Research on Lane Changing Control of Intelligent Vehicle based on Model Predictive Control 2019-01-0879
Aiming at the complexity, randomness, optimization problems with multiple objectives and constraints of lane changing, a path tracking controller based on model predictive control is designed in this paper. The virtual highway scene that accords with reality of China is built in PreScan. A three degree of freedom vehicle dynamics model with small angle assumption is established, and this model is used as the predictive model of the controller. Based on the relationship among the surrounding vehicles and the host vehicle, the minimum safety distance model is established. Fifth-degree polynomial is used for lane changing trajectory planning through analyzing and comparing the advantages and disadvantages of common lane changing trajectories. A reasonable objective function and constraints are designed, and they are transformed into quadratic programming problem for optimization. Finally, the designed controller algorithm is verified through the co-simulation of PreScan, CarSim and Simulink. The simulation results show that the host vehicle with different speed and attachment conditions shows great stability and effectiveness. In addition, the effects of two important parameters, prediction horizon and control horizon of the controller, are analyzed.
Liuhui Wang, Zhenfei Zhan, Yangxin Sun, Ling Zheng, Yonggang Liu, Gang Guo, Fan Ming
Chongqing University, GAC Research Institude