Dual-Mode Adaptive Cruise Control Strategy Based on Model Predictive Control and Neural Network for Pure Electric Vehicles

Abstract

The Adaptive Cruise Control (ACC) system is an important advanced driver assistance system. In order to improve the vehicle-following effect of ACC for pure electric vehicles to adapt to the complicated and changing traffic environment, reduce the cruise energy consumption and extend driving range, an ACC strategy based on Model Predictive Control (MPC) is developed for vehicle-following condition; an ACC strategy based on Neural Network (NN) using the driver behavior data is developed for the lead vehicle cutting-in/cutting-out condition; a dual-mode ACC strategy based on MPC and NN is proposed to realize intelligent ACC. The proposed ACC system for pure electric vehicles uses three-level control structure. The upper level adopts the above dual-mode ACC strategy. During braking, the middle level coordinates the regenerative braking and the hydraulic braking of the pure electric vehicle. The lower level controls the electric motor and the hydraulic braking device respectively. This three-level control structure is established in Matlab/Simulink software, the vehicle longitudinal dynamic model is established through Carsim software, and the traffic scenario is established through PreScan software, and then these three parts are co-simulated. The simulation results show that this proposed ACC strategy of the pure electric vehicle has better cruise control effect and brake energy recovery capability, and achieves the better safety, riding comfort and energy efficiency.