To prevent the large costs, both in money and time, of extensive experimental testing of a vehicle, it is increasingly common to develop and validate mathematical vehicle models and use them to analyze performance of the vehicle. Since complex models are also time-consuming to develop and to run, models should be as simple as possible, while still capturing the necessary characteristics for a particular purpose. In many cases, a simple model may be developed for a specific purpose or vehicle characteristic, such as one intended purely for analysis of a vehicle’s longitudinal motion. This paper presents the development of such a longitudinal model of an all-electric vehicle, the 2017 Chevrolet Bolt EV, which was implemented in MATLAB/Simulink.
The model assumes the calibration of battery-motor and takes propulsion/ brake torque as an input and delivers the vehicle speed as the output. The model was verified by testing a longitudinal controller on the model and comparing simulation results to on-road testing results. The model was validated in particular by comparing the time required to reach a specified speed, and was found to be an accurate reflection of the actual vehicle’s performance. Due to the fidelity of this simple model, it is possible to use the model in future work for development of vehicle controls.
Kedar Rege, Ajinkya A. Joshi, Diane Peters