A vehicle collision avoidance control method based on model predictive composite control

In order to reduce the increasing traffic safety problems, the development of vehicle collision early warning and collision avoidance control system is essential. This paper presents a compound control and collision avoidance method based on the longitudinal and lateral dynamics of vehicles to reduce traffic accidents. First, the vehicle inverse longitudinal and lateral dynamics models are established respectively. The longitudinal controller judges whether the vehicle is in a dangerous state and gives collision warning through the safety distance model. Then, the upper model prediction controller and the bottom single neuron PID controller are designed by using the layered control method. Finally, the model predictive controller is designed by combining the parameters constraint at different speeds. Simulation experiments under different working conditions were performed, and results show that the control system designed in this paper can avoid collision successfully and improve the safety, stability and comfort of the vehicle. © 2021, Jilin University Press. All right reserved.