System Identification and Nonlinear Observers

2009;():1-8. doi:10.1115/DSCC2009-2550.

It was shown recently that parameter estimation can be performed directly in the time-scale domain by isolating regions wherein the prediction error can be attributed to the error of individual dynamic model parameters [1]. Based on these single-parameter attributions of the prediction error, individual parameter errors can be estimated for iterative parameter estimation. A benefit of relying entirely on the time-scale domain for parameter estimation is the added capacity for noise suppression. This paper explores this benefit by introducing a noise compensation method that estimates the distortion by noise of the prediction error in the time-scale domain and incorporates it as a confidence factor when estimating individual parameter errors. This method is shown to further improve the estimated parameters beyond the time-filtering and denoising techniques developed for time-based estimation.

Commentary by Dr. Valentin Fuster
2009;():9-16. doi:10.1115/DSCC2009-2552.

A method of parameter estimation was recently introduced that separately estimates each parameter of the dynamic model [1]. In this method, regions coined as parameter signatures, are identified in the time-scale domain wherein the prediction error can be attributed to the error of a single model parameter. Based on these single-parameter associations, individual model parameters can then be estimated for iterative estimation. Relative to nonlinear least squares, the proposed Parameter Signature Isolation Method (PARSIM) has two distinct attributes. One attribute of PARSIM is to leave the estimation of a parameter dormant when a parameter signature cannot be extracted for it. Another attribute is independence from the contour of the prediction error. The first attribute could cause erroneous parameter estimates, when the parameters are not adapted continually. The second attribute, on the other hand, can provide a safeguard against local minima entrapments. These attributes motivate integrating PARSIM with a method, like nonlinear least-squares, that is less prone to dormancy of parameter estimates. The paper demonstrates the merit of the proposed integrated approach in application to a difficult estimation problem.

Commentary by Dr. Valentin Fuster
2009;():17-24. doi:10.1115/DSCC2009-2565.

Model validation is the procedure whereby the fidelity of a model is evaluated. The traditional approaches to dynamic model validation either rely on the magnitude of the prediction error between the process observations and model outputs or consider the observations and model outputs as time series and use their similarity to assess the closeness of the model to the process. Here, we propose transforming these time series into the time-scale domain, to enhance their delineation, and using image distances between these transformed time series to assess the closeness of the model to the process. It is shown that the image distances provide a more consistent measure of model closeness than available from the magnitude of the prediction error.

Topics: Dynamic models
Commentary by Dr. Valentin Fuster
2009;():25-32. doi:10.1115/DSCC2009-2574.

Three procedures for designing robust observers to estimate the state variables of nonlinear constrained systems have been developed in this work. All observers are based on the sliding mode methodology and assume that the number of transducers matches that of the degrees of freedom of the system. The conceptual differences between the proposed observer designs are in the number and selection of the sliding surfaces along with the formulations pertaining to their nominal models. The observers have been applied to estimate the state variables of a crank-slider mechanism of a single cylinder engine. The simulation results demonstrate the capabilities of the observers in accurately estimating the state variables of the system, including the superfluous ones, in the presence of significant structured and unstructured uncertainties. In addition, the results show that the nominal constraint equations are satisfied by the estimated state variables.

Commentary by Dr. Valentin Fuster
2009;():33-40. doi:10.1115/DSCC2009-2715.

This paper proposes a multiple model approach for a nonlinear multi-input, multi-output (MIMO) system identification (SI) of smart structures equipped with magnetorheological (MR) dampers. The proposed model is developed through integration of MIMO autoregressive exogenous (ARX) input models, Takagi-Sugeno (TS) fuzzy model, weighted linear least squares estimators, and data clustering algorithms. Nonlinear behavior of the structure-MR damper systems is represented by a set of linear MIMO ARX input models whose operating regions are blended by TS fuzzy sets. To demonstrate the effectiveness of the proposed MIMO ARX-TS fuzzy model, a 20-story high-rise building employing MR dampers is investigated. It is demonstrated that the proposed approach is effective in modeling nonlinear behavior of the structure-MR damper system subjected to a variety of disturbances. Comparison with high fidelity data proves the viability of the proposed approach in control engineering setting.

Commentary by Dr. Valentin Fuster
2009;():41-48. doi:10.1115/DSCC2009-2759.

This paper proposes a novel principal component analysis (PCA)-based sensor fault detection methodology for smart structures employing magnetorheological (MR) dampers. The MR damper is operated by a semiactive nonlinear fuzzy controller (SNFC) that is developed by integration of a set of Lyapunov optimal controllers, Kalman filters, and a semiactive converter with the fuzzy interpolation method. A numeric residual generator is found using the PCA analysis of ten measurements obtained from the structure-MR damper system for sensor fault detection. Using the matrix of this residual generator, the detectability and isolability of each sensor has been analyzed and the detection and isolation algorithm is applied to the smart structural system with different levels of artificially added faults. The simulation demonstrated that the proposed PCA-based sensor fault detection approach is effective in identifying the sensor faults of large smart structures employing MR dampers.

Commentary by Dr. Valentin Fuster


2009;():49-51. doi:10.1115/DSCC2009-2571.

In this paper, we report measurements of the multi-variable torque-displacement relation at the ankle. The passive behavior of the ankle in two degrees of freedom (inversion-eversion and dorsiflexion-plantarflexion) was quantified using the Anklebot. The measured torque-displacement relationship was represented as a vector field using thin-plate spline smoothing with generalized cross validation. Analysis of the experimental results showed that, when maximally relaxed, the ankle behaved like a mechanical spring. However, if muscles were active, the torque-displacement relation was not spring-like. Implications for the contribution of neural feedback to ankle impedance are discussed.

Topics: Torque , Displacement
Commentary by Dr. Valentin Fuster
2009;():53-55. doi:10.1115/DSCC2009-2573.

Though ankle mechanical impedance plays an important role in posture and locomotion, it has been inadequately characterized. Unlike previous studies, which confined themselves to measurements along the primary axes of the ankle in an isolated fashion, the study reported here characterized the static component of ankle impedance in two degrees of freedom. In addition, the effect of active muscle contraction on ankle static impedance was measured. We found that ankle static impedance varied significantly with direction under passive conditions. We further observed that, while muscle contraction increased the magnitude of ankle static impedance, its directional variation was essentially unchanged.

Commentary by Dr. Valentin Fuster
2009;():57-64. doi:10.1115/DSCC2009-2615.

Neurotransmitter homeostasis in and around synapses involves random processes such as diffusion, molecular binding and unbinding. A three-dimensional stochastic diffusion model of a synapse was developed to provide molecular level details of neurotransmitter homeostasis not predicted by alternative models based on continuum approaches. This framework was used to estimate effective diffusion and provide a more accurate prediction of geometric tortuosity in the perisynaptic region. The stochastic model was used to predict the relative contributions of non-synaptic sources to extracellular concentration in control, natural reward seeking, and chronic cocaine cases; and estimation of molecular influx rates required to maintain tone on presynaptic autoreceptors. Also, this was the first stochastic model to confirm the prediction of down-regulation of glutamate transporters by 40% after chronic cocaine. The model can be further generalized to study the role of diffusion path length in supporting neurotransmitter gradients and isolating the synapse.

Commentary by Dr. Valentin Fuster
2009;():65-72. doi:10.1115/DSCC2009-2621.

In this paper, we briefly review some fundamental qualitative features of ecological sign stability and transform these principles of ecology to a set of mathematical results in matrix theory with quantitative information, which is usually encountered in engineering sciences. This type of cross fertilization of ideas of life sciences and engineering sciences is deemed to be highly beneficial to both fields. In particular, we show in this paper what effect the signs of elements of a matrix have on the matrix properties such as eigenvalues and condition number. Similarly, it is also shown that under some assumptions on the magnitudes of the elements, predator-prey phenomenon in ecology renders some special properties like ‘normality’ to matrices. It is also shown that these predator-prey models have better robustness properties when compared to other matrices. The results presented in this paper can assist in the use of ecological system principles to build highly robust engineering systems.

Commentary by Dr. Valentin Fuster
2009;():73-80. doi:10.1115/DSCC2009-2640.

This work reports on the design and the feed forward stiffness control of bioinspired kinematic chains from a static and a dynamic point of view. While position control is clearly referred to common geometrical lagrangian coordinates for the considered system, in order to deal with the stiffness or compliance of the chain, especially in dynamic cases, global and less intuitive variables have to be defined and used. The advantage deriving from a similar control strategy can be important when the chain is part of a complex dynamic system or the computational resources are scarce. By defining and controlling stiffness or compliance for a certain position or trajectory, we can state that, even if the system is not continuously monitored in closed loop, a bounded perturbation cannot produce equilibrium point or trajectory variations greater than a fixed quantity. In a closed loop control strategy, the described methodology can be implemented during the time between two consecutive output sampling and feedback inputs. On the other hand, compliance control permits a kinematic chain to interact with objects without causing damages even if errors in position occur. In this work, the compliance and stiffness concepts, inspired to common reasoning in biological motor control theory, are generalized to a dynamic case and endowed with a mathematical architecture.

Commentary by Dr. Valentin Fuster
2009;():81-88. doi:10.1115/DSCC2009-2705.

Living cells stochastically switch their phenotypic states in response to environmental cues to maintain persistence and viability. Estimating the state transition probabilities from biological observations of cell populations gives valuable insight to the underlying processes, and gives insights as to how the transition statistics are influenced by external factors. In this work, we present two Bayesian estimation approaches. The first is applicable when individual cell state trajectories are observed. The second approach is applicable when only aggregate population statistics are available. Estimation of transition probabilities when individual cell state trajectories are available is a straightforward problem, whereas estimation from only aggregate statistics can be computationally expensive. In the latter case, we present an algorithm that relies on three key ideas to cut down computational time: i) approximating high-dimensional multinomial distributions with multi-variate Gaussians, ii) employing Monte-Carlo techniques to efficiently integrate over high dimensional spaces, and iii) explicitly incorporating sampling constraints by computing lower dimensional distributions over the constrained variable. Simulation results demonstrate the viability of the algorithm.

Commentary by Dr. Valentin Fuster

Intelligent Systems and Estimation

2009;():89-96. doi:10.1115/DSCC2009-2526.

A novel Condensed Hybrid Optimization (CHO) algorithm using Enhanced Continuous Tabu Search (ECTS) and Particle Swarm Optimization (PSO) is proposed. The proposed CHO algorithm combines the respective strengths of ECTS and PSO. The ECTS is a modified Tabu Search (TS), which has good search capabilities on large search spaces. In this study, ECTS is utilized to define smaller search spaces, which are used in a second stage by the basic PSO to find the respective local optimum. The ECTS covers the global search space by using a TS concept called diversification and then selects the most promising areas in the search space. Once the promising regions in the search space are defined, the proposed CHO algorithm employs another TS concept called intensification in order to search the promising area thoroughly. The proposed CHO algorithm is tested with the multi-dimensional Hyperbolic and Rosenbrock problems. Compared to other four algorithms, the simulations results indicate that the accuracy and effectiveness of the proposed CHO algorithm.

Commentary by Dr. Valentin Fuster
2009;():97-104. doi:10.1115/DSCC2009-2563.

The problem considered in this paper is the analysis of a battery State-of-Charge estimation algorithm: in particular the mix estimation algorithm. This algorithm provides the estimation mixing two estimation approaches: namely the Coulomb-Counting and the Model-Based. The mix algorithm is qualitatively able to provide a more robust and accurate estimation respect to the estimation provided by the approaches the algorithm mixes together. The aim of this paper is to analyze the differences between the three algorithms and the advantages produced by the mixing procedure. In particular the paper presents the comparison of the mix algorithm behavior with the Coulomb-Counting and the Model-Based behaviors in case of model errors.

Commentary by Dr. Valentin Fuster
2009;():105-110. doi:10.1115/DSCC2009-2694.

Allocation of a large number of resources to tasks in a complex environment is often a very challenging problem. This is primarily due to the fact that a large number of resources to be allocated results into an optimization problem that involves a large number of decision variables. Most of the optimization algorithms suffer from this issue of non-scalability. Further, the uncertainties and dynamic nature of environment make the optimization problem quite challenging. One of the techniques to overcome the issue of scalability that have been considered recently is to carry out the optimization in a distributed or decentralized manner. Such techniques make use of local information to carry out global optimization. However, such techniques tend to get stuck in local minima. Further, the connectivity graph that governs the sharing of information plays a role in the performance of algorithms in terms of time taken to obtain the solution, and quality of the solution with respect to the global solution. In this paper, we propose a distributed greedy algorithm inspired by market based concepts to optimize a cost function. This paper studies the effectiveness of the proposed distributed algorithm in obtaining global solutions and the phase transition phenomenon with regard to the connectivity metrics of the graph that underlies the network of information exchange. A case study involving resource allocation in wildland firefighting is provided to demonstrate our algorithm.

Topics: Optimization
Commentary by Dr. Valentin Fuster
2009;():111-118. doi:10.1115/DSCC2009-2707.

Every year all over the world, wildfires do extensive damages to the human lives, properties and natural resources. National Interagency Fire Center data provides a detailed description of the severe damages caused by the wildfires every year. Forest Fire Decision Support Systems (FFDSS) have been developed all over the world during the last thirty years with the purpose of fire detection, fire behavior prediction, and risk assessment. But optimized wildland fire containment strategies are largely lacking in these FFDSS. In this paper, decision making strategies have been formulated for wildland fire suppression so that the total burned area and hence the damage is minimized. This goal is achieved by the application of optimization tools such as the Genetic Algorithms (GA). For a given number of resources, the GA will determine their best utilization strategy so that the total area burnt is minimized. For generating optimal strategies for resource utilization, the Genetic Algorithm uses an advanced fire propagation model that predicts the propagation of wildland fires under given environmental conditions and topography. The fire-fighting strategy considered in this paper is fireline generation. Using the Genetic Algorithm, the optimal fireline is built that minimizes the area of land burned. GA also provides the proper locations of the attacking crews so that the fireline is built before the fire escapes. Using these intelligent decision making strategies, the damage caused due to a forest fire can be minimized significantly.

Commentary by Dr. Valentin Fuster

Energy Storage and Harvesting

2009;():119-121. doi:10.1115/DSCC2009-2542.

Energy harvesters are a promising technology for capturing useful energy from the environment or a machine’s operation. In this paper we highlight ideas that might lead to energy harvesters that more efficiently harvest a portion of the considerable vibrational energy that is present for human-made devices and environments such as automobiles, trains, aircraft, watercraft, machinery, and buildings. Specifically, we consider how to exploit ideas based on properties of nonlinear oscillators with negative linear stiffness driven by periodic and stochastic inputs to design energy harvesters having large amplitude response over a broad range of ambient vibration frequencies. Such harvesters could improve upon proposed harvesters of vibrational energy based on linear mechanical principles, which only give appreciable response if the dominant ambient vibration frequency is close to the resonance frequency of the harvester.

Commentary by Dr. Valentin Fuster
2009;():123-130. doi:10.1115/DSCC2009-2558.

Energy requirements for heating and cooling of residential, commercial and industrial spaces constitute a major fraction of end use energy consumed. Centralized systems such as hydronic networks are becoming increasingly popular to meet those requirements. Energy efficient operation of such systems requires intelligent energy management strategies, which necessitates an understanding of the complex dynamical interactions among its components from a mathematical and physical perspective. In this work, concepts from linear graph theory are applied to model complex hydronic networks. Further, time-scale decomposition techniques have been employed to obtain a more succinct representation of the overall system dynamics. Lastly, the usefulness of the proposed model for energy efficient operation of the system through advanced control techniques has been discussed.

Commentary by Dr. Valentin Fuster
2009;():131-138. doi:10.1115/DSCC2009-2679.

New micro renewable energy harvesting devices are being developed using the stable limit cycle response of aeroelastic systems to drive energy conversion. This paper analyzes such devices. This paper investigates devices that use two types of aeroelastic instability: galloping and multi-mode flutter. Since the generation of power can be stabilizing, resulting in no power generation at all, the analysis begins by analyzing the stability of such devices from the perspective of power generation. Next, the level of power generation is discussed, and peak levels of performance are found. The analysis suggests that with proper tuning the power generation of micro aeroelastic energy harvesters operating at representative speeds (∼4.5 m/s (10 mph) ) can produce power on the order of 10 mW .

Commentary by Dr. Valentin Fuster
2009;():139-147. doi:10.1115/DSCC2009-2724.

Recent advances in lithium ion battery modeling suggest degradation may be reduced by permitting unequal but controlled charging of individual cells. Hence, this paper compares anode-side film formation for a standard equalization scheme versus unequal charging through switches controlled by deterministic dynamic programming (DDP) and DDP-inspired algorithms. A static map for film growth rate is derived from a first-principles battery model adopted from the electrochemical engineering literature. Using this map, we consider two cells connected in parallel via relay switches. The key results we demonstrate are: (1) Enabling unequal charging indeed reduces film buildup; (2) A near optimal control method can be implemented using heuristic rules designed from the DDP solution and convexity properties of film growth rate. Simulation results indicate that the heuristic rules achieve near optimal performance relative to the DDP solution, with significant reduction in film growth compared to charging both cells equally.

Commentary by Dr. Valentin Fuster
2009;():149-156. doi:10.1115/DSCC2009-2729.

This paper illustrates the application of hybrid modeling and model predictive control techniques to the water purge management in a fuel cell with dead-end anode. The anode water flow dynamics are approximated as a two-mode discrete-time switched affine system that describes the propagation of water inside the gas diffusion layer, the spilling into the channel and consequent filling and plugging the channel. Using this dynamical approximation, a hybrid model predictive controller based on on-line mixed-integer quadratic optimization is tuned, and the effectiveness of the approach is shown through simulations with a high-fidelity model. Then, using an off-line multiparametric optimization procedure, the controller is converted into an equivalent piecewise affine form which is easily implementable even in an embedded controller through a lookup table of affine gains.

Commentary by Dr. Valentin Fuster
2009;():157-164. doi:10.1115/DSCC2009-2774.

This article provides an analysis of the effect on the overall energy bill of a commercial facility of active energy management. We first show the benefits of pure hedging, hedging when the facility has its own power source—we consider the use of co-generation in winter and the use of solar power in summer. We next show how active control of facility temperature set points augments the benefits of the hedging and use of local power. Our studies are based on real consumption data of a large commercial facility, the corresponding real time prices of grid power, prices of natural gas, intensity of solar radiation, and temperature history of the period under consideration. We show that the combination of hedging, local power generation and active control can reduce facility energy bills by up to 30%, and bill variance by up to 80%. Thus, we have a scenario where consumers save significantly while using power sources with a smaller carbon footprint.

Topics: Carbon , Cutting
Commentary by Dr. Valentin Fuster

Modeling and Control of Vehicle Dynamics

2009;():165-171. doi:10.1115/DSCC2009-2527.

Aerial tramways are a fast and efficient mean of transportation, especially found in mountainous regions where the ability of clearing steep inclines makes them a preferred solution. Often non-autonomous, these types of cabin have little or no power installed and thus must be drawn by the haulage rope. In case of failure, a safety braking system must prevent the cabin from rolling back along the track rope. The braking system under investigation is a fail-safe device composed by two braking units that clamp the track rope in case of failure. The aim of the present study was to estimate the braking time. Given the complexity and costs related to on-field experimental tests, simulations and laboratory tests were carried out. The braking unit was modelled in MSC ADAMS, making use of the MSC ADAMS/Hydraulics plug-in. Simulations and experiments were in a very good agreement, evidencing the system’s compliance to Italian regulation.

Topics: Safety , Simulation , Braking
Commentary by Dr. Valentin Fuster
2009;():173-180. doi:10.1115/DSCC2009-2572.

This paper describes the algorithms used for controlling an autonomous vehicle that operates at the limits of tire adhesion. The controller is designed to imitate a racecar driver by using both feedforward and feedback to command the steering, throttle, and brakes of the vehicle. The feedforward steering is based on the vehicle handling diagram, while the lanekeeping steering feedback is added to ensure vehicle stability and reduces tracking errors caused by disturbances or modeling errors. The feedforward speed is estimated based on the available friction, while the proportional speed feedback is introduced to mimic a race-car driver modulating the speed to trim the vehicle orientation. Two different speed feedback designs based on lookahead error and heading error are compared. The experiments demonstrate the superiority of heading error feedback, which enables the vehicle to operate at its limits while maintaining minimal lateral and heading errors from the desired trajectory.

Commentary by Dr. Valentin Fuster
2009;():181-188. doi:10.1115/DSCC2009-2626.

This paper presents an analysis of the dynamics of a rear wheel drive vehicle during cornering at high sideslip angles (“drifting”) using a three-state bicycle model. This model builds upon previous work with a two-state bicycle model by incorporating longitudinal dynamics and a nonlinear tire model with simplified lateral-longitudinal force coupling. Analysis of this model reveals the existence of unstable “drift equilibria” corresponding to steady state cornering at high sideslip angles with significant longitudinal force applied at the rear tire. These equilibria are saddle points, with characteristics that exhibit low sensitivity to friction and speed variation. The analysis of the equilibria provides insight into vehicle dynamics in an operating regime responsible for major safety concerns in everyday driving. It also sheds light upon aspects of the system dynamics that account for behavior observed in autonomous drift experiments and must be considered in future controller designs.

Commentary by Dr. Valentin Fuster
2009;():189-196. doi:10.1115/DSCC2009-2636.

The National Crash Analysis Center (NCAC) at the George Washington University (GWU) has been developing and maintaining a public domain library of LS-DYNA finite element (FE) vehicle models for use in transportation safety research. The recent addition to the FE model library is the 2007 Chevrolet Silverado FE model. This FE model will be extensively used in roadside hardware safety research. The representation of the suspension components and its response in oblique impacts into roadside hardware are critical factors influencing the predictive capability of the FE model. To improve the FE model fidelity and applicability to the roadside hardware impact scenarios it is important to validate and verify the model to multitude of component and full scale tests. This paper provides detailed description of the various component and full scale tests that were performed, specifically, to validate the suspension model of the 2007 Chevrolet Silverado FE model.

Commentary by Dr. Valentin Fuster
2009;():197-204. doi:10.1115/DSCC2009-2659.

Vehicle stability control systems have been widely and accurately cited as a significant influence in reducing the rate of severe injuries and fatalities in automotive crashes. However, these systems are purely reactive, providing additional control input only after undesired vehicle behavior is sensed. This paper presents a new approach to controlling the motion of a vehicle in highly dynamic situations. This approach solves a convex optimization problem over a finite time horizon to predict and prevent these hazardous situations. Thus, the controller determines input that simultaneously tracks the driver’s intended trajectory while preventing tire saturation. Simulation results are presented to demonstrate the efficacy of this control approach.

Commentary by Dr. Valentin Fuster
2009;():205-212. doi:10.1115/DSCC2009-2670.

This paper presents a kinematic yaw rate estimator which combines measurements obtained from a pair of accelerometers and ABS wheel speed sensors of the non-driven wheels. The Extended Kalman Filter methodology has been used for realization of this sensor fusion-based estimator. The estimator adapts to variations in reliability levels of individual sensors. The estimator performance is validated and corresponding estimation errors are analyzed by computer simulations for different driving maneuvers.

Commentary by Dr. Valentin Fuster

Hybrid Vehicles

2009;():213-221. doi:10.1115/DSCC2009-2538.

Energy management controllers for hybrid electric vehicles typically contain numerous parameters that must be tuned in order to arrive at a desired compromise among competing attributes, such as fuel economy and driving quality. This paper estimates the Pareto tradeoff curve of fuel economy versus driving quality for a baseline industrial controller, and compares it to the Pareto tradeoff curve of an energy management controller based on Shortest Path Stochastic Dynamic Programming (SPSDP). Previous work demonstrated important performance advantages of the SPSDP controller in comparison to the baseline industrial controller. Because the baseline industrial controller relies on manual tuning, there was always the possibility that better calibration of the algorithm could significantly improve its performance. To investigate this, a numerical search of possible controller calibrations is conducted to determine the best possible performance of the baseline industrial controller and estimate its Pareto tradeoff curve. The SPSDP and baseline controllers are causal; they do not rely on future drive cycle information. The SPSDP controllers achieve better performance (i.e., better fuel economy with equal or better driving quality) over a wide range of driving cycles due to fundamental structural limitations of the baseline controller that cannot be overcome by tuning. The message here is that any decisions that specify or restrict controller structure may limit attainable performance, even when many tunable parameters are made available to calibration engineers. The structure of the baseline algorithm and possible sources of its limitations are discussed.

Commentary by Dr. Valentin Fuster
2009;():223-227. doi:10.1115/DSCC2009-2549.

Direct Current (DC) Motors are one of the most common mechatronic actuators. They are important for electromechanical servo systems, drivers for battery powered appliances and tools as well as electric vehicles. Both brushless DC motors and wound DC motors are common in electric and hybrid vehicles. The series wound DC motor is commonly used for high torque vehicle applications. The literature has many papers discussing permanent magnet DC motors but a very limited number of publications on analytical models for series wound DC motors, especially motor models that fit series wound DC motor test data available in the market place. An analytical model for a series wound DC motor is developed here based on physical principles including energy conservation. The model developed will be compared with models developed by other investigators. Available commercial test data for a series motor will be used to find model parameters for the analytical model and the accuracy of this model evaluated against the original test data. The model developed displays excellent accuracy well within the accuracy of the test data available. Typical model rms deviation from test data is under 2% for the commercial series wound DC motors evaluated.

Commentary by Dr. Valentin Fuster
2009;():229-236. doi:10.1115/DSCC2009-2716.

A Fast Dynamic Programming (FDP) algorithm is proposed to optimize the fuel consumption of the Plug-in Hybrid Electronic Vehicles (PHEV) over a prescribed driving cycle. Firstly, an innovative DP mathematical model for PHEV with reduced dimension is created. It composes of a linear state transfer equation and a quadratic cost function with 1×1 dimension. Based on this model, two algorithms expressed as the simple analytic forms are derived. One is the control algorithm for the optimal power splitting ratio (PSR) between the internal combustion engine (ICE) power and demand power. Another is the recursive algorithm to calculate the optimization value of the fuel consumption. Then, the optimal output power of ICE (or electronic motor (EM)) and the optimal speed ratio of the gear position can be calculated rapidly as the consequence of using the algorithms with analytic forms. Finally, the simulation results confirm that the computational efficiency of the FDP control algorithm has been improved with a geometric ratio, while its control performance maintains in an acceptable range in contrast with conventional DP control algorithm.

Commentary by Dr. Valentin Fuster
2009;():237-244. doi:10.1115/DSCC2009-2741.

This paper uses lumped parameter dynamic equations to model the mass flow, piston dynamics, and control volume behavior inside a free-piston Stirling engine. A new model for a Stirling engine thermal regenerator that incorporates a dynamically changing temperature gradient is presented. The use of graphite as a regenerator matrix material is justified despite its limited background by comparing the functional requirements of regenerators to heat exchangers where graphite use is commonplace. Experimental results are used to characterize a graphite regenerator and validate the dynamic model.

Commentary by Dr. Valentin Fuster
2009;():245-251. doi:10.1115/DSCC2009-2745.

Battery has drawn much more attention due to the accelerating development of Electric Vehicle (EV) and Plugin Hybrid Electric Vehicles (PHEV) for sustainable mobility. It is worthwhile to study the dynamic model of battery in order to achieve better performance in power management. This paper proposes a nonlinear dynamic battery model for PHEV power management. First, an existing RC equivalent circuit is used to simulate the dynamics of the battery states with respect to the input current. Then, a concept of nonlinear hysteresis degree is introduced, and is adopted to implement the mathematical description of the battery nonlinear boundary and scanning hysteresis. Finally, the dynamics and the nonlinear boundary (scanning) hysteresis are integrated into a comprehensive battery model with an affine nonlinear state-space equation. The effectiveness of the proposed model is verified using some simulation tests.

Topics: Batteries
Commentary by Dr. Valentin Fuster
2009;():253-258. doi:10.1115/DSCC2009-2750.

This paper investigates the role of partial or complete knowledge of future driving conditions in fuel economy of Plug-in Hybrid Vehicles (PHEVs). We show that with the knowledge of distance to the next charging station only, substantial reduction in fuel use, up to 18%, is possible by planning a blended utilization of electric motor and the engine throughout the entire trip. To achieve this we formulate a modified Equivalent Consumption Minimization Strategy (ECMS) which takes into account the traveling distance. We show further fuel economy gain, in the order of 1–5%, is possible if the future terrain and velocity are known; we quantify this additional increase in fuel economy for a number of velocity cycles and a hilly terrain profile via deterministic dynamic programming.

Commentary by Dr. Valentin Fuster

ASME/Bath Fluid Power Symposium: Pump Design, Analysis and Application

2009;():259-266. doi:10.1115/DSCC2009-2543.

External market forces constrain the design of engineered systems in much the same way that environmental conditions constrain the evolution of biological systems. Dimensional analysis is used to find functional relationships within biological systems and in turn can be used determine design relationships for engineered ones. This work applies these concepts to axial-piston swashplate variable displacement pumps. After presenting a pump model suitable for dimensional analysis, the necessary transformations are performed to create a pump model with completely dimensionless parameters and dynamics. Using a set of industry data, scale independent design information is extracted from successful industry pumps. This information is used to develop dimensionless design guidelines and strategies which allow for the specification of new designs dynamically similar to the original successful products.

Topics: Design , Pumps , Displacement
Commentary by Dr. Valentin Fuster
2009;():267-274. doi:10.1115/DSCC2009-2604.

A stationary model is adopted to determine the critical condition for which the slipper moves away from the swashplate in an axial piston machine. The aim of the analysis is to find the critical speed, i.e. the value of the machine speed for which the slipper moves away from the swashplate; usually this condition may determine bad operating behaviour of the machine, although a retainer plate is used to maintain the slipper sufficiently near to the swashplate. The influences of the pressure transition in the cylinder, the swashplate angle and the radial clearance between piston and cylinder on the critical speed are depicted. Successively, the role of the position of the point of application of the resultant force due to the slipper-retaining plate contact is analyzed.

Topics: Motors , Bearings , Pumps , Pistons
Commentary by Dr. Valentin Fuster
2009;():275-282. doi:10.1115/DSCC2009-2693.

In this study, a valveless energy saving hydraulic position control servo system controlled by two pumps is investigated. In this system, two variable speed pumps driven by servomotors regulate the flow rate through a differential cylinder according to the needs of the system, thus eliminating the valve losses. The mathematical model of the system is developed in MATLAB Simulink environment. A Kalman filter is applied to reduce the noise in the position feedback signal. In the test set up developed, open loop and closed loop frequency response and step response tests are conducted by using MATLAB Real Time Windows Target (RTWT) module, and test results are compared with the model outputs.

Commentary by Dr. Valentin Fuster
2009;():283-290. doi:10.1115/DSCC2009-2719.

Hydraulic accumulators (HAs) have been used successfully in regenerative braking systems by companies such as Ford and Eaton Corp. to increase fuel efficiency of heavy vehicles by as much as 25–35%. However, the relatively low gravimetric and volumetric energy densities of conventional HAs prohibit their use in average-sized passenger vehicles. In an attempt to address this problem, an elastomer will be used to construct a HA that will use strain as the primary energy storing mechanism. By using strain in the composition material, as opposed to compression of a precharged gas, this accumulator should virtually eliminate heat losses due to extended holding times. Because its gravimetric and volumetric elastic energy storage density values are among the highest of any material, polyurethane was the elastomer chosen as the constituent material. Using a curable type of polyurethane, an α-prototype is currently being manufactured to provide empirical data for validation.

Commentary by Dr. Valentin Fuster
2009;():291-298. doi:10.1115/DSCC2009-2779.

This study is a part of a larger research project to predict noise sources of hydrostatic transmissions and investigating new methods for designing quieter systems. The aim of this study is to validate the developed model describing pump dynamics coupled with effects of a connecting line, thus validating a coupled pump-motor-line model for hydrostatic transmissions. This paper illustrates a numerical approach for evaluating pressure and flow oscillations generated by a hydraulic pump coupled with a connecting line. The presented model describes pump dynamics using a lumped parameter approach as well as one-dimensional unsteady compressible fluid flow by means of method of characteristics (MOC). Several lumped parameter models have been developed for hydraulic pumps and motors and the method of characteristics has been applied for many applications; however, the presented model uniquely utilizes both approaches and considers influence of pump dynamics and propagating pressure and flow pulsations throughout the line. Measurements of pressure ripple in the line at two different points were carried out at various loading conditions to validate the developed model. Comparisons between measurements, the developed model, and another more simplified model were conducted. Results indicate a reasonable match between the developed model and measurements as well as the importance of considering a line model based on method of characteristics.

Commentary by Dr. Valentin Fuster
2009;():299-306. doi:10.1115/DSCC2009-2780.

Power-split drive represents a class of Continuously Variable Transmission (CVT) that combines the convenience of CVT with the high overall transmission efficiency. In its hybrid configuration, a high pressure accumulator is used to capture the braking energy that is regenerated to aid the engine power during the next propulsion event. Output coupled power split drives are particularly suited for small and medium duty vehicle applications. In this work, optimal power management strategy has been designed based on Dynamic Programming approach. Although the control strategy obtained by Dynamic Programming is non-causal, it represents the benchmark solution against which other implementable power management schemes can be compared. Another control strategy based on instantaneous optimization is also discussed where a given cost function is minimized at every instant. It results in a sub-optimal solution that is practical and implementable. Finally, Dynamic Programming results are utilized to discuss the possible improvements that can be made to the instantaneous optimization based control strategy.

Commentary by Dr. Valentin Fuster

Modeling and Simulation

2009;():307-314. doi:10.1115/DSCC2009-2504.

A method for determining a closed-form expression for the hydrodynamic forces in finite plain journal bearings is introduced. The method is based on applying correction functions to the force models of the infinitely-long (IL) or infinitely-short (IS) bearing approximation. The correction functions are derived by modeling the ratio between the forces from the numerical integration of the two-dimensional Reynolds equation and the forces from either the IL or IS bearing approximation. Low-order polynomial models, dependent on the eccentricity ratio and aspect ratio, are used for the correction functions. The models are shown to outperform the standard limiting approximations in the steady-state analysis of the bearing system under static loading.

Commentary by Dr. Valentin Fuster
2009;():315-322. doi:10.1115/DSCC2009-2557.

A port based model is an external representation including pairs of input-output variables grouped into ports. Each port in a model is a representation of a mechanism for energy transfer into, or out of, the physical system modeled. Port based models always have an equal number of inputs and outputs because each port is composed of an input-output pair. The number of ports in such models must typically be reduced to decrease model size, protect proprietary internal topology or simply because the eventual user has no plans to make connections to them. The process of eliminating undesired model ports while maintaining useful ports is called condensation. The condensation of port-based linear system models was developed in [1]. This article extends the condensation process for port-based nonlinear models represented through Volterra series.

Topics: Condensation
Commentary by Dr. Valentin Fuster
2009;():323-330. doi:10.1115/DSCC2009-2580.

This paper introduces a new type of active fluid-film bearing and its feedback control. In particular, we propose to actively adjust the angular velocity of the pads of a tilting-pad bearing in response to changes in the operating conditions of the rotating machine. This is motivated by the observation that there is more control authority in the pad tilt motion than in its radial translation. To this end, we first develop a dynamic model for the bearing system, inclusive of the nonlinear hydrodynamic force for the infinitely-short bearing case. A model-based controller is then constructed, based on measurements of the journal position and velocity and pad tilt angles, to ensure that the journal is asymptotically regulated to the bearing center. Numerical simulations illustrate the performance of the active bearing under the proposed control in comparison with the bearing’s standard passive mode of operation.

Topics: Bearings
Commentary by Dr. Valentin Fuster
2009;():331-338. doi:10.1115/DSCC2009-2630.

Gyroscopes are commonly used to measure the angle of rotation and its rate of change in several critical systems like airplanes. Therefore, there is a never-ending desire for researchers to increase measurement precision of these devices. In order to achieve this goal, some new gyroscopes have been invented recently. Especially, advent of micro manufacturing has appeared some sophisticated to more précised gyroscopic systems. The widely-used gyroscopes are vibrating beam gyroscopes; however they face a very important drawback, called cross-coupling error. In presence of the secondary base rotations, significant errors will be produced in measurement of the gyroscope output. In order to deal with this issue, this paper addresses a novel gyroscopic system, called rocking-mass gyroscope. It is consist of four beams attached to a rigid substrate, undergoing coupled flexural and torsional vibrations with a finite mass attached in the middle. This configuration is such that, it does not encounter the same problems as vibrating beam gyroscopes. This configuration makes the vibration analysis very complicated. Despite this fact, a thorough analysis is performed in this paper. Using Extended Hamilton’s principle, eight governing partial differential equations of motion along with their corresponding boundary conditions are derived. Further attempt is made to find the closed-form frequency equation of the system. Solving this equation needs high computational costs and gives the natural frequencies of the system. In spite of this fact, the system is analysed in the frequency domain using an exact method in full detail, for two cases of fixed and rotating base support. Furthermore, a detailed parameter sensitivity analysis is carried out to determine the effects of different parameters on the natural frequencies of the system. The contributions of this research are very important from two viewpoints. Firstly, determination of natural frequencies and resonance conditions are essential for design of the system, and design of appropriate control strategies. Secondly, frequency domain analysis forms the basis of time domain analysis, followed by exact mode superposition method.

Commentary by Dr. Valentin Fuster
2009;():339-345. doi:10.1115/DSCC2009-2696.

A wind turbine generator offers a green renewable alternative to the traditional fossil and nuclear fuel processes to generate electrical power. Both wind energy technology and wind turbine farm designs remain in demand given the current growth in energy requirements and the public’s preference for clean sources. Simply put, wind energy offers a safe, relatively cost effective solution for global energy production. For example, the energy demand for populated coastal cities encourages offshore farms to fulfill future electrical needs. Similarly, wind turbines may be placed in land-locked regions and power transmitted through electric grids to population centers. Most wind models available to engineers offer superb capabilities for predicting wind velocities on land and far offshore (5 km and greater). However, near shore winds have proven difficult to determine due to surface roughness, thermal stratification, and abrupt displacement height variances. This paper discusses the model comparison of two foremost wind speed prediction tools, AWS Truewinds’ MASS and WindPro’s WAsP. The model comparison is related to measured South Carolina coastal data and suggests AWS Truewind’s MASS wind shear model is the more effective near-shore wind speed prediction tool. In arriving at these results, several areas of future work are discussed.

Commentary by Dr. Valentin Fuster
2009;():347-354. doi:10.1115/DSCC2009-2786.

Passive fluid mounts have been in use for the purpose of cabin noise and vibration reduction in the automotive and the aerospace industry. Cabin noise and vibration isolation is provided at a frequency coined “notch frequency”. To obtain the greatest cabin noise and vibration reduction at any desired frequency, the notch frequency needs to be close to that desired frequency. But, due to tolerances on all the fluid mount dimensions, and elastomer material properties, the notch frequency never ends up at the right location on the first manufacturing pass. To resolve notch frequency tuning cycle time, a new fluid mount design is proposed which consists of a conventional single-pumper fluid mount and a 3-layer piezoelectric cantilever beam resulting in a tunable notch frequency mount. Since this new design involves multi energy domains, bond graph modeling technique is used. This new design concept, its mathematical model and simulation results are presented.

Commentary by Dr. Valentin Fuster

Biologically Inspired Design, Control and Planning for Robotics

2009;():355-361. doi:10.1115/DSCC2009-2530.

In this paper, we develop an analytical basis for designing the locomotion of mobile robots with a spherical or circular core and equispaced diametral legs. The mechanism has resemblance with certain cellular locomotions. Locomotion is generated by actuation of the legs in the radial direction. Two elementary regimes of motion are first developed using the geometry of the mechanism. The overall motion of the robot is generated by repeated switching between the two regimes. The paper addresses both the kinematics and dynamics of the mechanism enabling the prediction of trajectories and computation of constraint and actuation forces. Simulation results are provided in support of the theory developed.

Topics: Robots
Commentary by Dr. Valentin Fuster
2009;():363-370. doi:10.1115/DSCC2009-2570.

In order to autonomously navigate in an unknown environment, a robotic vehicle must be able to sense obstacles, determine their velocities, and follow a clear path to a goal. However, the perceived location and motion of the obstacles will be uncertain due to the limited accuracy of the robot’s sensors. Thus, it is necessary to develop a system that can avoid moving obstacles using uncertain sensor data. The method proposed here is based on a certainty occupancy grid—which has been used to avoid stationary obstacles in an uncertain environment—in conjunction with the velocity obstacle concept—which allows a robot to avoid well-known moving obstacles. The combination of these two techniques leads to velocity occupancy space: a search space which allows the robot to avoid moving obstacles and navigate efficiently to a goal using uncertain sensor data.

Commentary by Dr. Valentin Fuster
2009;():371-378. doi:10.1115/DSCC2009-2675.

A functionality test at the level of individual muscles by investigating the activity of a muscle of interest on various tasks may enable muscle-level force grading. This paper proposes a new method for muscle function tests using an exoskeleton robot for obtaining a wider variety of muscle activity data than standard motor tasks, e.g., pushing a handle by his/her hand. A computational algorithm systematically computes control commands to a wearable robot with actuators (an exoskeleton robot, or a power-assisting device) so that a desired muscle activation pattern for target muscle forces is induced. This individual muscle control method enables users (e.g., therapists) to efficiently conduct neuromuscular function tests for target muscles by arbitrarily inducing muscle activation patterns. Simulation results justify the use of an exoskeleton robot for muscle function testing in terms of the variety of muscle activity data.

Topics: Robots , Testing , Muscle
Commentary by Dr. Valentin Fuster
2009;():379-386. doi:10.1115/DSCC2009-2699.

The goal of this research is to develop energetically efficient robots that can produce stable gait. One cannot ignore the promise of robots that can walk by using very little energy. Gravity powered bipeds provide ample proof that this is possible. It is conceivable that mechanisms that are much simpler than legged robots can also produce gravity powered locomotion. It is also plausible that one can get deeper understanding of the dynamics of such simple mechanisms. In addition, evolutionary biology tells us that one can learn a great deal by studying the hereditary traits of organisms that evolve over generations. Thus, we study a chain of mechanisms that span from the very simple to the progressively more complicated. Each member derived from the previous one by making well defined incremental changes. Hence, our goal is to generate a family of planar mechanisms, with members that can be as simple as a bouncing ball or as complex as a five link biped. This presentation will be the first step of this study.

Topics: Gravity (Force)
Commentary by Dr. Valentin Fuster
2009;():387-393. doi:10.1115/DSCC2009-2702.

The paper presents a swarm based lane detection system which uses cooperating agents acting on different regions of the images obtained from an onboard robot camera. These agents act on the image and communicate with each other to find out the possible location of the lane in the image. The swarm agents finalize their locations based on a set of rules which includes each other’s relative position and their previous locations. The swarm agents place themselves on the lane and generate a guidance path for the robot. This proposed lane detection method is is fast and robust to noises in the image. It is faster than the regression methods commonly used and can overcome the problem of noisy image to a good extent.

Commentary by Dr. Valentin Fuster
2009;():395-401. doi:10.1115/DSCC2009-2703.

In this paper the use of support vector machines (SVM) for path planning has been investigated through a Player/Stage simulation for various case studies. SVMs are maximum margin classifiers that obtain a non-linear class boundary between the data sets. In order to apply SVM to the path planning problem, the entire obstacle course is divided in to two classes of data sets and a separating class boundary is obtained using SVM. This non-linear class boundary line determines the heading of the robot for a collision-free path. Complex obstacles and maps have been created in the simulation environment of Player/Stage. The effectiveness of SVM for path planning on unknown tracks has been studied and the results have been presented. For the classification of newly detected data points in the unknown environment, the k-nearest neighbors algorithm has been studied and implemented.

Commentary by Dr. Valentin Fuster

Robust Control

2009;():403-409. doi:10.1115/DSCC2009-2551.

The brief instability concept in Linear Parameter Varying (LPV) systems allows the linear system to be unstable for some values of the LPV parameters so that instability occurs only for a short period of time. The present paper takes advantage of an extension of the notion of the brief instability to the LPV systems with time-delay in their dynamics to examine the performance degradation in Fault Tolerant Control (FTC) systems in the presence of false identification of the fault signals. The paper provides tools for the stability and performance analysis of such systems, where performance is evaluated in terms of induced L 2 -gain. The results presented in the paper demonstrate that stability and performance can be evaluated by examining the feasibility of a parameterized set of Linear Matrix Inequalities (LMIs). The paper provides the analysis conditions to guarantee the asymptotic stability and H ∞ performance for FTC systems, in which instability, due to the false identification of the fault signals, is allowed to take place for a short period of time. A numerical example is presented to illustrate the qualifications of the proposed analysis and synthesis conditions for treating brief instability in the delayed FTC systems.

Commentary by Dr. Valentin Fuster
2009;():411-418. doi:10.1115/DSCC2009-2592.

In this paper a graphical technique is introduced for finding all continuous-time or discrete-time proportional integral derivative (PID) controllers that satisfy a weighted sensitivity constraint of an arbitrary order transfer function with time delay. These problems can be solved by finding all achievable PID controllers that simultaneously stabilize the closed-loop characteristic polynomial and satisfy constraints defined by a set of related complex polynomials. The key advantage of this procedure is that this method depends only on the frequency response of the system. The delta operator is used to describe the controllers in a discrete-time model, because it not only possesses numerical properties superior to the discrete-time shift operator, but also converges to the continuous-time controller as the sampling period approaches zero. A unified approach allows us to use the same procedure for discrete-time and continuous-time weighted sensitivity design of PID controllers.

Commentary by Dr. Valentin Fuster
2009;():419-426. doi:10.1115/DSCC2009-2623.

In this work, a Linear Parameter-Varying (LPV) control method is used to compensate the hysteretic behavior of a Shape Memory Alloy (SMA) wire. Controller is implemented on an experimental system which consists of a pre-tension spring and a mass actuated with a thin SMA wire. The hysteretic characteristic of the SMA wire is modeled using the Preisach model and the model is verified both for the major and minor hysteresis loops. The small signal linear gain of the Preisach model is used as a scheduling stiffness variable. The parameter-dependent controller is scheduled based on the real time measurement of the stiffness variable. An H∞ controller is also synthesized by representing the hysteresis as a parametric uncertainty and comparisons are made with LPV gain scheduling controllers using similar weights for both controllers. Experimental trajectory tracking results show that the LPV Gain Scheduling controller has a better response and the hysteresis uncertainty is compensated for the full range of stiffness variability.

Topics: Wire , Gain scheduling
Commentary by Dr. Valentin Fuster
2009;():427-435. doi:10.1115/DSCC2009-2685.

This paper presents a methodology for analyzing the H 2 guaranteed cost performance of a discrete-time LTI system with unstructured dynamic uncertainty. Using the methods of guaranteed cost control, an upper bound on H 2 guaranteed cost performance over unstructured parametric uncertainty is formulated in terms of feasibility of a linear matrix inequality. It is then shown that the feasibility of this inequality also guarantees the same level of performance also over unstructured dynamic uncertainty. This is then used to formulate the problem of finding the best upper bound on H 2 guaranteed cost performance over unstructured causal dynamic uncertainty as a semi-definite program. Finally, it is shown that this optimization problem can be solved efficiently and accurately using discrete algebraic Riccati equations.

Commentary by Dr. Valentin Fuster
2009;():437-443. doi:10.1115/DSCC2009-2747.

For sampled-data control systems, where a continuous-time plant is under digital control, one of the most important design parameter is the sample rate/period. Higher sample rate typically is associated with the need of high performance components and processors that results in higher system cost. In this paper, we propose an approach to determine the slowest sample rate for a sampled-data control system that will achieve the desired performance and robustness specifications. An optimization problem can be formulated using lifting technique to parameterize sample period for a sampled-data control system. The utility of the proposed approach is numerically verified through the control systems design of the media advance system of an inkjet printer.

Topics: Control systems
Commentary by Dr. Valentin Fuster

Ionic Polymer-Metal Composite (IPMC) Sensors and Actuators: Modeling, Control, and Applications

2009;():445-452. doi:10.1115/DSCC2009-2587.

Low power consumption and activation voltage combined with high flexibility and minimal weight make Ionic Polymer Metal Composites (IPMCs) well-suited for miniaturized underwater propulsion systems. In this series of papers, we comprehensively discuss the flow field induced by an IPMC strip vibrating in a quiescent aqueous environment by performing complementary physical experiments and numerical simulations. The experimental results are presented in this paper. Planar particle image velocimetry is used to measure the time-averaged flow field of a vibrating IPMC. The momentum transferred to the fluid is computed to estimate the mean thrust generated by the vibrating actuator. We find that the mean thrust increases with the Reynolds number, defined by the maximum tip speed and IPMC length, and is only marginally affected by the relative vibration amplitude. Detailed understanding of the flow environment induced by a vibrating IPMC can guide the optimization of IPMC-based propulsion systems for bio-mimetic robotic swimmers.

Commentary by Dr. Valentin Fuster
2009;():453-460. doi:10.1115/DSCC2009-2588.

Ionic Polymer Metal Composite (IPMC) actuators have shown promise as miniature underwater propulsors due to their high flexibility, reduced weight, and low activation voltage and power consumption. In this second of two papers, we discuss numerical simulations of the flow of a viscous fluid generated by a two-dimensional cantilever IPMC actuator vibrating along its fundamental mode shape. We compute the thrust produced by the actuator as a function of its oscillation frequency and maximum tip displacement and show that it is correlated to vortex shedding. We find that vorticity production is prominent at the IPMC tip and increases as the oscillation frequency increases. We analyze the lateral force and the moment exerted by the IPMC on the surrounding fluid. Further, we study the power transferred by the vibrating IPMC to the encompassing fluid. The findings are validated via comparison with the experimental results presented in part 1 of this series.

Commentary by Dr. Valentin Fuster
2009;():461-468. doi:10.1115/DSCC2009-2609.

Reported are advances made in connection with modeling of ionic polymeric metal composite (IPMC) plates undergoing large deformation under an imposed dynamic electric field. Analysis, design and prototyping of sensing or/and actuating plates made with IPMCs requires analytical models of the utilized materials and structures. This paper presents recent advances made towards the development of a computational implementation of a general theory for describing such systems in a way that allows accurate prediction of their behavior within their state space. Continuum mechanics, irreversible thermodynamics, and electrodynamics are utilized to derive the general four dimensional multi-physics field equations of materials used for artificial muscle applications. These applications are particularly important in terms of creating data sheets, thin data keyboards as well as flat speakers made with IPMC plates. The system of governing partial differential equations describing the state evolution of large deflection IPMC plates is derived. The system of these electro-hygro-thermally modified Von-Karman non-linear equations are solved numerically through an adaptive finite element approach through perspectives of geometrical and material nonlinearities. The preliminary results are presented for the case of finite deformation of an IPMC plate.

Commentary by Dr. Valentin Fuster
2009;():469-476. doi:10.1115/DSCC2009-2612.

Ionic polymer-metal composites (IPMCs) are soft materials that can generate large deformation under a low voltage. IPMCs have many potential applications in biomedical, robotic and micro/nano manipulation systems. In this paper, we first present a distributed, nonlinear circuit model for IPMC, which incorporates the nonlinear capacitance, the nonlinear DC resistance, and the effect of surface resistance. The bending displacement is proportional to the total stored charge in IPMC. After discretizing the model in the length direction, we obtain a multiple-segment model which can be represented in the state space for nonlinear control design. The model is validated using experimental data, and we show that a one-segment model can predict the current and displacement response reasonably well. A model-based nonlinear controller is proposed for IPMC actuators, where feedback linearization is applied. Simulation results show that model-based nonlinear controller delivers better performance than a traditional PI controller in terms of the tracking error, control effort, and robustness to sensing noises.

Commentary by Dr. Valentin Fuster
2009;():477-484. doi:10.1115/DSCC2009-2660.

In this paper, we study the free-locomotion of a miniature bio-mimetic underwater vehicle inspired by carangiform swimming fish. The vehicle is propelled by a vibrating Ionic Polymer Metal Composite (IPMC) attached to a compliant passive fin. The IPMC vibration is remotely controlled through the vehicle’s onboard electronics that consists of a small-sized battery pack, an H-Bridge circuit, and a wireless module. The planar motion of the vehicle body is described using rigid-body dynamics. Hydrodynamic effects, such as added mass and damping, are included in the model to enable a thorough description of the vehicle’s surge, sway, and yaw motions. The time-varying actions exerted by the vibrating IPMC on the vehicle body, including thrust, lift, and moment, are estimated by combining force and vibration measurements with reduced order modeling based on modal analysis. The model predictions are validated through experimental results on the planar motion of the fish-like robotic swimmer.

Commentary by Dr. Valentin Fuster
2009;():485-491. doi:10.1115/DSCC2009-2700.

An ionic polymer-metal composite (IPMC) is an electroactive material that bends when electrically stimulated and generates electric current when bent. In this paper we investigate a coupled IPMC sensor-actuator using both the sensing and actuation properties of these electroactive materials. We describe the design of a coupled IPMC sensor-actuator, the feedback controller and the experimental evaluation of the system. Experimental results show the feasibility of closed-loop control of IPMC actuator with a mechanically coupled IPMC sensor.

Commentary by Dr. Valentin Fuster

Vehicle Dynamics Control

2009;():493-500. doi:10.1115/DSCC2009-2548.

We present new development of a hybrid physical/dynamic tire/road friction model for real-time friction estimation and control. We extend the LuGre tire/road friction model by considering the physical model-based deformation distribution on the tire/road contact patch. Relationship between the physical friction model and the LuGre dynamic friction model has been built and developed. We have shown that the LuGre dynamic friction model predicts the similar deformation and stress characteristics of the physical model, and therefore the friction model parameters can be interpreted with physical meaning and estimated experimentally. We demonstrate preliminary model comparison study through the “smart tire” sensor measurements on a mobile robot platform.

Topics: Friction , Roads , Tires
Commentary by Dr. Valentin Fuster
2009;():501-508. doi:10.1115/DSCC2009-2600.

This paper describes a 2D bristle contact friction model which is capable of modeling and simulating frictional behavior in both sliding and sticking regimes occurring in general 3D rigid-body contact. The model extends the 1D integrated bristle friction model to a 2D space by allowing the “bristle spring” to not only stretch along the direction of the relative velocity but also rotate due to the direction change of the velocity or motion trend in the common tangential plane of the contacting surfaces involved at the contact point of interest. With such an extension, the resulting friction model can be readily used to compute 3D contact friction forces in both sticking and sliding regimes for a general 3D contact dynamics model working with a multibody dynamics simulation application. Several simulation examples are provided to demonstrate the effectiveness of the model for predicting the experimentally seen frictional behavior such as sticking, stickslip, and sliding.

Commentary by Dr. Valentin Fuster
2009;():509-516. doi:10.1115/DSCC2009-2717.

We present a nonlinear analysis of vehicle motion using a hybrid physical/dynamic tire/road friction model. The advantage of the proposed LuGre dynamic tire/road friction model is the simple and attractive structural properties for real-time friction estimation and control. Moreover, the model provides a property of capturing coupling effects between the longitudinal and lateral friction forces. We take advantages of these properties and analyze the vehicle lateral motion stability. We have shown that the existence of longitudinal slip affects the lateral motion stability. The quantitative analysis and relationship are also demonstrated through numerical simulation examples.

Commentary by Dr. Valentin Fuster
2009;():517-524. doi:10.1115/DSCC2009-2728.

The topic of this paper is the design and analysis of a control strategy for a hill starting assistance for commercial vehicles equipped with an electric parking brake (EPB). The present hill start assistance system is thought to be installed in a vehicle with manual transmission and it is based on a very simple and low complexity hardware layout. The hill start strategy herein proposed is based on events detection: the detection of the torque transmission from the engine to the wheels and the detection the vehicle movement before the EPB -release. Although all the possible scenarios cannot be tackled, because of the layout constraints and of the presence of the driver in the control loop, the effectiveness of the proposed framework has been experimentally tested.

Commentary by Dr. Valentin Fuster
2009;():525-532. doi:10.1115/DSCC2009-2737.

This article deals with the nonlinear feedback regulation of the longitudinal traction forces for high-speed vehicles, possibly over a low friction surface. Hybrid models of the longitudinal vehicle dynamics incorporating load transfer effects, a crucial element in advanced driving techniques, are derived. The designed hybrid regulator allows the tracking of a given friction force profile in the presence of known disturbances and unknown model uncertainties. Simulations show good performance of the proposed hybrid regulator under all operating conditions.

Commentary by Dr. Valentin Fuster
2009;():533-540. doi:10.1115/DSCC2009-2773.

A tire model is an essential element in the vehicle controller development and various complexities of tire models have been developed and used. It is highly desirable for the control systems to use a tire model that is computationally efficient and easy to implement in control algorithms while providing desired performance. In this paper, a wheel torque controller was developed using a non-linear predictive control theory, 8 degree of freedom vehicle model, and a simplified nonlinear tire model in order to control the vehicle yaw rate and side slip angle. The performance of this controller was compared to that based on well known Magic Formula tire model. The effectiveness and limitations of the proposed controller were discussed through simulation.

Commentary by Dr. Valentin Fuster

Engine Control and Identification

2009;():541-548. doi:10.1115/DSCC2009-2523.

Cylinder pressure is one of the most important parameters characterizing the combustion process in an internal combustion engine. The recent developments in piezoelectric pressure transducers and progress in on-line computational throughput facilitate the use of cylinder pressure as a feedback signal for engine combustion control. However, a typical production cylinder pressure sensor is subject to noise and offset issues that require signal processing methods, including averaging over several engine cycles, in order to extract a pressure trace sufficiently accurate for combustion characterization. This limits the application of cylinder pressure sensing to off-line applications. In order to enable closed-loop combustion control using cylinder pressure feedback, this study proposes a real-time estimation algorithm that extracts the pressure signal on a crank-angle basis. A simplified thermodynamic model for Diesel engine combustion is derived to predict the in-cylinder pressure. The model is then adapted to model-based estimation, by applying an Extended Kalman Filter in conjunction with a recursive least squares estimation. The resulting estimator is tested on a high-fidelity Diesel engine model for different operating conditions. The results obtained show the effectiveness of the estimator in reconstructing the cylinder pressure and in rejecting measurement noise and modeling errors.

Commentary by Dr. Valentin Fuster
2009;():549-556. doi:10.1115/DSCC2009-2553.

Homogeneous charge compression ignition (HCCI) is one of the most promising piston-engine concepts for the future, providing significantly improved efficiency and emissions characteristics relative to current technologies. This paper presents a framework for controlling a multi-cylinder HCCI engine with exhaust recompression and direct injection of fuel into the cylinder. A physical model is used to describe the HCCI process, with the model states being closely linked to the thermodynamic state of the cylinder constituents. Separability between the effects of the control inputs on the desired outputs provides an opportunity to develop a simple linear control scheme, where the fuel is used to control the work output and the valve timings are used to control the phasing of combustion. Experimental results show good tracking of both the work output and combustion phasing over a wide operating region. In addition, the controller is able to balance out differences between cylinders, and reduce the cycle-to-cycle variability of combustion.

Commentary by Dr. Valentin Fuster
2009;():557-563. doi:10.1115/DSCC2009-2649.

All vehicle manufacturers implement an air-to-fuel ratio (AFR) control system for emissions reduction in gasoline engines. When using a model based control structure, it is vital to capture the underlying dynamics of the plant as accurately as possible, thus facilitating a robust control design that meets the emissions regulation requirements. One of the leading sources of uncertainty in the engine model is the variable plant delay. Although the delay could be modeled using a look-up table of steady-state delay values, during transients when AFR control is most important the steady-state delay poorly approximates the true delay. An exhaust geometry based delay model was developed previously within the framework of a model based control design for AFR control of stoichiometric engines. In this paper, it is shown that using this model the delay can be predicted with a significantly higher accuracy especially during transients, thus improving emissions performance. Because the plant delay plays a destabilizing role in feedback control, the utility of such a model is also to minimize phase errors between the predicted and measured equivalence ratio (EQR) in a reference tracking control setting.

Commentary by Dr. Valentin Fuster
2009;():565-572. doi:10.1115/DSCC2009-2697.

Closed-loop control of diesel combustion is of great interest for improving conventional Diesel engine combustion as well as facilitating new combustion modes such as Homogenous Charge Compression Ignition and other low NOx regimes. Most generalized feedback control systems that can be applied to this problem require a reference or set-point which the control attempts to achieve. Diesel engines are well known for having many degrees of freedom which poses a problem in generating valid set-points for all possible conditions encountered in practice. This problem is compounded by the fact that these set-points are usually determined in steady state operation further limiting the space where set-points can be defined. Kernel-based methods are applied to this problem as a method of generating valid setpoints when operating in regions outside of the space where set-points are defined. This is most useful during transient conditions where conditions such as exhaust gas recycle level, manifold air flow, and fuel mass are far from the steady state values.

Topics: Combustion , Engines
Commentary by Dr. Valentin Fuster
2009;():573-580. doi:10.1115/DSCC2009-2698.

At present, Diesel engine combustion in most production engines is controlled via open-loop control. Increasing pressure from tightening emissions standards and on-board diagnosis requirements has made closed-loop combustion a possibility for production engines in the near future. For new combustion concepts, such as Homogeneous Charge Compression Ignition and other low NOx regimes, the need for closed-loop combustion control is very strong. In this work, the applicability of closed-loop combustion control for controlling the variability between cylinders in conventional Diesel combustion is explored through the use of a high-fidelity engine model. The problem is formulated such that the optimal performance of two different closed-loop control concepts can be evaluated through optimization rather than via control design. It is found that, for the types of disturbances occurring in a non-faulty engine, that control of individual cylinders leads to small performance gains compared to fuel bank control.

Commentary by Dr. Valentin Fuster
2009;():581-588. doi:10.1115/DSCC2009-2708.

Flexible fuel vehicles (FFVs) can operate on a blend of ethanol and gasoline in any volumetric concentration of up to 85% ethanol (93% in Brazil). Existing FFVs rely on ethanol sensor installed in the vehicle fueling system, or on the ethanol-dependent air-to-fuel ratio (AFR) estimated via an exhaust gas oxygen (EGO) or λ sensor. The EGO-based ethanol detection is desirable from cost and maintenance perspectives but has been shown to be prone to large errors during mass air flow sensor drifts [1, 2]. Ethanol content estimation can be realized by a feedback-based fuel correction of the feedforward-based fuel calculation using an exhaust gas oxygen sensor. When the fuel correction is attributed to the difference in stoichiometric air-to-fuel ratio (AFR) between ethanol and gasoline, it can be used for ethanol estimation. When the fuel correction is attributed to a mass air flow (MAF) sensor error, it can be used for sensor drift estimation and correction. Deciding under which condition to blame (and detect) ethanol and when to switch to sensor correction burdens the calibration of FFV engine controllers. Moreover, erroneous decisions can lead to error accumulation in ethanol estimation and in MAF sensor correction. In this paper, we present a cylinder air flow estimation scheme that accounts for MAF sensor drift or bias using an intake manifold absolute pressure (MAP) sensor. The proposed fusion of the MAF, MAP and λ sensor measurements prevents severe mis-estimation of ethanol content in flex fuel vehicles.

Topics: Sensors , Fuels , Vehicles , Ethanol
Commentary by Dr. Valentin Fuster

ASME/Bath Fluid Power Symposium: Valve Design and Analysis

2009;():589-596. doi:10.1115/DSCC2009-2535.

This paper reports on an initial investigation of a switched inertance device (‘SID’). Using this device, flow and pressure can be varied by a means that does not rely on dissipation of power. The device can provide a step-up or step-down of pressure or flowrate, analogous to a hydraulic transformer. Simulated and experimental results on a prototype device show a promising performance. The device could potentially provide very significant reduction in power consumption over conventional valve-controlled systems, provided that noise issues and some other practical problems can be overcome.

Commentary by Dr. Valentin Fuster
2009;():597-604. doi:10.1115/DSCC2009-2556.

For the unmanned underwater vehicles (UUVs) with power between 20kW and 60kW, the electric motor is too heavy, while the power of variable displacement motor is too high. In this paper, the high speed on/off hydraulic control propulsion system for a UUV is designed. The three main parts of the system, the hydraulic subsystem, the propeller, and the amplifier are described in detail. An integrated valve-manifold is designed, which can be installed in the rear of motor to simplify the structure and to improve the response performance of system. A novel quasi PFM amplifier which can make sure the enough switching time of on/off valve and good propeller control performance in the entire working conditions is developed. Since the propeller is driven in switching mode, the speed fluctuation cannot be avoided. Therefore, the relations between the system parameters and speed fluctuation are analyzed by mathematical method, from which the approximate analytical solutions of speed fluctuation and average speed are obtained. Based on these, a flywheel is added on the propeller shaft for the purpose of filtering the speed fluctuation. To choose the inertia of the flywheel, a trade-off between the fluctuation degree and the response speed has been done. Experimental results prove the system has good dynamic performance that the respond time of on/off valve is only about 3.5ms and the fluctuation degree is 12.55% while the response speed of propeller is quick. Simulation results show that the speed closed-loop control system has satisfactory transient and steady-state performance and good tracking ability.

Commentary by Dr. Valentin Fuster
2009;():605-612. doi:10.1115/DSCC2009-2611.

This communication proposes an original method to define a low-cost manufacturing process for hydraulic spool valves in the case of very low production rate. A valve is usually completely defined by three characteristic curves obtained by two tests: a pressure gain test and a flow gain test. The aim is to limit the number of tests and in particular to avoid the flow gain test that is quite long to achieve. The method only uses a single metrology operation and the pressure gain curve to manage the spool flanks grinding while ensuring the specified hydraulic characteristics. The method is established by a model-based approach. It is then validated by experimental tests on two different valves.

Topics: Manufacturing , Valves
Commentary by Dr. Valentin Fuster
2009;():613-620. doi:10.1115/DSCC2009-2617.

A method for significantly reducing the losses associated with an on/off controlled hydraulic system is proposed. There has been a growing interest in the use of on/off valves to control hydraulic systems as a means of improving system efficiency. While on/off valves are efficient when they are fully open or fully closed, a significant amount of energy can be lost in throttling as the valve transitions between the two states. A soft switching approach is proposed as a method of eliminating the majority of these transition losses. The operating principle of soft switching is that fluid can temporarily flow through a check valve or into a small chamber while valve orifices are partially closed. The fluid can then flow out of the chamber once the valve has fully transitioned. Thus, fluid flows through the valve only when it is in its most efficient fully open state. A model of the system is derived and simulated, with results indicating that the soft switching approach can reduce transition and compressibility losses by 79%, and total system losses by 66%. Design equations are also derived. The soft switching approach has the potential to improve the efficiency of on/off controlled systems and is particularly important as switching frequencies are increased. The soft switching approach will also facilitate the use of slower on/off valves for effective on/off control; in simulation, a valve with soft switching matched the efficiency an on/off valve that was 5 times faster.

Topics: Valves
Commentary by Dr. Valentin Fuster
2009;():621-628. doi:10.1115/DSCC2009-2631.

This paper considers the feasibility of a new type of voice coil motor direct drive flow control servo valve. The proposed servo valve controls the flow rate using only a direct measurement of the spool position. A neural network is used to estimate the flow rate based on the spool position, velocity and coil current. The estimated flow rate is fed back to a closed loop controller. The feasibility of the concept is established using simulation techniques only at this point. All results are validated by computer co-simulation using AMESim and Simulink. A simulated model of a VCM-DDV (Voice Coil Motor-Direct Drive Valve) and hydraulic test circuit are built in an AMESim environment. A virtual digital controller is developed in a Simulink environment in which the feedback signals are received from the AMESim model; the controller outputs are sent to the VCM-DDV model in AMESim (by interfacing between these two simulation packages). A LQR (Linear Quadratic Regulator) state feedback and nonlinear compensator controller for spool position tracking is considered as this is the first step for flow control. A flow rate control loop is subsequently included via a neural network flow rate estimator. Simulation results show that this method could control the flow rate to an acceptable degree of precision, but only at low frequencies. This kind of valve can find usage in open loop hydraulic velocity control in many industrial applications.

Commentary by Dr. Valentin Fuster
2009;():629-636. doi:10.1115/DSCC2009-2763.

Efficient high-speed on/off valves are a critical technology for enabling digital control of hydraulic systems via pulse-width-modulation (PWM). High-speed valves, when used in virtually variable displacement pumps (VVDP), increase system bandwidth and reduce output pressure ripple by enabling higher PWM frequencies. Our approach to achieving high speed and large flow area with low actuation power is a unidirectional rotary valve designed specifically for PWM. In comparison to conventional valves, the rotary valve reduces valve actuation power from a cubic dependence on PWM frequency to a square dependence by eliminating motion reversals during transition. This paper presents experimental data that validates the rotary valve concept, valve design equations, and dynamic model of a rotary valve based VVDP. Our unoptimized prototype exhibits 65 % efficiency at 50 % displacement and 15 Hz PWM frequency while the validated model projects that an optimized valve is capable of achieving 85 % efficiency at 15 Hz and 73 % at 75 Hz.

Topics: Modeling , Valves
Commentary by Dr. Valentin Fuster

Fault Diagnostics and Prognostics

2009;():637-643. doi:10.1115/DSCC2009-2667.

This paper presents a comparative evaluation of two classification schemes that can be used to accurately diagnose the health of machines in the presence of sensor failure. In the developed approach, multiple sensors acquire vibration and sound signals from a machine and the signals are represented using the Wavelet Packet Transform (WPT). A “wrapper” feature selection procedure is used to reduce the size of the feature set without sacrificing the classification accuracy. The performance of a Radial Basis Function Network (RBFN) is compared with that of a Support Vector Machine (SVM) by simulating and monitoring machine and sensor faults in an industrial fish cutting machine. Initial results show an 85% reduction in feature set size for an RBFN and a 92.5% reduction in feature set size for a SVM.

Commentary by Dr. Valentin Fuster
2009;():645-652. doi:10.1115/DSCC2009-2680.

Prediction of a bearing service life is traditionally achieved by empirical or physical models, which have their own strengths and limitations. In an effort to combine the strengths of these modeling approaches, this research investigates the concept of Multi-Time Scale Modeling (MTSM). Specifically, a MTSM strategy for bearing life prognosis is developed by correlating experimentally acquired bearing vibration data with physics based model of microscopic growth of crack size. The strategy is composed of a fast scale empirical model (e.g., root mean square value of vibration), a slow scale physical model (e.g., change of crack length over one loading cycle), and a model coupling mechanism (e.g., bidirectional mapping functions). The fast and slow scale models are obtained by polynomial regression analysis and using the concept of the Paris Law, respectively. The coupling mechanism is established through introduction of dynamic mass into the model. The improvement in bearing service life prediction, obtained by the presented MTSM strategy is experimentally validated.

Topics: Bearings , Modeling
Commentary by Dr. Valentin Fuster
2009;():653-660. doi:10.1115/DSCC2009-2692.

This paper describes the model-based fault detection and diagnosis for electromagnetic actuators. Due to hysteresis behavior, the model of Jiles and Atherton is used to capture these hysteretic effects. It is shown that the use of this model results in superior model fidelity and allows to detect even tiny, incipient faults at the actuator. The method has been applied to a direct-driven proportional valve of an electro-hydraulic servo axis and has been tested extensively on a real testbed.

Commentary by Dr. Valentin Fuster
2009;():661-668. doi:10.1115/DSCC2009-2731.

The driveline components of engine cold-test cells undergo large torsional vibrations during transient tests as the system resonances are excited by various engine harmonics. The excessive torsional vibrations not only compromise the structural reliability of the system but also make the fault detection process difficult by preventing accurate measurement of gear noise and by compromising the quality of diagnostic torque waveforms. This paper presents modeling of an engine cold-test cell and a methodology to quantify signal distortion levels by using the proposed distortion metrics which are based on harmonic order amplitude ratios. A rigorous validation of the simulation model using both torque amplitude and waveform distortion comparisons of experimental and simulation data is conducted. The model is used to identify the driveline inertia and stiffness parameters that can help reduce high torsional resonant amplitudes as well as waveform distortion. The design modifications are implemented in a production test cell which helped to control the torsional vibration and diagnostic signal degradation issues with a corresponding increase in the sensitivity to faults.

Commentary by Dr. Valentin Fuster
2009;():669-676. doi:10.1115/DSCC2009-2787.

In this paper we present a formulation of an aircraft systems model for prognosis-based control. An aircraft systems model with hydraulic actuators for control surfaces is developed in SIMULINK and is connected with a model for predicting degradation of an actuator. An LQR based feedback control is employed to compensate for changes in the plant dynamics and a feedforward controller is used to achieve better performance through set-point tracking, while taking into account the current actuator degradation rate. The advantage of the combination of feedback and feedforward control strategies is then demonstrated through reduction of degradation of the actuator while achieving a minimum loss of performance using constrained optimization.

Topics: Actuators , Aircraft
Commentary by Dr. Valentin Fuster
2009;():677-684. doi:10.1115/DSCC2009-2790.

With requirements for on-board diagnostics on diesel engines becoming more stringent for the coming model years, diesel engine manufacturers must improve their ability to identify fault conditions that lead to increased exhaust emissions. This paper proposes a statistical classifier model to identify the state of the engine, i.e. healthy or faulty, using an optimal number of sensors based on the data acquired from the engine. The classification model proposed in this paper is based on Sparse Linear Discriminant Analysis. This technique performs Linear Discriminant Analysis with a sparseness criterion imposed such that classification, dimension reduction and feature selection are merged into one step. It was concluded that the analysis technique could produce 0% misclassification rate for the steady-state data acquired from the diesel engine using five input variables. The classifier model was also extended to transient operation of the engine. The misclassification rate in the case of transient data was reduced from 31% to 26% by using the steady-state data trained classifier using thirteen variables.

Topics: Diesel engines
Commentary by Dr. Valentin Fuster

Sensing, Modeling, and Control of the Human Body

2009;():685-687. doi:10.1115/DSCC2009-2541.

The optimal input current for a reduced neuron model and a specific target spiking time is obtained. The objective of optimization is to minimize the total input energy to the system subject to a zero net input integral over the time horizon. This “charge-balance constraint” ensures that no net external charge is injected into the neuron. The results are compared to optimal currents for which the charge-balance constraint is not imposed.

Commentary by Dr. Valentin Fuster
2009;():689-696. doi:10.1115/DSCC2009-2575.

The goal of this work is to present methodology to first evaluate the performance of an in vivo spine system and then to synthesize optimal neuromuscular control for rehabilitation interventions. This is achieved 1) by determining control system parameters such as static feedback gains and delays from experimental data, 2) by synthesizing the optimal feedback gains to attenuate the effect of disturbances to the system using modern control theory, and 3) by evaluating the robustness of the optimized closed-loop system. We also apply these methods to a postural control task, with two different control strategies, and evaluate the robustness of the spine system with respect to longer latencies found in the low back pain population. This framework could be used for rehabilitation design as discussed at the end of the paper.

Commentary by Dr. Valentin Fuster
2009;():697-704. doi:10.1115/DSCC2009-2598.

Developed in this paper is the set up of a spatiotemporal (multi-domain) numerical model for intravascular ultrasound (IVUS). The transmission line matrix method (TLM) is used to mimic the propagation of the sound wave through the circular cross-section of an artery. Different propagation scenarios (healthy and abnormal arteries) are illustrated in this work showing how this model captures the behavior of these waves. The model is evaluated by means of regular TLM models established in the literature.

Topics: Ultrasound
Commentary by Dr. Valentin Fuster
2009;():705-712. doi:10.1115/DSCC2009-2642.

Neuromuscular electrical stimulation (NMES) is a promising technique that has the potential to restore functional tasks in persons with movement disorders. Clinical and commercial NMES products exist for this purpose, but a pervasive problem with current technology is that overstimulation of the muscle (among other factors) leads to muscle fatigue. The objective of the current effort is to develop a NMES controller that incorporates the effects of muscle fatigue through an uncertain function of the calcium dynamics. A neural network-based estimate of the fatigue model mismatch is incorporated in a nonlinear controller through a backstepping based method to control the human quadriceps femoris muscle undergoing non-isometric contractions. The developed controller is proven to yield uniformly ultimately bounded stability for an uncertain nonlinear muscle model in the presence of bounded nonlinear disturbances (e.g., spasticity, delays, changing load dynamics).

Commentary by Dr. Valentin Fuster
2009;():713-720. doi:10.1115/DSCC2009-2658.

Processing electromyographic (EMG) signals for force estimation has many unknown variables that can influence the outcome or interpretation of the recorded EMG signal significantly. An array of filtering methods have been proposed over the past few years with the objective to classify motion for use in prosthetic hands. In this paper, we explore the optimal parameter settings of a set of Bayesian based EMG filters with the objective to use the filtered EMG data for system identification. System identification is utilized to establish a relationship between the measured EMG data and the generated force developed by fingers in a human hand. The proposed system identification is based on nonlinear Hammerstein-Wiener models. Optimization is also applied to find the optimal parameter settings for these nonlinear models. Genetic Algorithm (GA) is used to conduct the optimization for both, the optimal parameter settings for the Bayesian filters as well as the Hammerstein-Wiener model. The experimental results and optimization analysis indicate that the optimization can yield significant improvement in data accuracy and interpretation.

Commentary by Dr. Valentin Fuster
2009;():721-723. doi:10.1115/DSCC2009-2690.

Traditional electromyopgrahic (EMG) measurements are based on single sensor information. Due to the arrangement of skeletal muscle fibers for hand motions, cross talk is an inherent problem when inferring motion/force potentials from EMG data. This paper studies means of using sensor arrays to infer better motion/force potential for prosthetic hands. In particular, a surface electromyographic (sEMG) sensor array is used to investigate multiple model fusion techniques. This paper provides a comparison between three statistical model selection criteria. The sEMG signals are pre-processed using four filters, Butterworth, Chebyshev type-II, as well as Bayesian filters such as the Exponential and Half-Gaussian filter. Output Error (OE) models were extracted from sEMG data and hand force data and compared using a Bayesian based fusion model. The four different filters effect were quantified based on the OE models performance in matching the actual measured data. The comparison indicates a preference for using the sensor fusion technique with preprocessed EMG data using the Half-Gaussian Bayesian filter and the Kullback Information Criterion (KIC).

Commentary by Dr. Valentin Fuster

Adaptive Control

2009;():725-732. doi:10.1115/DSCC2009-2568.

A novel method of controller tuning is introduced to achieve a desired closed-loop response. It uses the same strategy as Iterative Feedback Tuning (IFT), but instead of relying on a scalar cost function of the performance error between the desired response and system response, it utilizes the expanded version of the performance error in the time-scale domain for estimating the suitable controller parameters. The proposed method relies on the enhanced delineation of output sensitivities in the time-scale domain to identify regions in the time-scale domain wherein the performance error can be attributed to individual controller parameters [1]. It then relies on the error association in each region to estimate the corresponding controller parameter. It is shown that given a realistic desired response for the closed-loop system, the proposed method can lead to satisfactory controller parameters. It is also shown that the results from this method can be integrated with those from IFT to represent the best of the two solutions from the time and time-scale domains.

Commentary by Dr. Valentin Fuster
2009;():733-740. doi:10.1115/DSCC2009-2687.

To obtain a higher level of contouring motion control performance for linear-motor-driven multi-axes mechanical systems subject to significant nonlinear cogging forces, both coordinated control of multi-axes motions and effective compensation of cogging forces are necessary. In addition, the effect of unavoidable velocity measurement noises needs to be carefully examined and sufficiently attenuated. To solve these problems simultaneously, in this paper, a discontinuous projection based desired compensation adaptive robust contouring controller is developed by explicitly taking into account the specific characteristics of cogging forces in the controller designs and employing the task coordinate formulation for coordinated motion controls. Specifically, based on the largely periodic nature of cogging forces with respect to position, design models consisting of known sinusoidal functions of positions corresponding to the main harmonics of the force ripple waveforms with unknown weights are used to approximate the unknown cogging forces. Theoretically, the resulting controller achieves a guaranteed transient performance and final contouring accuracy in the presence of both parametric uncertainties and uncertain nonlinearities. In addition, the controller also achieves asymptotic output tracking when there are parametric uncertainties only. Comparative experimental results obtained on a high-speed industrial biaxial precision gantry driven by linear motors are presented to verify the excellent contouring performance of the proposed control scheme and the effectiveness of the cogging force compensations.

Topics: Force
Commentary by Dr. Valentin Fuster
2009;():741-747. doi:10.1115/DSCC2009-2688.

It was shown in an earlier paper that the authors’ recursive open loop adaptive control strategy can be used as the basis for an adaptive multi-objective control strategy to satisfy design criteria for rotor-bearing systems involving the relative weighting of forces transmitted to the foundation and rotor vibration. This paper describes a complementary-layer adaptive control strategy to limit vibration at critical locations along the rotor, where contact may occur, following changes in operating conditions, e.g. transient disturbances. Experiments are presented to explore the effectiveness of this approach.

Commentary by Dr. Valentin Fuster
2009;():749-756. doi:10.1115/DSCC2009-2721.

The Asymmetric Cell Transmission model can be used to simulate traffic flows in freeway sections. The model is specified by fundamental diagram parameters—determined from mainline data, and on-ramp and off-ramp flows. The mainline flow/density data are efficiently archived and readily available, but the ramp flow data are generally found missing. This paper presents an imputation technique based on iterative learning control to determine these flows. The imputation technique is applied sequentially on all the segments of the freeway, and the ramp flows, which minimize the error between the model calculated densities/flows and measurements are investigated. The stability and convergence of the density and flow errors using the imputation updates is also presented. Finally an example is shown to illustrate its use in a practical scenario.

Commentary by Dr. Valentin Fuster
2009;():757-764. doi:10.1115/DSCC2009-2727.

Cross Coupled Iterative Learning Control (CCILC) has previously been applied to contour tracking problems with planar robots in which both axes can be characterized as similar systems; having similar dynamics and identical hardware. However, there are many repetitive applications in which dissimilar systems cooperate to pursue a primary performance objective. This paper introduces a novel framework to couple dissimilar systems while applying Iterative Learning Control (ILC), showing the ability to noncausally compensate for a slow system with a fast system. In this framework, performance requirements for a primary objective can more readily be achieved by emphasizing an underutilized fast system instead of straining a less-capable slow system. The controller is applied in simulation and experimentally to a micro-Robotic Deposition (μRD) manufacturing system to coordinate a slow extrusion system axis and a fast positioning system axis to pursue the primary performance objective, dimensional accuracy of a fabricated part. Experimental results show a 30 % improvement in fabrication dimensional accuracy with only marginal changes in actuator effort in the slow system, as compared to independently controlled axes.

Commentary by Dr. Valentin Fuster
2009;():765-772. doi:10.1115/DSCC2009-2764.

This paper considers the application of repetitive control in improving the comfort of electric bicycle riders under uphill riding condition. A particular problem addressed is the nonuniform torque input from the rider. The human torque input can be decomposed into two parts: the DC or near-DC local average torque and the fluctuating torque. The latter is due to nonuniform pedal torque production by the rider and is near periodic. It makes the acceleration and the velocity of the bicycle non-uniform, which is pronounced at low speeds. This makes uphill climbing a difficult task for the rider. Repetitive control will be used in this paper to compensate for this component during low speed uphill riding. We will consider both sinusoidal and non-sinusoidal periodic models for the non-uniform torque generated by the rider. The time-varying characteristic of the rider input will be dealt with by both time domain and pedal-angle domain methods. Simulation results for both cases will be shown and compared.

Topics: Bicycles
Commentary by Dr. Valentin Fuster

Modeling and Control of Micro and Nano Systems

2009;():773-780. doi:10.1115/DSCC2009-2595.

Bottom-up fabrication presents the potential applications of improved materials and highly effective devices. Matter at the microscale can be manipulated using micro-placing techniques that rely solely on surface forces. These techniques can be used on rough substrates and allow for fast manipulation. In the present paper, we will investigate generalizing these techniques to manipulate nanoscale samples. The use of dynamic oscillations of the manipulator is proposed to allow for controlled release at the nanoscale. The effects of the adhesion and sample mass on the manipulation outcome are investigated. Simulations are conducted to help select the amplitude and the frequency of the vibrations and to improve the accuracy of the manipulation.

Commentary by Dr. Valentin Fuster
2009;():781-788. doi:10.1115/DSCC2009-2712.

This paper presents an Euler-Bernoulli microcantilever beam model utilized in non-contact Atomic Force Microscopy (AFM) systems. A distributed-parameters modeling is considered for such system. The motions of the microcantilever are studied in a general Cartesian coordinate with an excitation at the base such that beam end with a tip mass is subject to a general force. This general force comprising of two attractive and repulsive parts with high power terms is taken as the atomic intermolecular one which has a relation with the displacement between the tip mass and the surface such that the total general force will be in the form of an implicit nonlinear equation. It is most desired to observe the effects of the base excitation in high frequencies on the tip van der Waals interaction force. Hence, the general force will produce a peak in the FFT spectrum corresponding to the frequency of the base.

Commentary by Dr. Valentin Fuster
2009;():789-794. doi:10.1115/DSCC2009-2742.

To reduce the cost and improve the speed of Atomic Force Microscopy (AFM) in molecular scale imaging of materials, we propose a laser-free AFM scheme augmented with an accurate control strategy for its scanning axes. It employs a piezoresistive sensing device with a high level of accuracy to avoid using the bulky and expensive laser interferometer. Change in the resistance of piezoelectric layer due to the deflection of microcantilever caused by the variation of surface topography is monitored through a Wheatstone bridge. Hence, it captures the surface topography without the use of laser and with nanometer scale accuracy. To improve the speed of imaging, however, a Lyapunov-based robust adaptive control strategy is implemented in the 2-DOF scanning stage. It has been demonstrated in an earlier publication that this control framework has superior performance over the conventional PID controllers typically used in commercial AFMs. The paper, then, demonstrates a set of experiments on a standard AFM calibration sample with 200 nm stepped topography. Results indicate accurate imaging of the sample up to the frequency of 30 Hz, for a 16 μm×16 μm scanning area, proving the feasibility of less costly and high speed AFM-based metrology.

Commentary by Dr. Valentin Fuster
2009;():795-802. doi:10.1115/DSCC2009-2752.

Previously, precise positioning of electrostatic MEMS devices has been achieved using mechanical contact between movable and fixed components. A disadvantage of this approach is that stiction at the contact may eventually lead to device failure. This paper presents an alternative approach, whereby a desired configuration is defined by the intersection of two comb structures—one movable and the other fixed. Extremum-seeking control drives the movable electrode to this desired configuration by maximizing the mutual capacitance between the two comb structures. As in the case of mechanical contact, the device structure primarily determines the actuated configuration rather than precise sensing or actuation. However, because the comb structures never physically contact each other, stiction failure is eliminated.

Commentary by Dr. Valentin Fuster
2009;():803-810. doi:10.1115/DSCC2009-2761.

High precision motion is critical in the semiconductor industry and as feature size continues to decrease, the need for higher precision increases. This paper presents the control system design and integration of a novel multi-degree of freedom precision stage for nano-manufacturing. It is composed of a 6 DOF wafer holder and a 3 DOF module holder. The modules can be chosen for a desired task, such as a nano-imprint lithography module. Capacitance gauges, interferometers, and photo detectors provide position feedback of the stage. Piezo-electric actuators and linear motors, that produce two orthogonal forces each, produce the desired output. Modeling of the coordinate transformation among the spaces of sensors, stage position, and actuators along with dynamic modeling of the stage is presented. Controller design, hardware, and software is described and results comparing simulation and implementation show a closed-loop positioning of less than 1 nm error in x and y and 0.01 arc seconds in θ z with a sensor noise level of less than 0.2 nm.

Commentary by Dr. Valentin Fuster
2009;():811-817. doi:10.1115/DSCC2009-2785.

Piezoelectric tube scanners are commonly used in Scanning Probe Microscopy (SPM) to provide the scanning motion of the tip or sample. Oscillations due to the weakly damped resonances of the scanner are a major source of image distortion in SPM-imaging. In this contribution, multiple input-multiple output (MIMO) self-sensing actuation of a piezoelectric tube scanner is presented, allowing to actively dampen the scanner oscillations. By connecting the tube scanner in a capacitive bridge-circuit, the piezoelectric tube is simultaneously used as a sensor and actuator. In order to enable the use of low-order decentralized controllers, the cross-talk between both axes is reduced by compensating for the capacitive coupling. The MIMO self-sensing actuation allows to actively dampen the scanner’s fundamental resonance by 18dB, while simultaneously reducing the resonance induced coupling by 30dB. Experimental results verify a significant reduction of the scanner oscillations in both positioning axes during fast scanning.

Commentary by Dr. Valentin Fuster

Vehicle State Estimation

2009;():819-826. doi:10.1115/DSCC2009-2531.

In advanced vehicle stability control systems, the availability of online estimates of the vehicle attitude is essential. In four-wheeled vehicles, attitude information is deeply linked to vehicle sideslip angle and sideslip rate, since these variables are strictly related to instability phenomena, safety and handling performances. As direct measurements of such quantities cannot be performed with standard sensors equipment, the design of robust and efficient estimators is needed. In this paper, we tackle the problem by proposing a sideslip rate observer and by demonstrating how an existing sideslip estimation algorithm can be made robust and reliable by online compensation of the sensors bias via a recursive identification approach. The proposed method does not require any additional sensor, making the final solution suitable for industrial applications. Experimental results confirm the effectiveness of the sensors offset compensation and witness satisfactory results of the overall vehicle attitude estimation scheme.

Topics: Sensors , Vehicles
Commentary by Dr. Valentin Fuster
2009;():827-833. doi:10.1115/DSCC2009-2534.

This paper addresses the dynamics identification and servo-controller design for a dual-stage electronic throttle body (ETB) designed for ride-by-wire applications in racing motorcycles. The dynamics of the system is identified and a model of the friction—which is the main nonlinear phenomenon affecting the control of the mechanism—is identified from experimental data. The identified model is used to design a controller composed of a linear part and a nonlinear friction compensator. Experimental tests are employed to validate the controller design and comparison with a linear controller is carried out in order to quantify the advantages brought by the friction compensation.

Commentary by Dr. Valentin Fuster
2009;():835-840. doi:10.1115/DSCC2009-2554.

For vehicle dynamics applications, automotive companies are interested in determining the precise vehicle state in every driving situation in real-time. Part of the vehicle state is the side slip angle—the angle between the vehicle heading and its direction of movement. Currently the side slip angle is not measured in stock cars. To fill the gap this paper presents a basic proof of concept to measure the side slip angle using stock car components for sensing. These include an automotive camera and additional movement information provided in current production passenger cars. A basic computer vision algorithm allows determination of camera movement through the identification of static objects in consecutive camera images. In conjunction with a kinematical model, this data is then used to derive the car’s side slip angle. Finally, the method is evaluated on a real vehicle, with dGPS providing ground truth.

Topics: Automobiles
Commentary by Dr. Valentin Fuster
2009;():841-848. doi:10.1115/DSCC2009-2562.

In this work it is proposed an analysis of the foremost inertial measurements to take into account for the estimate of the lean angle in two-wheeled vehicles. As it is well-known, the roll angle is a crucial variable in the dynamic of a two-wheeled vehicle, since it greatly affects the behaviour of the tire-road contact forces, especially in the lateral direction. Hence, the capability of providing a real time and reliable measure of such quantity allows to evaluate the dynamic properties of the vehicle and its tires and represents the enabling technology for the design of advanced ABS systems and stability control systems. The aim of the analysis proposed in this work is to identify which are the inertial measurements (accelerations and angular rates) that are strongly related to the lean angle and that can be used to estimate it in a urban path. Ideally, the measured lateral acceleration and the vertical angular velocity are the only crucial variables to estimate the tilt angle of a motorcycle, but in a real condition characterized by the presence of slopes, banks and variable speed, also the longitudinal measurement axis becomes relevant.

Commentary by Dr. Valentin Fuster
2009;():849-856. doi:10.1115/DSCC2009-2627.

This paper introduces a wireless piezoelectric tire sensor whose readings can be utilized for the estimation of various tire variables such as slip angle, slip ratio, tire forces and tire road friction coefficient. In this paper, the proposed sensor is demonstrated for the estimation of tire slip angle. Lateral deformation of the tire is decoupled from radial and longitudinal tire deformations using a special sensor design. The decoupled lateral deflection profile of the tire is employed to estimate the slip angle. A new tire test rig is constructed to experimentally evaluate the performance of the developed sensor. Results show that the tire sensor can accurately estimate slip angles up to values of 5.0 degrees.

Topics: Deformation , Sensors , Tires
Commentary by Dr. Valentin Fuster
2009;():857-864. doi:10.1115/DSCC2009-2641.

This paper presents a novel method for identifying in real-time the sprung mass of a 2-DOF quarter-car suspension model. It does so by uniquely combining the base-excitation concept with polynomial chaos estimation. This unique combination of the two methods provides two important benefits. First, the base-excitation concept makes it possible to estimate the sprung mass without explicitly measuring or knowing the terrain profile prior to estimation. Second, the polynomial chaos estimation strategy makes it possible to perform such mass estimation using sprung and unsprung acceleration measurements without pseudo-integration filters that can be difficult to tune. This paper derives the proposed method in detail and presents computer simulations to evaluate its convergence speed and accuracy. The simulation results consistently converge to within 10% of the true mass value typically within 120 seconds.

Commentary by Dr. Valentin Fuster

Engine Control

2009;():865-872. doi:10.1115/DSCC2009-2501.

This paper presents a diesel engine selective catalytic reduction (SCR) control design based on a novel model predictive control (MPC)-assisted approach, which utilizes the advantages of MPC while keeping the computation demand under an acceptable level. The SCR control problem is featured by the challenges of time delay, significant time-varying characteristics, and limited control authority. Based on the understanding of the SCR reactions, the NH3 surface coverage ratio was selected as the control objective. The proposed MPC-assisted method was compared with conventional controllers such as PID and linear MPC (LMPC). Simulation results exhibited that the MPC-assisted approach can achieve a SCR ammonia surface coverage ratio control with much smaller root mean square error compared to these of other controllers while maintaining a manageable computational demand, and in turn better control of tailpipe NOx and ammonia emissions.

Commentary by Dr. Valentin Fuster
2009;():873-880. doi:10.1115/DSCC2009-2506.

This paper explores the possibility of using a cost-effective air-path system that includes a dual-loop (exhaust gas recirculation) EGR and a (variable geometry turbocharger) VGT to achieve independent control of the main in-cylinder charge conditions (i.e. in-cylinder oxygen, inert gas amounts, and gas temperature at the intake valve closing) for HCCI engine combustion transient operation. An engine simulation model consisting of the air-path system and a HCCI combustion model was developed and synthesized to evaluate the control authority of the air-path system on the in-cylinder charge conditions as well as their effects on combustion. A variety of simulations unveiled that such an air-path system can enable independent control of the main in-cylinder charge conditions and active compensation of the effects of the wall temperature variations on HCCI combustion.

Commentary by Dr. Valentin Fuster
2009;():881-887. doi:10.1115/DSCC2009-2519.

Air-to-fuel (A/F) ratio is the mass ratio of air-to-fuel mixture trapped inside a cylinder before combustion begins, and it affects engine emissions, fuel economy, and other performances. Using an A/F ratio and dual-fuel ratio control oriented engine model, a multi-input-multi-output (MIMO) sliding mode control scheme is used to simultaneously control the mass flow rate of both port fuel injection (PFI) and direct injection (DI) systems. The control target is to regulate the A/F ratio at a desired level (e.g., at stoichiometric) and fuel ratio (ratio of PFI fueling vs. total fueling) to a given desired level between zero and one. A MIMO sliding mode controller was designed with guaranteed stability to drive the system A/F and fuel ratios to the desired target under various air flow disturbances. The performance of the sliding mode controller was compared with a baseline multi-loop PID (Proportional-Integral-Derivative) controller through simulations and showed improvements over the baseline controller.

Commentary by Dr. Valentin Fuster
2009;():889-896. doi:10.1115/DSCC2009-2567.

This paper presents an ammonia surface coverage ratio control approach based on the backstepping concept for diesel engine selective catalytic reduction (SCR) systems. SCR models with multiple cells connected in cascade provide more accurate representations of the actual SCR system dynamics by considering the spatial distribution. Control of SCR system ammonia coverage ratio is critically important and effective in terms of ensuring low tailpipe NOx and ammonia emissions. However, such a task is also very challenging primarily due to the nonlinearities of the SCR dynamics and limited ammonia injection control authority. Grounded in the understanding of the SCR nonlinear dynamic characteristics, a backstepping-based nonlinear control law is then proposed to regulate the ammonia surface coverage ratio of the last SCR cell in order to tightly control the tailpipe NOx and ammonia emissions. Lyapunov-based analyses show the stability of the designed control law. FTP75 test cycle simulation results based on a full-vehicle (including engine, chassis, and aftertreatment systems) model illustrated that, compared with a conventional PID controller, the nonlinear backstepping control law can more appropriately handle the SCR system dynamics and exhibits superior ammonia coverage ratio control capability.

Commentary by Dr. Valentin Fuster
2009;():897-903. doi:10.1115/DSCC2009-2704.

Precise control of the air-fuel ratio in a spark ignition (SI) engine is important to minimize emissions. The emission reduction strongly depends on the performance of the air-fuel ratio controller for the SI engine in conjunction with the Three Way Catalytic (TWC) converter. The TWC converter acts as a buffer to any variations occurring in the air-fuel ratio. It stores oxygen during a lean operation and releases the stored oxygen during a rich transient phase. The stored oxygen must be maintained close to the current storage capacity to yield maximum benefits from the TWC converter. Traditionally this is achieved using a simple PI control or a gain-scheduled PI control to address the variability in the operating conditions of the engine. This, however, does not guarantee closed-loop system stability and/or performance. In this work a model-based linear parameter varying (LPV) approach is used to design an H∞ controller. The design goal is to minimize the effect of disturbances on the air-fuel ratio and hence the relative storage level of oxygen in the TWC, over a defined operating range for the SI engine. The design method formulated in terms of Linear Matrix Inequalities (LMIs) leads to a convex optimization problem which can be efficiently solved using existing interior-point optimization algorithms. Simulations performed validate the proposed control design methodology.

Commentary by Dr. Valentin Fuster
2009;():905-912. doi:10.1115/DSCC2009-2706.

In this paper, a framework for isolating unprecedented faults for an EGR valve system is presented. Using normal behavior data generated by a high fidelity engine simulation, the recently introduced Growing Structure Multiple Model System (GSMMS) is used to construct models of normal behavior for an EGR valve system and its various subsystems. Using the GSMMS models as a foundation, anomalous behavior of the entire system is then detected as statistically significant departures of the most recent modeling residuals from the modeling residuals during normal behavior. By reconnecting anomaly detectors to the constituent subsystems, the anomaly can be isolated without the need for prior training using faulty data. Furthermore, faults that were previously encountered (and modeled) are recognized using the same approach as the anomaly detectors.

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