Neural Network Sliding Mode Control of Intelligent Vehicle Longitudinal Dynamics

Abstract

Longitudinal dynamics control is the basis for autonomous driving of intelligent vehicles, which have great significance to the development of intelligent transportation system (ITS). To solve the problems of traditional sliding mode control method when applied to intelligent vehicle longitudinal dynamics, such as large velocity tracking errors, strong chattering phenomenon and so on, a new sliding mode control strategy based on RBF (Radical Basis Function) neural network is presented in this paper. Firstly, a nonlinear mathematical model of the intelligent vehicle longitudinal motion is established by considering the dynamics of the engine, the torque converter, the automatic transmission and the brake system. On the basis of the system model, a variable structure control system with sliding mode is introduced to design a sliding mode variable controller with RBF neural network. This controller can adaptively adjust the switching gain and its stability is proved based on the Lyapunov theory. Finally, the effectiveness of the designed longitudinal velocity control strategy is verified by simulation under typical driving conditions. The simulation results show that the improved control algorithm can effectively suppress chattering, obtain the higher precision and stronger robustness than the traditional sliding mode control. Thus, the longitudinal motion control performance of intelligent vehicles is improved effectively.