Sai S. V. Rajagopalan

Control-oriented model for an automotive PEM fuel cell system with imbedded 1+1D membrane water transport

Fuel-cell driven vehicles have become the focal point of research and development in all automotive and academic research institutions. The proton exchange membrane fuel cell system which seems to be the forerunner for traction applications, poses a significant challenge for modelling and control design due to a highly nonlinear behavior attributed to the complex interaction of the fluid, thermal, electro-chemical and mechanical mechanisms. Efficient control of the system is required to meet drivability requirements and maintain stack health. In this paper, a nonlinear model exhibiting the important dynamics and behavior of such a system is presented. The fuel cell model is based on static maps obtained from a higher order 1&1D model. Compressible fluid flow equations, conservation of mass and energy principles and data from experiments conducted on a test bench are used to model the fuel cell system.

A high fidelity starter model for engine start simulations

Engine start is a very crucial phase in the operation of automotive engines, and the starter motor plays a vital role in this short transient period. However, a complete and exhaustive model of the combined engine-starter system has not appeared to date in the open literature. Although many researchers have modeled the starter system in engine models, the role of the roller one-way clutch is typically omitted. The roller one-way clutch not only protects the starter motor from damage due to high (transient) engine RPM, but it also enables the starter motor to deliver only positive torque to the engine crankshaft. As the trend for model-based calibration grows, the use of computer simulation in engine cold start and crank-to-run transition problems demand high-fidelity simulations and modeling, which in turn requires the inclusion of an accurately modeled oneway clutch mechanism. In this paper, a novel roller one-way clutch model is developed based on the principles of the oneway clutch. Simulation results are presented for a cylinder pressure resolved engine model and results for starter current characteristics and engine RPM, are verified against actual engine data obtained during the starting phase.

Multi-variable control for an automotive traction PEM fuel cell system

This paper describes an approach for the control of pressurized PEM fuel cell systems used in automotive traction applications. This model-based controller design approach is based on a 13-states non-linear dynamic model. The focus of the paper is in controlling the excess air ratio while tracking an optimum variable pressurization for maximum system efficiency during load transients. The control approach combines a feed-forward approach based on the steady state plant inverse response, coupled to a multi-variable LQR feedback controller. The controller shows excellent performance over severe load transients with both actual states and state observers feedback

Application of an Exhaust Geometry Based Delay Prediction Model to Internal Combustion Engines

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.Copyright

Control oriented modeling of a three way catalyst coupled with oxygen sensors

Modeling of three-way catalyst behavior in stoichiometric engines is a tovpic with significant depth of research which encompasses complex kinetics based models through highly simplified control-oriented models. For model based control design, one must consider the behavior of the catalyst in conjunction with the feedback oxygen sensors. These sensors have well known influences from exhaust gas species due to interaction with the catalyst which, if ignored, can cause significant difficulties in modeling and control. These effects have often been addressed by calibrating and validating catalyst models under simplified conditions in order to minimize errors. In this work, the root cause of many of these errors is investigated and experimental evidence presented. Additionally, ARMA and Hammerstein models are used to find a model capable of predicting the post-catalyst oxygen sensor response over realistic validation data.Copyright

System Identification for Air/Fuel Ratio Modeling Using Switching Sensors

Modeling the internal combustion engine for air-to-fuel ratio (AFR) control has been widely studied and several methodologies have been adopted toward the end goal of applying model based control schemes. In this paper, an online binary sensor identification (BID) technique using switching sensors is adopted for modeling the response from fuel input to AFR output of a spark-ignited, internal combustion engine, to be used in AFR control. In general terms, the algorithm identifies the impulse response of a linear time invariant (LTI) system by choosing an optimal sequence of inputs. The entire modeling process is done online with a four-cylinder engine in a test cell, using typical production switching sensors. Finite impulse response (FIR) linear time invariant (LTI) models are identified at prescribed operating points of the engine (specified by engine speed and the manifold air pressure). The validity of the resulting model is then tested on separate data streams with AFR measured from a wide-range sensor output. By scheduling the coefficients of the FIR models based on the operating condition, it is possible to identify a linear parameter varying AFR model for the appropriate operating regions of the engine.Copyright