputs. Given these types of applications, accuracy and energy-saving in trajectory control is of importance for mission successes. In this work, we propose a nonlinear model predictive control (NMPC) approach to path following of a tail-actuated robotic fish. In this design, we use bias and amplitude of the tail-beat as the input to be determined by the NMPC. The effectiveness of the proposed approach is demonstrated via simulation.

Commentary by Dr. Valentin Fuster

Advances in Wind Energy Systems

2016;():V001T04A001. doi:10.1115/DSCC2016-9716.

A variable ratio gearbox (VRG) can enable small wind turbines to operate at discrete variable rotor speeds. This reliable, low-cost, alternative does not require power conversion equipment, as is the case with conventional variable speed. Previous work conducted by the author has demonstrated that a VRG can increase the power production for a fixed-speed system with passive blades. The current study characterizes the performance of a wind turbine equipped with a VRG and active blades. The contribution of this work is an integrative framework that optimizes power production with blade root stress. It works by defining a set of control rules that specify the VRG gear ratio and pitch angle that will be used in relation to wind speed. Three ratios are selected through the proposed procedure. A case study based on the simulation of a 300-kW wind turbine model is performed to demonstrate the proposed technique. The model is constructed with aerodynamic, mechanical, and electrical submodels. These drivetrain components work together to simulate the conversion of moving air to electrical power. The blade element momentum (BEM) technique is used here to compute the blade loading. The resulting torque and rotor speed are reduced and increased, respectively, through the mechanical system gearbox. The output from this is then applied to the electrical generator. The BEM technique is also used here to determine the bending and thrust and loads that are applied to the blade. The stress in the root of the blade is then determined based on these loads, and that caused by centrifugal force and gravity. The proposed method devises a VRG design and control algorithm based on the unique wind conditions at a given installation site. Two case studies are conducted using wind data sets provided by the National Renewable Energy Laboratory (NREL). Low and high-speed data set are selected as inputs to demonstrate the versatility of the proposed method. Dynamic programming is used to reduce the computational expense. This enables the simulation of an exhaustive set of potential VRG combinations over each set of recorded wind data. Each possible combination is evaluated in terms of the total energy production and blade-root stress produced over the simulation period. A set of weights is applied to a multi-objective function that computes the cost associated with each combination. A Pareto analysis is then used to identify the VRG combination and establish the control algorithm for both systems. The results suggest that the VRG can improve energy production in the partial-load region by roughly 10% in both cases. Although stress increases in Region 2, it decreases in Region 3, and overall, through the optimal selection of gear combinations.

Commentary by Dr. Valentin Fuster
2016;():V001T04A002. doi:10.1115/DSCC2016-9719.

The increasing size of modern wind turbines also increases the structural loads on the turbine caused by effects like turbulence or asymmetries in the inflowing wind field. Consequently, the use of advanced control algorithms for active load reduction has become a relevant part of current wind turbine control systems. In this paper, an H-norm optimal multivariable control design approach for an individual blade-pitch control law is presented. It reduces the structural loads both on the rotating and non-rotating parts of the turbine. Classical individual blade-pitch control strategies rely on single control loops with low bandwidth. The proposed approach makes it possible to use a higher bandwidth since it takes into account coupling at higher frequencies. A controller is designed for the utility-scale 2.5 MW Liberty research turbine operated by the University of Minnesota. Stability and performance are verified using a high-fidelity nonlinear benchmark model.

Commentary by Dr. Valentin Fuster
2016;():V001T04A003. doi:10.1115/DSCC2016-9737.

This paper presents an original experimental setup for controlling and measuring the crosswind flight of airborne wind energy systems in a laboratory environment. Execution of cross-wind flight patterns, which is achieved in this work through the asymmetric motion of three tethers, enables dramatic increases in energy generation compared with stationary operation. Achievement of crosswind flight in the 1:100-scale experimental framework described herein allows for rapid, inexpensive, and dynamically scalable characterization of new control algorithms without recourse to expensive full-scale prototyping. This work is the first example of successful lab-scale control and measurement of crosswind motion for an airborne wind energy system. Specifically, this paper presents the experimental setup, crosswind flight control strategy, and experimental results for a model of the Altaeros Buoyant Airborne Turbine (BAT). The results demonstrate that crosswind flight control can achieve nearly 50 percent more power production then stationary operation, while also demonstrating the potential of the experimental framework for further algorithm development.

Commentary by Dr. Valentin Fuster
2016;():V001T04A004. doi:10.1115/DSCC2016-9771.

The goal of this paper is to control the dynamics of an islanded microgrid, a small-scale power system with distributed generation. An islanded microgrid is disconnected from the larger, main grid, and must maintain voltage and frequency standards using only local generation. As a result, islanded microgrids are more vulnerable to fluctuations in power supply and demand; this is especially relevant for intermittent renewable sources like wind turbines.

The system is stabilized with static-output-feedback γ-suboptimal control. This is a multiple-input multiple-output (MIMO) controller in which the measured data is used as the direct input to a static gain matrix, whose output is in turn used to control the closed-loop system. In order to judge the performance of the decentralized controllers, the micgorid is controlled first in a centralized manner, where each controller has access to all measured state variables. Decentralized controllers are then synthesized by casting the problem as a convex program, where each controller only has access to locally measured variables. Control performance is compared with respect to a switched-on wind turbine, where we see that the decentralized controller effectively mitigates the system disturbance due to the renewable intermittency.

Topics: Microgrids
Commentary by Dr. Valentin Fuster
2016;():V001T04A005. doi:10.1115/DSCC2016-9825.

In an isothermal compressed air energy storage (CAES) system, it is critical that the high pressure air compressor/expander is both efficient and power dense. The fundamental trade-off between efficiency and power density is due to limitation in heat transfer capacity during the compression/expansion process. In our previous works, optimization of the compression/expansion trajectory has been proposed as a means to mitigate this trade-off. Analysis and simulations have shown that the use of optimized trajectory can increase power density significantly (2–3 fold) over ad-hoc linear or sinusoidal trajectories without sacrificing efficiency especially for high pressure ratios. This paper presents the first experimental validation of this approach in high pressure (7bar to 200bar) compression. Experiments are performed on an instrumented liquid piston compressor. Correlations for the heat transfer coefficient were obtained empirically from a set of CFD simulations under different conditions. Dynamic programming approach is used to calculate the optimal compression trajectories by minimizing the compression time for a range of desired compression efficiencies. These compression profiles (as function of compression time) are then tracked in a liquid piston air compressor testbed using a combination of feed-forward and feedback control strategy. Compared to ad-hoc constant flow rate trajectories, the optimal trajectories double the power density at 80% efficiency or improve the thermal efficiency by 5% over a range of power densities.

Commentary by Dr. Valentin Fuster

Aerospace Applications

2016;():V001T05A001. doi:10.1115/DSCC2016-9774.

As small rotorcraft grow in capability, the possibilities for their application increase dramatically. Many of these new applications require stable outdoor flight, necessitating a closer look at the aerodynamic response of the aircraft in windy environments. This paper develops the equations of motion for a single-propeller test stand by analyzing the blade-flapping response of a small-stiff propeller in wind. The system dynamics are simulated to show behavior under various wind conditions, and stable system equilibria are identified. Experiments with a rotor-pendulum validate the simulations, including system equilibria and gust response.

Commentary by Dr. Valentin Fuster
2016;():V001T05A002. doi:10.1115/DSCC2016-9812.

In this paper, the uncertainty and disturbance estimator (UDE)-based position controllers are developed to achieve the robust position control of a quadrotor using only onboard sensing. Firstly, in order to accurately estimate the positions of the quadrotor in GPS-denied environments, an open source high speed optical flow sensor PX4FLOW is adopted. Secondly, the UDE-based controllers are developed to handle the challenges brought by the highly nonlinear quadrotor position dynamics, including underactuation, coupling, nonaffine inputs, model uncertainties and external disturbances. Real flight experiments, including hover and disturbance rejection are carried out to demonstrate the effectiveness of the developed UDE-based controllers.

Commentary by Dr. Valentin Fuster
2016;():V001T05A003. doi:10.1115/DSCC2016-9842.

The main contribution of this paper is to introduce a computationally efficient iterative closest line (ICL) algorithm for determining indoor position drift of a quadcopter using minimal lidar data. In addition, we present the system-level design and implementation of a new quadcopter both as hardware and flight control algorithms. Such a platform allows us to develop and experiment new control and system optimization techniques. As an example, we discuss how the proposed ICL algorithm is used for position hold and control purposes by plugging it into the low level implementation of the flight control algorithm of the quadcopter. For testing and validation we use simulations with real world data. As part of the system-level design aspects, we present an investigation of the quadcopter power consumption. We are interested in power consumption because it is the major factor that determines the flight time of a typical quadcopter. We believe that this work is a contribution toward achieving better quadcopter design and control for indoor autonomous navigation.

Commentary by Dr. Valentin Fuster
2016;():V001T05A004. doi:10.1115/DSCC2016-9896.

Unmanned Air Vehicles (UAVs) have become more applicable in several military and civilian domains during the last decade due to their enhanced capabilities. For outdoor environments, one of the most reliable methods for navigation is waypoint following. Usually, the path or trajectory to be followed is decided based on the behavior of the vehicle during path following such as time and energy consumption. However, feasibility of the trajectory is based on the vehicle dynamics and the ability of UAV to follow the path generated based on the way-point setup. Moreover, the paths obtained from minimizing time or energy consumption are often contradictory. This paper investigates two cases where the objective of the path planning based on the Continuous Cubic C1 Bezier Curve (C1CBC) method and 4 other first degree Bezier curves is: i) minimizing the time consumption, and ii) minimizing the energy consumption. At the end, the quad-copters were simulated through the generated path to reveal the effects of the path UAV follows to reach to the goal position.

Topics: Splines , Navigation
Commentary by Dr. Valentin Fuster
2016;():V001T05A005. doi:10.1115/DSCC2016-9897.

In this paper, stability and control of tilting-rotor quadcopters is presented upon failure of one propeller during flight. A tilting rotor quadcopter provides advantage in terms of additional stable configurations and maneuverability. Upon failure of one propeller, a traditional quadcopter has a tendency of spinning about the primary axis fixed to the vehicle as an outcome of the asymmetry about the yaw axis. This forces the quadcopter to land abandoning its mission. The tilting-rotor configuration is an over-actuated form of a traditional quadcopter and it is capable of handling a propeller failure, thus making it a fault tolerant system. In this paper, a dynamic model of tilting-rotor quadcopter with one propeller failure is derived and a controller is designed to achieve hovering and navigation capability. The simulation results showing the effectiveness of the proposed controller is presented using the translational and hovering motion.

Topics: Rotors , Failure , Propellers
Commentary by Dr. Valentin Fuster
2016;():V001T05A006. doi:10.1115/DSCC2016-9913.

Enforcing safety is critical for aerial robotics. In this paper we consider the safety control problem for a 3D quadrotor with limited sensing range subject to avoiding collisions with time-varying obstacles. By using the concepts of Control Lyapunov Functions (CLFs) and Control Barrier functions (CBFs), we propose a control algorithm that explicitly considers the nonlinear and underactuated dynamics of a quadrotor to strictly guarantee time-varying safety-critical constraints. We demonstrate the feasibility of our proposed control design through numerical validation of (a) aerial flight through a region of dense cluttered obstacles, and (b) aerial flight through a dense time-varying obstacle field.

Topics: Safety
Commentary by Dr. Valentin Fuster

Assistive and Rehabilitation Robotics

2016;():V001T06A001. doi:10.1115/DSCC2016-9660.

Recently, the authors designed and fabricated an Instrumented Walkway for the estimation of the ankle mechanical impedance in the sagittal and frontal planes during walking in arbitrary directions [1]. It consists of a powered platform; therefore, the users do not need to wear or carry any measurement device or actuation system other than reflective markers used to record the ankle kinematics with a motion capture camera system. This paper describes the continuous development of the Instrumented Walkway and presents an experimental preliminary validation of its capability to estimate the impedance of a system with time-varying dynamics. To validate the system, a mockup with mechanical characteristics similar to a human lower-leg and controllable time-varying stiffness was used. The stiffness of the mockup was estimated with fixed and time-varying stiffness. With fixed stiffness, a stochastic system identification method was used to estimate the mockup’s impedance. When the mockup presented a time-varying stiffness, a second order parametric model was used. The RMS error between the two methods was 2.81 Nm/rad (maximum 4.12 Nm/rad and minimum of −3.41 Nm/rad). The results show that the proposed approach can estimate the stiffness of systems with time-varying dynamics or static dynamics with similar accuracy. Since the setup was already validated for systems with time-invariant dynamics, it concluded the system’s applicability for time-varying systems such as the human ankle-foot during the stance phase.

Commentary by Dr. Valentin Fuster
2016;():V001T06A002. doi:10.1115/DSCC2016-9699.

There are over three million wheelchair users within the United States and that number is growing. This paper is concerned with improving the safety of wheelchair operation by the on-line estimation of tire slip. Wheelchair tire slip is a result of icy or low friction surfaces, often representative of dangerous conditions. In this research, wheel slip is detected by estimating the instantaneous center of rotation (ICR) locations of wheelchair wheels relative to the ground surface. Any departure of the estimated ICR positions from the wheel contact point indicates slippage is occurring.

An Extended Kalman Filter (EKF) algorithm uses inputs of position and orientation obtained via map-based localization to detect changes in wheelchair ICR location estimates. The ICR EKF algorithm is verified in simulation. A robotic wheelchair is used for testing the presented algorithms under conditions inducing tire slip. The results show that the ICR locations do not vary significantly when the wheelchair is operated under normal conditions, i.e. low slip surfaces; however, they change significantly under slip conditions. Implementing this method with electric wheelchairs can improve the prediction of wheelchair motion on slippery surfaces, enabling warning systems and safe operational modes that can enhance the safety of wheelchair users.

Commentary by Dr. Valentin Fuster
2016;():V001T06A003. doi:10.1115/DSCC2016-9706.

The first step to study and develop a two Degrees of Freedom (DOF) prosthesis is to derive a dynamic model for simulation and control design. In this paper, the ankle-foot prosthesis has controllable Dorsi-Plantarflexion (DP) and Inversion-Eversion (IE) DOF. We derive a compliant dynamic model for a recently developed ankle-foot prosthesis followed by identification of the actuators, transmission, and prosthetic foot parameters. The resulting model is then verified experimentally and in simulation. Dynamic decoupling of the actuators to the ankle’s DP and IE DOF is also investigated using Bode plots. The code used for simulating the prosthesis is provided on GitHub for the community.

Commentary by Dr. Valentin Fuster
2016;():V001T06A004. doi:10.1115/DSCC2016-9778.

This paper builds on prior investigations of the electromyogram (EMG) control of a single degree-of-freedom (DOF) transfemoral prosthetic limb, but augmented with mechanical haptic feedback of prosthetic limb state. Preliminary studies were conducted where quasi-static and vibratory cutaneous haptic feedback was provided to subjects performing nonweight-bearing motion tracking tasks with the EMG controlled transfemoral prosthesis. The results of these studies showed that the subjects exhibited improved tracking performance when following pseudo-random step commands under EMG control augmented with static and vibratory haptic feedback cues. The work to be discussed in this paper augments the EMG control architecture to foster improved co-contraction of the instrumented antagonist muscle pair. Using the modified EMG control architecture, experimental studies were conducted to investigate the efficacy of two haptic feedback modalities in conveying information pertinent to single-DOF nonweight-bearing sinusoidal motion tracking tasks. The two haptic feedback modalities investigated were quasi-static pressure feedback provided with pneumatic actuation and vibratory feedback provided by a vibrotactile motor array. Able-bodied test subjects were asked to control the prosthetic knee to follow sinusoidal trajectories with and without visual and haptic feedback. Experimental results show that EMG-control performance in tracking sinusoidal trajectories significantly improves in visually devoid environments when haptic feedback in the form of error-based and pacemaking stimulation patterns are presented to the user.

Commentary by Dr. Valentin Fuster
2016;():V001T06A005. doi:10.1115/DSCC2016-9802.

Electric wheelchair users depend on a reliable power system in order to regain mobility in their daily lives. If a wheelchair’s battery power depletes without the user being aware, the individual may become stranded, further limiting their freedom of mobility and potentially placing the user in a harmful situation. This research seeks to develop a State-of-Charge (SOC) estimator for the batteries of an electric wheelchair. A second-order equivalent circuit battery model is developed and parameterized for a wheelchair’s lead-acid battery pack. To simplify the SOC estimation, this algorithm models a vehicle’s fuel gauge. A coulomb accumulator is incorporated to estimate energy usage in the non-linear region of the OCV-SOC curve, while a Kalman filter is used to estimate SOC in the linear region of the curve. The estimator is verified using experimentally collected data on-board a robotic wheelchair. The implementation of these algorithms with powered wheelchairs can significantly improve the estimation of wheelchair battery power and can ultimately be coupled with warning systems to alert users of depleting battery life, as well as enable low-power modes to increase wheelchair user safety.

Topics: Gages , Fuels , Wheelchairs
Commentary by Dr. Valentin Fuster
2016;():V001T06A006. doi:10.1115/DSCC2016-9811.

As a muscle activation treatment, magnetic stimulation can elicit current flow through appropriate amplitude and frequency to generate electrical currents, which will prompt activation in the muscle tissue. The design of an electromagnetic system serves as an alternative to direct electrical stimulation for treating muscles located deep inside the tissue, such as the laryngeal muscles, stimulated via output-based control systems. Through magnetic induction, we can implement electrical stimulation and target specific muscle activation. In light of this approach, our goal is to incorporate feedback control using an electromagnetic stimulation system. These studies, therefore, focus on assessing a linear Proportional-Integral (PI) controller and two nonlinear controllers, Model Reference Adaptive Controller (MRAC) and an Adaptive Augmented PI (ADP-PI) system, to identify the most appropriate controller providing effective stimulation of the muscle. Our work focuses on the feasibility of our system to carry out in vitro experiments on mouse skeletal muscle and the characterization of the two-coil system for future design of implantable systems in humans.

Topics: Muscle
Commentary by Dr. Valentin Fuster
2016;():V001T06A007. doi:10.1115/DSCC2016-9851.

Although dynamic walking methods have had notable successes in control of bipedal robots in the recent years, still most of the humanoid robots rely on quasi-static Zero Moment Point controllers. This work is an attempt to design a highly stable controller for dynamic walking of a human-like model which can be used both for control of humanoid robots and prosthetic legs. The method is based on using time-based trajectories that can induce a highly stable limit cycle to the bipedal robot. The time-based nature of the controller motivates its use to entrain a model of an amputee walking, which can potentially lead to a better coordination of the interaction between the prosthesis and the human. The simulations demonstrate the stability of the controller and its robustness against external perturbations.

Commentary by Dr. Valentin Fuster
2016;():V001T06A008. doi:10.1115/DSCC2016-9894.

This paper presents a 5 degree-of-freedom (DoF) low inertia shoulder exoskeleton that was developed using two novel technologies with a broad range of application. The first novelty is a 3-DoF spherical parallel manipulator (SPM) that uses three linear actuators. Each actuator is designed using a method of motion coupling such that the pitch and linear stroke DoF are dependent. By using an SPM, this shoulder exoskeleton takes advantage of the inherent low effective inertia property of parallel architecture. The second novelty is a 2-DoF passive slip mechanism that couples the user’s upper arm to the SPM. This slip mechanism increases system mobility and prevents joint misalignment caused by the translational motion of the user’s glenohumeral joint from introducing mechanical interference that could affect the device’s kinematic solution or harm the user. An experiment to validate the kinematics of the SPM was performed using motion capture. A computational slip model was created to quantify the slip mechanism’s response for different conditions of joint misalignment. In addition to offering a low inertia solution for the rehabilitation or augmentation of the human shoulder, the presented device demonstrates the technologies of actuator motion coupling and passive slip for use in exoskeletal systems. The use of motion coupling could be applied to other types of parallel actuated architectures in order to constrain the kinematics or improve stiffness characteristics. Passive slip mechanisms could have application in either serial or parallel actuated systems as a means of negating the adverse effects of joint misalignment.

Commentary by Dr. Valentin Fuster

Assistive Robotics

2016;():V001T07A001. doi:10.1115/DSCC2016-9731.

We analyzed a new class of passive devices that can help individuals regain their independence. A new walker design has been commercialized under the name of UP’N FREE®1. This new design can lift the user from a seated posture to a standing position through the employment of a four bar mechanism and gas piston. When the user is standing, the mechanism can provide partial compensation from gravity to help the user regain mobility. By using this system, users can sit and stand without another person’s assistance and can also compensate for gravity loads during walking both indoor and outdoor. This paper illustrates two examples of how to implement the dynamic lifting support and discusses the pros and cons of different system configuratons of a a gas spring cylinder as a passive actuator.

Commentary by Dr. Valentin Fuster
2016;():V001T07A002. doi:10.1115/DSCC2016-9781.

The hydraulic human power amplifier (HPA) is a tool similar to exoskeleton that uses hydraulic actuation to amplify the applied human force. The control objective is to make the system behave like a passive mechanical tool that interacts with the human and the environment passively with a specified power scaling factor. In our previous work, a virtual velocity coordination approach recasts the single degree-of-freedom human power amplifier control problem into a velocity coordination with a fictitious reference mechanical system. Force amplification becomes a natural consequence of the velocity coordination. In this paper, this control approach is extended for fully coupled multi-DoF systems. A passivity based control approach that uses the natural energy storage of the hydraulic actuator to take full account of the nonlinear pressure dynamics is used to define the flow requirement. Additional passive assistance dynamics are designed and implemented to enable the user to perform specific tasks more easily. Guidance is achieved using a passive velocity field controller (PVFC), and obstacle avoidance is achieved using a potential field. Experimental results demonstrate good performance on a 2-DoF Human Power Amplifier.

Commentary by Dr. Valentin Fuster
2016;():V001T07A003. doi:10.1115/DSCC2016-9823.

This paper presents an experimental setup and results on enhancing sensations of a common haptic effect -a virtual wall-induced via neuromuscular electrical stimulation (NMES). A single degree of freedom (DOF) elbow platform with position sensing was constructed. This platform supports the arm in the horizontal plane while elbow flexion and extension torques are generated by stimulation of triceps brachii or the biceps brachii muscles. The response of the system was experimentally characterized by determining the latency, and the relationship between stimulation pulse width, stimulation current, joint position and generated output torques. After system characterization, stimulation control methods to enhance haptic sensations were designed, implemented and pilot tested under a variety of virtual wall hit scenarios. Our results indicate that the wall hit trajectories and interaction were improved by control laws that initiated low intensity stimulation prior to the wall hit and utilized co-contraction for damping. The “priming” of the muscle with low intensity stimulation prior to the main stimulation improved the responsiveness of muscle contractions.

Commentary by Dr. Valentin Fuster
2016;():V001T07A004. doi:10.1115/DSCC2016-9891.

Nowadays a large number of individuals suffer from lower-limb weaknesses caused by multiple reasons, such as the gradual degeneration of musculoskeletal structure in elderly population, and the pathological losses of neuromuscular functions in stroke and spinal cord injury patients. In this paper, the design and control of a new robotic knee orthosis is presented, with the objective of assisting the user’s locomotion (primarily sit-to-stand motion) by applying an assistive torque on the knee, the largest joint in the human body. The orthosis consists of an orthosis shell and an actuation unit. The former functions as the user interface that transfers the assistive torque to the human body, while the latter generates the desired assistive torque with a motor-ball screw assembly. Through detailed design calculation, it has been demonstrated that the actuated orthotic joint is able to provide 20% of the required knee torque in the sit-to-stand motion. A controller for the robotic orthosis has also been developed by studying and emulating the knee biomechanics in the sit-to-stand motion. Benchtop testing conducted on a surrogate limb system demonstrated that the joint motion powered by the robotic orthosis is stable, smooth, and similar to the biological knee motion in the human sit-to-stand motion.

Topics: Robotics , Orthotics , Knee
Commentary by Dr. Valentin Fuster
2016;():V001T07A005. doi:10.1115/DSCC2016-9893.

This paper describes the design, development, modeling, and control of a robotic platform developed for the characterization of mechanical impedance and reflex responses of the ankle in 2 degrees-of-freedom (DOF): sagittal (dorsiflexion-plantarflexion) and frontal plane (inversion-eversion). The platform, controlled actively along both DOFs, is capable of producing rapid and strong perturbations up to an angular speed of 200°/s and peak torque of 400 Nm. This enables study of ankle impedance characteristics even during extreme task conditions such as running. The platform is designed to provide perturbations to the ankle up to an angle of 20° in sagittal plane and 10° in frontal plane, which is sufficient for all possible configurations of the ankle during postural balance and stance phase of walking. These characteristics of the platform make it ideal for both impedance and reflex characterization along both DOFs of the ankle. The platform’s performance of position control is validated under varying loads simulating various conditions of posture and locomotion. The platform’s orientation accuracy in both sagittal and frontal planes is established for various input signals including slow sinusoid inputs and rapid ramp perturbations. Implications for future ankle studies to estimate neuro-mechanical characteristics and applications to assistive and prosthetic devices are discussed.

Topics: Robotics
Commentary by Dr. Valentin Fuster
2016;():V001T07A006. doi:10.1115/DSCC2016-9895.

Recent progresses in powered lower-limb prostheses have the potential of enabling amputee users to conduct energetically demanding locomotive tasks, which are usually beyond the capability of traditional unpowered prostheses. Realizing such potential, however, requires responsive and reliable control of the power provided by prosthetic joints. In this paper, an integrated walking-stair climbing control approach is presented for transfemoral prostheses with powered knee joints. Leveraging the similarities between walking and stair climbing, this new approach adopts the general finite-state impedance control framework. Furthermore, important modifications are introduced to model the biomechanical characteristics that are beyond the capability of standard impedance control. The transition between the walking and stair-climbing modes is triggered through the real-time measurement of the spatial orientation of the user’s thigh, which provides a reliable indicator of the user’s intention of making such transition. This new control approach has been implemented on a powered knee prosthesis, and its effectiveness was demonstrated in human subject testing.

Commentary by Dr. Valentin Fuster

Battery and Oil and Gas Systems

2016;():V001T08A001. doi:10.1115/DSCC2016-9730.

This paper introduces a method to monitor battery state of health (SOH) by estimating the number of cyclable Li-ions, a health-relevant electrochemical variable. SOH monitoring is critical to battery management in particular for balancing the trade-off between maximizing system performance and minimizing battery degradation. However, SOH-related electrochemical variables cannot be directly measured non-invasively. Hence, estimation algorithms are needed to track those variables non-destructively while the battery is in use. In this paper, the extended Kalman filter (EKF) is used to estimate the number of cyclable Li-ions as an unknown battery parameter. Simulations are performed using an example parameter set for a hybrid-electric-vehicle battery whose cathode material is LiMn2O4 mixed with other Li-compounds to obtain estimation results under a typical electric vehicle current profile that consists of a 1 C constant current charge mode and a discharge current profile for an electric vehicle subject to the Urban Dynamometer Driving Schedule cycle. The simulations show promising results in estimation of the number of cyclable Li-ions using the EKF under the ideal conditions. Next, robustness of the algorithm under non-ideal conditions (i.e., with SOC estimation error, modeling error, and measurement noise) is analyzed, and it is shown that estimation of the number of cyclable Li-ions using the EKF preserves high accuracy even under these non-ideal conditions. The proposed estimation technique for the number of cyclable Li-ions can also be applied to other parameter sets and batteries with other cathode materials to monitor the SOH change resulting from any degradation mechanism that consumes cyclable Li-ions.

Topics: Ions , Batteries
Commentary by Dr. Valentin Fuster
2016;():V001T08A002. doi:10.1115/DSCC2016-9736.

Estimating the remaining useful life of lithium-ion batteries is crucial for their application as energy storage devices in stationary and automotive applications. It is therefore important to understand battery degradation based on chemistry, usage patterns, and operating environment. Different degradation mechanisms that affect performance and durability of lithium-ion batteries have been identified over the past decades. Amongst them, the solid-electrolyte interface (SEI) layer growth has been observed to be the most influential cause of capacity fading. In this paper, we introduce for the very first time, a framework that evaluates the predictive ability of physics-based macroscopic models in capturing battery dynamics as function of their state-of-health (SoH). Using data from accelerated aging experiments, we identify the applicability conditions of classical electrochemical models. This analysis is performed using a phase diagram approach that involves parameters controlling the micro-scale dynamics inside the lithium-ion cell.

Commentary by Dr. Valentin Fuster
2016;():V001T08A003. doi:10.1115/DSCC2016-9848.

Hydraulic fracturing is one of the key technologies for producing shale oil and gas. During hydraulic fracturing, a blender is used to mix sand with water and chemicals to obtain a fluidic mixture that will be pumped down a well to frack rocks. In order to achieve high-quality fracturing during a job, the blender needs to maintain its tub level as well as the density of the fluidic mixture. In this paper, an auto-tuning proportional-integral (PI) control is developed for the blender automation system to maintain the tub level of its fluidic mixture. The control system adopts a single-loop PI with gains that can be auto-tuned during a job. A relay feedback test is conducted for auto-tuning the PI gains online. The auto-tuning PI control has been successfully tested in a blender simulator. Experimental results have shown that the control performance was improved after auto-tuning and that the control system was adaptive to variation in system parameters.

Commentary by Dr. Valentin Fuster
2016;():V001T08A004. doi:10.1115/DSCC2016-9877.

This paper presents state estimation for a system of diffusion equations coupled in the boundary appearing in reduced electrochemical models of lithium-ion batteries with multiple active materials in single electrodes. The observer is synthesized from a single particle model and is based on the backstepping method for partial differential equations. The observer is suitable for state of charge estimation in battery management systems and is an extension of existing backstepping observers which were derived only for cells with electrodes of single active materials. Observer gains still can be computed analytically in terms of Bessel and modified Bessel functions. This extension is motivated by the trend in cell manufacturing to use multiple active materials to combine power and energy characteristics or reduce degradation.

Commentary by Dr. Valentin Fuster
2016;():V001T08A005. doi:10.1115/DSCC2016-9931.

The risk of kick and lost circulation at the wellbore open-hole increases with water depth due to narrow pressure margins. The safety of drilling operations and mitigation of risk to drilling personnel, equipment, and the environment hinges on the ability of the drilling crew to detect these undesirable events in their early stages and quickly bring the well under control. This paper presents an approach for the estimation of unobserved bottom-hole phenomena during drilling ahead operations by means of combining multiple surface measurements with predictions from a hydraulic model of the well. Bond graph technique is used to formulate a lumped-parameter hydraulic model of the drilling ahead process, the model is linearized, and an estimation method is applied to the proposed stochastic model. This methodology was tested offline with drilling ahead data from a well where a kick occurred and the results showed kick detection earlier than traditional methods allowed.

Commentary by Dr. Valentin Fuster

Bioengineering Applications

2016;():V001T09A001. doi:10.1115/DSCC2016-9667.

This paper introduces a new design for individually-controlled magnetic artificial cilia for use in microfluidic devices. The design has been implemented using a low-cost prototype that can be fabricated using polydimethylsiloxane (PDMS) and off-the-shelf parts, and achieves large cilium deflections (59% of the cilium length). Experimental results show that exploiting the individual control leads to faster mixing (38% reduction in mixing time) than when operating the device in a simultaneous-actuation mode with the same average cilium beat frequency.

Commentary by Dr. Valentin Fuster
2016;():V001T09A002. doi:10.1115/DSCC2016-9742.

Physical activity is an important physiological information which should be taken into account by artificial pancreas to achieve optimal control of blood glucose in Type 1 Diabetes patients. An accurate glucose dynamic model with physical activity as an additional input is highly desirable for the next generation artificial pancreas. In this paper, we present a nonlinear data-driven model that captures both the insulin-independent and -dependent effect of physical activity, especially the prolonged effect of physical activity on insulin sensitivity that can last 24–48 hours post exercise. The model was identified and validated using data sets generated by a physiological glucose-exercise model under a clinical training protocol. Compared to modeling the effect of physical activity as a linear additive term only in a glucose dynamic equation, the proposed nonlinear model showed significant improvement of prediction accuracy in all three metrics, particularly in large prediction horizons (P < 0.05). Further investigation in time-series data indicates that the improvement mainly resulted from the better prediction of glucose around the first meal time after exercise (6 to 8 hours after the meal was taken).

Commentary by Dr. Valentin Fuster
2016;():V001T09A003. doi:10.1115/DSCC2016-9755.

This work considers a 3-state nonlinear model with two inputs for the controlled dynamics of the human immunodeficiency virus (HIV). The three states represent the number of healthy and unhealthy T-cells as well as the number of free virus particles in a person’s body. The two inputs represent two different types of anti-retroviral drugs that are available to treat a person infected with the virus. These inputs can be used to create a stable nominal point that is much healthier for the patient than the open-loop stable equilibrium point. The goal of this paper is to use the inputs, which are subject to constraints, to efficiently and accurately reach the desired nominal point despite parameter uncertainties. We have designed an integral sliding mode control law to achieve this goal, and simulations are presented to demonstrate the performance of the controller.

Commentary by Dr. Valentin Fuster
2016;():V001T09A004. doi:10.1115/DSCC2016-9767.

Insect flight has gained wide interests in both biology and engineering communities in the past decades regarding its aerodynamics, sensing and flight control. However, studying insect flight experimentally remains a challenge in both free-flight and tethered-flight settings. In free flight experiments, due to highly unpredictable and fast flight behavior of flying insects, it is difficult to apply controlled sensory inputs to their flight system for system identification and modeling analyses. In tethered flight experiments, constrained whole body movement results in silenced proprioceptive feedback therefore breaks the flight control loop and does not reveal any flight dynamics. Therefore, this work aims to develop a novel insect tether system using magnetic levitation. Such a system magnetically fixes an insect in space but allows it to rotate freely about yaw axis with minimal interference from mechanical constraints. This paper presents the development, analysis and feedback control of this system and finally test its performance using a hawkmoth (Manduca Sexta). In addition, a system identification of the magnetic levitation system and detailed analysis in closed-loop stability and performance are provided. In the future, the insect tether system will be applied to study the insect flight aerodynamics, sensing and control.

Commentary by Dr. Valentin Fuster
2016;():V001T09A005. doi:10.1115/DSCC2016-9768.

This article presents the investigation of the dynamic behavior of the cytoskeleton of live human cells, enabled by a recently-developed control-based approach on scanning probe microscope (SPM). Mechanical behaviors of live cells play an important role in various cell physiological and pathological activities, and have been studied via various techniques and approaches. Studies of evolutions of mechanical properties of live cell, however, are still rather limited and scarce, due to the limitations of current instruments including SPM for single cellular measurements. Particularly, currently nanomechanical measurements using SPM is too slow to excite the mechanical behavior and then measure the corresponding response of life biological species over a large frequency range (broadband). Moreover, large uncertainty is induced in the in-liquid nanomechanical measurement using SPM, as in the indentation quantification, the effects of the acceleration force from the cantilever motion and the hydrodynamic force are not accounted for. The main contribution of this article is the use of a control-based nanomechanical protocol to interrogate the viscoelasticity oscillation of live human prostate cancer cell (PC-3 cells) and its dependence on myosin activities. The experiment results show that as the oscillation of static elastic modulus reported earlier in the literature, the oscillation of dynamic viscoelastic modulus measured is also periodic with a 200-second period. Moreover, as the elastic modulus oscillation, both the amplitude and the period of the viscoelasticity oscillation also strongly depend on the myosin activities, and closely regulated by the calcium density of the cytoskeleton.

Topics: Viscoelasticity
Commentary by Dr. Valentin Fuster
2016;():V001T09A006. doi:10.1115/DSCC2016-9920.

This paper reviews the design of smart shoes, a wearable device that measures ground contact forces (GCFs) for gait analysis. Smart shoes utilize four coils of silicone tubes adhered directly underneath the shoe insole at key points of interest. Air pressure sensors connect to each tube coil to measure pressure changes caused by compression. This paper presents static and dynamic calibration performed on each sensing coil to establish a model of internal pressure and the GCF. Based on the model, a phase lead filter is designed to account for the hysteresis effect and visco-elastic properties of the silicone tube in order to provide accurate GCF measurements. To design this filter, the air bladder is modeled using a standard linear solid (SLS) model. The prediction error minimization (PEM) algorithm is then implemented to identify the continuous-time transfer function of this SLS model, which is then transformed to discrete time domain to implement in a digital processor. Mechanical characterization and testing on a healthy subject are performed to validate the model and its capability to compensate for hysteresis in GCF measurement.

Commentary by Dr. Valentin Fuster

Biomedical and Neural Systems Modeling, Diagnostics and Healthcare

2016;():V001T10A001. doi:10.1115/DSCC2016-9785.

This paper presents a model-based system identification approach to estimation of central aortic blood pressure waveform from non-invasive cuff pressure oscillation signals. First, we developed a mathematical model that can reproduce the relationship between central aortic blood pressure waveform and non-invasive cuff pressure oscillation signals at diametric locations by combining models to represent wave propagation in the artery, arterial pressure-volume relationship, and mechanics of the occlusive cuff. Second, we formulated the problem of estimating central aortic blood pressure waveform from non-invasive cuff pressure oscillation signals into a system identification problem. Third, we showed the proof-of-concept of the approach using simulated central aortic blood pressure waveform and cuff pressure oscillation signals. Finally, we illustrated the feasibility of the approach using central aortic blood pressure waveform and cuff pressure oscillation signals collected from a human subject. We showed that the proposed approach could estimate central aortic blood pressure waveform with accuracy: the root-mean-squared error associated with the central aortic blood pressure waveform was 1.7 mmHg (amounting to 1.6 % of the underlying mean blood pressure) while the errors associated with central aortic systolic and pulse pressures were −0.4 mmHg and −1.5 mmHg (amounting to −0.3 % and −1.4 % of the underlying mean blood pressure).

Commentary by Dr. Valentin Fuster
2016;():V001T10A002. doi:10.1115/DSCC2016-9828.

The application of surface electromyography (sEMG) technique for muscle fatigue studies is gaining importance in the field of clinical rehabilitation and sports medicine. These sEMG signals are highly nonstationary and exhibit scale-invariant self-similarity structure. The fractal analysis can estimate the scale invariance in the form of fractal dimension (FD) using monofractal (global single FD) or multifractal (local varying FD) algorithms. A comprehensive study of sEMG signal for muscle fatigue using both multifractal and monofractal FD features have not been established in the literature. In this work, an attempt has been made to differentiate sEMG signals recorded nonfatigue and fatigue conditions using monofractal and multifractal algorithms, and machine learning methods. For this purpose, sEMG signals have been recorded from biceps brachii muscles of fifty eight healthy subjects using a standard protocol. The signals of nonfatigue and fatigue region were subjected to eight monofractal (Higuchi, Katz, Petrosian, Sevcik, box counting, multi-resolution length, Hurst and power spectrum density) and two multifractal (detrended fluctuating and detrended moving average) algorithms and 28 FD features were extracted. The features were ranked using conventional and genetic algorithms, and a subset of FD features were further subjected to Naïve Bayes (NB), Logistic Regression (LR) and Multilayer Perceptron (MLP) classifiers. The results show that all fractal features are statistically significant. The classification accuracy using feature subset of conventional method is observed to be from 83% to 88%. The highest accuracy of 93.96% was achieved using genetic algorithm and LR classifier combination. The result demonstrated that the performance of multifractal FD features to be more suitable for sEMG signals as compared to monofractal FD features. The fractal analysis of sEMG signals appears to be a very promising biomarker for muscle fatigue classification and can be extended to detection of fatigue onset in varied neuromuscular conditions.

Commentary by Dr. Valentin Fuster
2016;():V001T10A003. doi:10.1115/DSCC2016-9836.

The purpose of this study is to numerically evaluate the performance of information entropy in electroencephalography (EEG) signal analysis. In particular, we use EEG data from an Alzheimer’s disease (AD) pilot study and apply several wavelet functions to determine the signals’ time and frequency characteristics. The wavelet entropy and wavelet sample entropy of the continuous wavelet transformed data are then determined at various scale ranges corresponding to major brain frequency bands. Non-parametric statistical analysis is then used to compare the entropy features of the EEG data obtained in trials with AD patients and age-matched healthy normal subjects under resting eyes-closed (EC) and eye-open (EO) conditions. The effectiveness and reliability of both choice of wavelet functions and the parameters used in wavelet sample entropy calculations are discussed and the ideal choices are identified. The result shows that, when applied to wavelet transformed filtered data, information entropy can be effective in determining EEG discriminant features, after selecting the best wavelet functions and window size of the sample entropy.

Commentary by Dr. Valentin Fuster
2016;():V001T10A004. doi:10.1115/DSCC2016-9849.

Sleepiness has been considered as one of the major contributors to driver error that causes many automobile accidents. Among various technologies developed to address this issue, the electrooculography (EOG) signal is considered most suitable for sleepiness detection. It is simple, and resilient to environmental factors such as light intensity and driver movement. Most importantly, the physiological signal changes in an early stage and can be used to detect the on-set of human sleepiness. In this paper, we introduce the development of a wearable sleepiness detection system based on analyzing EOG signal dynamics. The system includes wearable sensors, amplifying and transmitting circuits, and a smart phone that could alarm the driver if sleepiness is detected. In this system, the EOG signal is considered as a neurophysiological response of the oculomotor system. Blink signatures are extracted from the EOG signal. It was found that the poles of a linearized blinking motion associated with an alert state are different from those associated with a sleepy state. Based on this understanding, an algorithm to detect the driver’s sleepiness was developed. A proof of concept device design has been completed. This system will help a driver to correct the behavior, and ultimately saves lives.

Commentary by Dr. Valentin Fuster
2016;():V001T10A005. doi:10.1115/DSCC2016-9863.

Recent mathematical models of human posture have been explored to better understand the space of control parameters that result in stable upright balance. These models have demonstrated that there are two types of instabilities — a leaning instability and an instability leading to excessive oscillation. While these models provide insight into the stability of upright bipedal stance, they are not sufficient for individuals that require the aid of assistive technologies, such as a passive-cane or a walker. Without a valid model one is unable to understand the control parameters required for maintain upright posture or if similar instabilities even exist when assistive technologies are used. Therefore in this study, we developed a mathematical model of human posture while using a passive-cane to examine the nonlinear dynamics of stance. First, we developed a simple mathematical model of cane assisted human stance by adapting the inverted pendulum model of Chagdes et al., [1]. We modeled the human body, upper arm, forearm, cane, and ground as a two-degree-of-freedom, five-bar-linkage with pin joints representing the ankle, shoulder, elbow, and wrist joints. Second, we investigate upright stability in the parameter space of feedback gain and time-delay. We hypothesize that the analysis will show similar instabilities compared to that of a human standing without assistive technology. We also hypothesize that the space of control parameters which stabilize upright equilibrium posture will increase when a cane is incorporated. This study has two potential applications. First, the developed mathematical model could allow clinicians to better assess technology assisted balance and if needed help clinicians to customize a treatment plan for an individual that allows them to avoid unstable postural dynamics. Second, the mathematical model can be used to design customized assistive technology for people of difference physical properties and impairments.

Commentary by Dr. Valentin Fuster

Control and Monitoring of Vibratory Systems

2016;():V001T11A001. doi:10.1115/DSCC2016-9617.

A new hybrid control methodology is presented for vibration suppression in flexible structures, where an active actuator is used to assist a nearby semi-active device to achieve a control performance close to that of a fully active system. The clipping phenomenon, typical of semi-active control, is reduced to a large extent by the proposed hybrid controller. The immersion and invariance methodology along with sliding mode control is used to create the hybrid controller. The result is that as the semi-active controller switches off in the hybrid controller, the active actuator injects the required energy into the system. A two degree of freedom system with cubic stiffness is used as an example system. Both simulation and experiment data are presented to demonstrate the usefulness of the proposed idea. The proposed hybrid controller shows robust results as compared to just using a semi-active controller.

Commentary by Dr. Valentin Fuster
2016;():V001T11A002. doi:10.1115/DSCC2016-9711.

This work investigates the voltage response of superharmonic resonance of second order of electrostatically actuated Micro-Electro-Mechanical Systems (MEMS) resonator cantilevers. The results of this work can be used for mass sensors design. The MEMS device consists of MEMS resonator cantilever over a parallel ground plate (electrode) under Alternating Current (AC) voltage. The AC voltage is of frequency near one fourth of the natural frequency of the resonator which leads to the superharmonic resonance of second order. The AC voltage produces an electrostatic force in the category of hard excitations, i.e. for small voltages the resonance is not present while for large voltages resonance occurs and bifurcation points are born. The forces acting on the resonator are electrostatic and damping. The damping force is assumed linear. The Casimir effect and van der Waals effect are negligible for a gap, i.e. the distance between the undeformed resonator and the ground plate, greater than one micrometer and 50 nanometers, respectively, which is the case in this research. The dimensional equation of motion is nondimensionalized by choosing the gap as reference length for deflections, the length of the resonator for the axial coordinate, and reference time based on the characteristics of the structure. The resulting dimensionless equation includes dimensionless parameters (coefficients) such as voltage parameter and damping parameter very important in characterizing the voltage-amplitude response of the structure. The Method of Multiple Scales (MMS) is used to find a solution of the differential equation of motion. MMS transforms the nonlinear partial differential equation of motion into two simpler problems, namely zero-order and first-order. In this work, since the structure is under hard excitations the electrostatic force must be in the zero-order problem. The assumption made in this investigation is that the dimensionless amplitudes are under 0.4 of the gap, and therefore all the terms in the Taylor expansion of the electrostatic force proportional to the deflection or its powers are small enough to be in the first-order problem. This way the zero-order problem solution includes the mode of vibration of the structure, i.e. natural frequency and mode shape, resulting from the homogeneous differential equation, as well as particular solutions due to the nonhomogeneous terms. This solution is then used in the first-order problem to find the voltage-amplitude response of the structure. The influences of frequency and damping on the response are investigated. This work opens the door of using smaller AC frequencies for MEMS resonator sensors.

Commentary by Dr. Valentin Fuster
2016;():V001T11A003. doi:10.1115/DSCC2016-9714.

This paper investigates the frequency response of microplates under electrostatic actuation. The microplate is parallel to a fixed ground plate. The electrostatic force that actuates the system is given by both Alternate Current (AC) and Direct Current (DC) voltages. The AC frequency is set to be near half natural frequency of the structure. Damping influence is also investigated in this paper. The method of investigation is Reduced Order Model. The effects of various parameters on the response of the structure are reported.

Commentary by Dr. Valentin Fuster
2016;():V001T11A004. doi:10.1115/DSCC2016-9749.

A new consensus based active vibration controller for distributed parameter system is developed and presented in this paper. The controller is made of multiple parallel first-order integrator filters that have consensus goal of suppressing resonant modes of an oscillating structure. Each actuator on the structure is designed to work as a node of the network. Together through the consensus communication topology of the controller the structural vibrations are suppressed. To validate the results a numerical study has been performed on a flexible cantilever beam structure. The results presented in this paper will show that the new controller works effectively, and is robust to failures and inconsistencies in the control system.

Commentary by Dr. Valentin Fuster
2016;():V001T11A005. doi:10.1115/DSCC2016-9874.

Identifying damages in mechanical structures in advance is essential part of preventing catastrophic losses. Among several non-destructive methods, the vibration-based method, which utilizes global characteristics of the structures, has several advantages such as not requiring prior information on possible damage location and physical access to it. In the meantime, the mechanical structures are inevitably subject to uncertainties, whose distribution is often unknown in practical situations due to such as limited amount of available data. Uncertainties are treated as interval uncertainty in such cases. In this regard, this study presents vibration-based damage identification under interval uncertainty. To obtain reliable result, this research does not assume any random distribution, e.g., uniform distribution, inside interval. Since detected damage is not assumed to be monotonic function with respect to interval uncertainty either, traditional fuzzy interval arithmetic is not applicable. Instead, we first carry out exhaustive search to see the effect of the interval uncertainty on the identified damage; i.e., discretizing interval uncertainty into sub-intervals and executing damage identification under all possible combinations to see the effect of the interval uncertainty on the identified damage. We then develop the unique algorithm based on M-H algorithm to facility computational efficiency.

Commentary by Dr. Valentin Fuster
2016;():V001T11A006. doi:10.1115/DSCC2016-9881.

The objective of this paper is to analytically study the nonlinear behavior of variable cross-section beam flexures interconnecting an eccentric rigid body. Hamilton’s principle is utilized to obtain the partial differential equations governing the nonlinear vibration of the system as well as the corresponding boundary conditions. Using a single mode approximation, the governing equations are reduced to a set of two nonlinear ordinary differential equations in terms of end displacement components of the beam which are coupled due to the presence of the transverse eccentricity. The method of multiple scales are employed to obtain parametric closed-form solutions. The obtained analytical results are compared with the numerical ones and excellent agreement is observed. These analytical expressions provide design insights for modeling and optimization of more complex flexure mechanisms for improved dynamic performances.

Commentary by Dr. Valentin Fuster

Diagnostics and Detection

2016;():V001T12A001. doi:10.1115/DSCC2016-9631.

In this paper, we consider the general problem of data classification, and focus on the development of suitable approaches for automotive calibration applications. Specifically, we propose two methods for nonlinear data classification: the first one is to find the smallest piecewise linear (PWL) region for a given data set and the second one is to construct piecewise quadratic (PWQ) boundary to separate two data sets . In both methods, the construction of the boundary curve is formulated as convex optimization problems that can be solved efficiently. Our approaches can incorporate prior information about data distribution and allow fixed structure of the decision boundary from different data sets with similar generating sources. We demonstrate the efficiency and effectiveness of the approaches with an application to calibration identification of vehicle rollover detection algorithm.

Topics: Vehicles , Calibration
Commentary by Dr. Valentin Fuster
2016;():V001T12A002. doi:10.1115/DSCC2016-9725.

With the motivation to develop Condition Based Maintenance (CBM) strategies for the automotive vehicles, this paper considers a data-driven approach to the prognostics of the automotive fuel pumps. Focusing on the returnless type fuel delivery systems, our approach is based on estimating the fuel pump workload based on the model learned from the past driving history. Statistical reliability models are then exploited to estimate failure probability. These models are formulated in terms of the workload and updated from data available from vehicles in the field. Numerical examples which illustrate the proposed methodology are reported. Compared to alternative approaches, which are based on detailed physics-based degradation modeling and/or electrical signal analysis, our approach is data-driven, exploits connected vehicle analytics and reliability-based modeling, and has a potential to lead to simpler implementations.

Commentary by Dr. Valentin Fuster
2016;():V001T12A003. doi:10.1115/DSCC2016-9858.

The alarmingly degrading state of transportation infrastructures combined with their key societal and economic importance calls for automatic condition assessment methods to facilitate smart management of maintenance and repairs. In particular, scalable data-driven approaches is of great interest, because it can deal with large volume of streaming data without requiring models that can be inaccurate and computationally expensive to run. Properly designed, a data-driven methodology could enable fast and automatic evaluation of infrastructures, discovery of causal dependencies among various sub-system dynamic responses, and inference and decision making with uncertainties and lack of labeled data. In this work, a spatiotemporal pattern network (STPN) strategy built on symbolic dynamic filtering (SDF) is applied to explore spatiotemporal behaviors in bridge network. Data from strain gauges installed on two bridges are simulated by finite element method, and the causality among strain data in spatial and temporal resolutions is analyzed. Case studies are conducted for truck identification and damage detection from simulation data. Results show significant capabilities of the proposed approach in: (i) capturing spatiotemporal features to discover causality between bridges (geographically close), (ii) robustness to noise in data for feature extraction, and (iii) detecting and localizing damage via the comparison of behaviors within the bridge network.

Commentary by Dr. Valentin Fuster
2016;():V001T12A004. doi:10.1115/DSCC2016-9888.

A significant cost in industrial fluid systems is associated with system maintenance and unexpected downtime due to the troubleshooting and repair of system faults. In large complex systems, faults are often difficult to identify and predict, but recent advances in technology have enabled low-cost wireless-capable sensors, which can provide an unprecedented amount of data to system owners. Combining this data with knowledge of the system dynamics, and a methodical structural analysis-based fault diagnosis approach, provides new opportunities in the field of fault diagnostics. The goal of the project described in this paper is to develop methods to detect and diagnose faults in industrial fluid systems in order to minimize downtime costs and energy losses, with minimum system capital cost in mind.

This paper summarizes a portion of the first phase of this project, which focuses on fault modeling of a small-scale compressed-air test bench. Specifically, the design of the test bench, creation and validation of a flow model used to understand the test bench system dynamics, fault and structural analysis of an element of the test bench (an orifice-pipe combination), and the generation of a sensor installation guide that indicates what type and where to place sensors to detect and isolate faults that may occur. The sensor installation guide provided insight that all faults in the orifice-pipe element were detectable and isolable to the desired level with only a pressure sensor in the orifice and pipe. Orifice and orifice pressure sensor faults were uniquely isolable, but pipe and pipe pressure sensor faults were only isolable.

Commentary by Dr. Valentin Fuster
2016;():V001T12A005. doi:10.1115/DSCC2016-9907.

Flame dynamics and combustion instability is a complex problem involving different non-linearities. Combustion instability has several detrimental effects on flight-propulsion dynamics and structural integrity of gas turbines and any such spaces where combustion takes places internally, primarily in internal combustion engines. The description of coherent features of fluid flow in such cases is essential to our understanding of the flame dynamics and propagation processes. A method that is able to extract dynamic information from flow fields that are generated by a direct numerical simulation or visualized in a physical experiment (like in the case discussed in this paper) is Dynamic Mode Decomposition. This paper presents such a feature extraction and stability analysis of hi-speed combustion flames using Dynamic Mode Decomposition and it’s sparsity promoting variant. Extensive experimental data was collected in a swirl-stabilized dump combustor at various operating conditions (e.g. premixing level and flow velocity) for analysing the flame stability conditions.

Topics: Combustion , Flames
Commentary by Dr. Valentin Fuster

Energy Harvesting

2016;():V001T13A001. doi:10.1115/DSCC2016-9723.

An electromagnetic energy harvester for railroad application featuring anchorless mounting is presented along with system modeling and in-field testing results. The spring-reset mechanism allows harvester to be installed on any railroad sleepers without any change to the existing ballast and foundation. A railway train-track-harvester model is established for design validation and risk assessment. Within harvester, train-pass induced track vibration is translated to rotation and then conditioned thru mechanical motion rectifier mechanism for generator to operate in one direction with a relatively steady speed. Particularly, in-field test demonstrates that harvester can harness electric at an average power of 6.9 Watts and a peak power of 60 Watts when test cargo train running at 40 mph (64 km/h). Dynamic modeling of the vehicle-rail system are also developed to estimate the displacement of the sleeper under various train weight and speed. Simulation result based on typical train parameters are provided and compared with infield test result.

Commentary by Dr. Valentin Fuster
2016;():V001T13A002. doi:10.1115/DSCC2016-9748.

Recovering and regenerating power in automotive applications has drawn significant interest recently. A car-suspension system can be modeled as a 2-DOF mass-spring-damper system. Active control used for the car suspension system produces results superior to other methods. In this study, a 3-phase linear generator is used to harvest energy and suppress vibration on a quarter-car suspension setup. The suspension system is analyzed to estimate the harvestable power and damping capability of the generator. Analysis for the generator and its efficiency are presented. Harvestable power of around 105 mW was achieved at a 3.5-Hz input disturbance. The regenerative suspension system can reduce the vibration of the sprung-mass acceleration by up to 22% in an indexed performance. Around 8.4 W used to drive the motor in active control was saved when the regenerative system was used. As a result, much energy can be saved by switching from the active to the energy-harvesting mode. A more efficient system can be designed by matching the mechanical and electromagnetic (EM) damping.

Commentary by Dr. Valentin Fuster

Estimation and Identification

2016;():V001T14A001. doi:10.1115/DSCC2016-9686.

Citizen science projects are becoming increasingly popular, yet they typically rely on only a small portion of users for the majority of contribution. In this paper, we propose a model for citizen scientist contribution in an online image tagging task. The model describes participant contribution in response to the performance of a virtual peer, the behavior of which can be controlled by the experimenter. Experimental trials where the virtual peer behaves independent of the participant are used to calibrate the model. The model’s ability to predict participant performance is then verified in a closed-loop condition, where the behavior of the virtual peer is explicitly dependant on the performance of the participant. We foresee this model being a useful tool in the design of web-based citizen science projects, where the behavior of a virtual peer can be used to modulate the performance of contributors in an effort to increase overall levels of contribution.

Topics: Design
Commentary by Dr. Valentin Fuster
2016;():V001T14A002. doi:10.1115/DSCC2016-9689.

The magnetic field of a moving piston can be used for real-time estimation of its position. Piston position estimation is useful for a number of automation and performance improvement applications in hydraulic actuators, pneumatic cylinders, and I.C. engines. A significant challenge to magnetic field based position estimation comes from disturbances due to unexpected ferromagnetic objects coming close to the sensors.

This paper develops a new disturbance estimation method based on modeling the magnetic disturbance as a dipole with unknown location and magnitude. A Truncated Interval Unscented Kalman Filter (TIUKF) is used to estimate all the parameters of this unknown dipole. Experimental data from a pneumatic actuator is used to verify the performance of the developed estimator. Experimental results show that the developed estimator is significantly superior to a linear magnetic field model based disturbance estimator.

Topics: Pistons
Commentary by Dr. Valentin Fuster
2016;():V001T14A003. doi:10.1115/DSCC2016-9715.

This paper presents an algorithm that estimates a measured output, which has been converted to a binary representation in order to transmit the information via a digital communication channel. The purpose of the estimator is to mitigate the effect of the delay due to transmitting the information across the communication channel. The algorithm uses each bit of the output sample to estimate the output sample value as it is being transmitted, and updates the control signal based on this estimate. Some basic analysis of this algorithm is presented along with simulations and a discussion on situations in which it could be beneficial.

Commentary by Dr. Valentin Fuster
2016;():V001T14A004. doi:10.1115/DSCC2016-9746.

This work elucidates another theoretical property of the ubiquitous extended Kalman filter by analyzing the energy gain of the continuous-time extended Kalman filter used as a nonlinear observer in the presence of finite-energy disturbances. The analysis provides a bound on the ratio of estimation error energy to disturbance energy, which shows that the extended Kalman filter inherently has the H-property along with being the locally optimal minimum variance estimator. A special case of this result is also shown to be the H2-property of the extended Kalman filter.

Topics: Kalman filters
Commentary by Dr. Valentin Fuster
2016;():V001T14A005. doi:10.1115/DSCC2016-9854.

Automated methods for deriving dynamic models from frequency response data of high-order dynamics systems are the default choice of most engineers. However, these methods themselves often require manual tuning of weighting parameters, a priori selection of system order, and even by hand removal of extraneous dynamics. On the other hand, manually matching complicated features in the Bode plot of the frequency response of high-order system is difficult with conventional first and second order numerators and denominators. In this papers we present a manual technique for systematically creating a dynamic model from Bode plots of frequency response data with complicated features. We apply the method to identifying dynamics of a piezoelectric stage holding the sample of an atomic force microscope (AFM). We show the manual method works better than the tfest command of Matlab™ for this example system.

Commentary by Dr. Valentin Fuster

Fuel Cells/Energy Storage

2016;():V001T15A001. doi:10.1115/DSCC2016-9750.

This paper derives analytic expressions for both the mean and variance of battery state of charge (SOC) estimation error, assuming a least squares estimation law. The paper examines three sources of estimation error, namely: (i) voltage measurement errors (both bias and noise), (ii) current measurement bias, and (iii) mismatch between the order of the battery model used for estimation and the true order of the battery’s dynamics. There is already a rich literature on quantifying battery SOC estimation errors for different estimator designs. The novelty of this paper stems from its extensive examination of both the expected SOC estimation bias and noise, for a least squares estimation algorithm, in the presence of three different fundamental sources of these estimation errors. We show, both analytically and using Monte Carlo simulation, that under reasonable operating conditions, the expected bias in SOC estimation for lithium-ion batteries is dominant compared to the expected estimation variance. This leads to the important insight that quantifying SOC estimation variance using Fisher information furnishes overly optimistic predictions of achievable SOC estimation accuracy.

Topics: Errors , Batteries
Commentary by Dr. Valentin Fuster
2016;():V001T15A002. doi:10.1115/DSCC2016-9754.

This paper proposes a novel approach for integrating battery storage into photovoltaic (PV) arrays. The approach relies on the integration of PV arrays with individual batteries to form “hybrid cells” that are then assembled into series strings. We use Lyapunov analysis to show that the proposed hybrid strings are globally asymptotically self-balancing, meaning that initial variations in state of charge (SOC), no matter how large, converge to zero. The PV subsystem serves as a negative feedback path that guarantees self-balancing without requiring dedicated balancing circuits. This significantly reduces the cost of the power electronics needed for integrating batteries into PV farms, compared to typical integration topologies. The paper uses local linearization to approximate the balancing rate, thereby highlighting its independence of battery pack length and elucidating its dependence on subsystem sizing. Finally, a simulation study validates the paper’s theoretical insights regarding self-balancing, and examines its sensitivity to parameter heterogeneities.

Commentary by Dr. Valentin Fuster
2016;():V001T15A003. doi:10.1115/DSCC2016-9800.

This paper presents a dynamic programming approach to optimize energy cost of multiple interacting household appliances such as air conditioning systems and refrigerators with temperature flexibility, under time varying electricity price signals. We adopt a first order differential equation model with a binary (ON-OFF) switching control function for each load. An energy cost minimization problem is then formulated with a pair of constraints on the temperature lower and upper bounds, as well as an equality condition on the initial and final temperature states. We use dynamic programming to compute cost-optimal control inputs and temperature trajectories for a given electricity price profile and ambient temperature condition. To account for temperature deviation from its desired setpoint, a quadratic temperature deviation penalty is added to the cost function. Moreover, to minimize the control input chattering for equipment protection, the cost function is expanded to also minimize the number of on-off switching events. Results for the different weighting combinations of the optimization objectives provide useful insights on the optimal operation of individual and multiple interacting HVAC loads. In particular, we observe that the loads are desynchronized under the cost-optimal operation, in the presence of local (renewable) power generation. The presented optimization algorithm and observed results can lead to the development of novel model predictive and rule-based feedback control policies for optimal energy management in households.

Commentary by Dr. Valentin Fuster
2016;():V001T15A004. doi:10.1115/DSCC2016-9866.

In this paper, a semi-empirical aging model of lithium-ion pouch cells containing blended spinel and layered-oxide positive electrodes is calibrated using aging campaigns. Sensitivity analysis is done on this model to identify the effect of parameter variations on the State of Health (SOH) prediction. The sensitivity analysis shows that the aging model alone is not robust enough to perform long term predictions, hence we propose to use online parameter estimation algorithms to adapt the model parameters. Four different estimation methods are compared using aging campaign. It is demonstrated that the estimation algorithms improve aging model leading to significant improvement in Remaining Useful Life (RUL) prediction.

Commentary by Dr. Valentin Fuster
2016;():V001T15A005. doi:10.1115/DSCC2016-9884.

For a compressed air energy storage (CAES) system to be competitive for the electrical grid, the air compressor/expander must be capable of high pressure, efficient and power dense. However, there is a trade-off between efficiency and power density mediated by heat transfer, such that as the process time increases, efficiency increases at the expense of decreasing power. This trade-off can be mitigated in a liquid (water) piston air-compressor/expander with enhanced heat transfer. However, in the past, dry air has been assumed in the design and analysis of the compression/expansion process. This paper investigates the effect of moisture on the compression efficiency and power. Evaporation and condensation of water play contradictory roles — while evaporation absorbs latent heat enhancing cooling, the tiny water droplets that form as water condenses also increase the apparent heat capacity. To investigate the effect of moisture, a 0-D numerical model that takes into account the water evaporation/condensation and water droplets have been developed. Results show that inclusion of moisture improves the efficiency-power trade-off minimally at lower flow rates, high efficiency cases, and more significantly at higher flow rates, lower efficiency cases. The improvement is primarily attributed to the increase in apparent heat capacity due to the increased propensity of water to evaporate.

Commentary by Dr. Valentin Fuster

Intelligent Transportation

2016;():V001T16A001. doi:10.1115/DSCC2016-9622.

Feedback controls are important to the improvement of dynamic performance of high-speed trains. However, designing an active control for these vehicles is a very challenging task because the control system is usually under-actuated and has to meet multiple conflicting objectives. Examples of conflicting objectives include designing a highly relative stable system while minimizing the control efforts or maximizing the capability of the system to reject external disturbances. In addition, the mathematical models of these systems are not completely controllable and observable. This paper studies multi-objective optimal design of feedback controls for a sub-system of high-speed trains, i.e. the bogie system.

The bogie system can be decomposed such that the observable and controllable components of the model are used to stabilize the internal states and therefore the overall system. A linear mathematical model of the system is used in the design. The controllable and the observable states of the model are separated to form a state-feedback control to drive the internal modes and the whole system to stability. A multi-objective genetic algorithm is used to search for the feedback control gains to optimize three objectives: the Frobenius norm of the control law, relative stability and the disturbance rejection. The solutions of the multi-objective optimization provide various trade-offs among the objectives. Numerical simulations show that the proposed control designs can stabilize the system even at a high critical speed of 500 km/h.

Commentary by Dr. Valentin Fuster
2016;():V001T16A002. doi:10.1115/DSCC2016-9641.

In this paper, we present a fuel efficient control strategy for a group of connected hybrid electric vehicles (HEVs) in urban road conditions. A hierarchical control architecture is proposed where the higher level controller is located at traffic signal light while the lower level controllers are equipped on each HEV. The higher level controller utilizes Signal Phase and Timing (SPAT) information from the traffic lights to generate target velocities for every HEV, which allows a maximum number of vehicles pass the intersection at given green light window. Model Predictive Control (MPC) is used to track the target velocity and evaluate the energy efficient velocity profile for every vehicle for a given horizon. Each lower level controller then follows the velocity profile (from the higher level controller) in a fuel efficient fashion using adaptive equivalent consumption minimization strategy (A-ECMS). The lower level controller also feeds the average recuperation efficiency in a certain time window back to the higher level controller, thus affects the future velocity profile evaluation from the higher level controller, which is the major contribution of this paper. In this paper, the HEV model is developed based on Autonomie software and the simulation results show the effectiveness of our proposed approach.

Commentary by Dr. Valentin Fuster
2016;():V001T16A003. doi:10.1115/DSCC2016-9822.

Connected vehicles are one of the promising technologies for future Intelligent Transportation Systems (ITS). Despite being the potentially beneficial in creating an efficient, sustainable and green transportation system, connected vehicles presents a set of specific challenges from safety and reliability standpoint. The first challenge arises from the information lost due to unreliable communication network which affects the control/management system of the individual vehicles and the overall system. Secondly, faulty sensors can affect the individual vehicle’s safe operation and in turn will create a potentially unsafe node in the vehicular network. Therefore, it is of utmost importance to take these issues into consideration while designing the control/management algorithms of the individual vehicles as a part of connected vehicle system. In this paper, we consider a connected vehicle system under Co-operative Adaptive Cruise Control (CACC) and propose a diagnostic scheme that deals with these aforementioned challenges. The effectiveness of the overall diagnostic scheme is verified via simulation studies.

Commentary by Dr. Valentin Fuster
2016;():V001T16A004. doi:10.1115/DSCC2016-9880.

Exhaust gas recirculation (EGR) has become an integral part of the NOx emission reduction mechanisms utilized in modern combustion engines. However, capabilities to recirculate the processed gas are oftentimes compromised by the inability to surmount the pressure differential between the intake and exhaust manifolds. The issue is dealt with successfully by the turbocharged EGR system discussed in this article. The increased complexity of such an EGR system requires a multivariable control system in order to achieve the EGR set-point tracking across the desired operating range. In this work, model predictive control is used to naturally incorporate the information about system internal couplings and constraints via the prediction model. The states of the partially observed EGR system required for the feedback control are recovered by using the unscented Kalman filter. Finally, the designed unscented MPC (UMPC) system is validated by numerical simulations.

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