Anti-Jerk Dynamic Modeling and Parameter Identification of an Electric Vehicle Based on Road Tests

Model-based design facilitates quick development of vehicle controllers early in the development cycle. The goal is to develop simple, accurate, and computationally efficient physics based models that are capable of real-time simulation. We present models that serve the purpose of both plant and anti-jerk control design of electric vehicles (EVs). In this research, we propose a procedure for quick identification of longitudinal dynamic parameters for a high-fidelity plant and control-oriented model of an EV through road tests. Experimental data were gathered on our test vehicle, a Toyota Rav4EV, using an integrated measurement system to collect data from multiple sensors. A MATL4B/SIMULINK nonlinear least square parameter estimator with a trust-reflective algorithm was used to identify the vehicle parameters. The models have been validated against experimental data.