Autonomous Lane-change Vertical and Horizontal Control for Vehicle Considering Driver Dissatisfaction
Research on assisted driving functions of lane-change decision-making and trajectory tracking control considering driver dissatisfaction is carried out for autonomous lane-change scenarios on highways.First,the psychological factors of drivers were taken into account in lane-change decision-making,and a lane-change decision-making model based on driver dissatisfaction and minimum safe distance was established.Then,a lateral Linear Quadratic Regulator(LQR)controller was established based on the tracking error model to obtain the optimal feedback front wheel steering angle.Due to the problem that the weight matrices Q and R needed to be debugged repeatedly,a lateral LQR controller based on the Genetic Algorithm(GA)was proposed to obtain the optimal weight matrices Q and R,which in turn improves the robustness of the algorithm.The longitudinal LQR controller was also established to maintain the desired vehicle speed.Autonomous lane-change technology was validated using PreScan,MATLAB/Simulink and CarSim simulation platforms,and the robustness and reliability of the lateral LQR controller were verified under different fixed-speed lane-changing conditions.Under the condition of unsatisfied lane-change safety,the variable desired speed could be followed stably by the longitudinal LQR controller with a absolute value of maximum speed error of 0.69 km/h.Further comparing the lateral control performances of the LQR controller and GA+LQR controller,the lateral and heading tracking accuracies were improved by 66.7%and 27%,respectively,using the GA+LQR controller.The results confirm that the longitudinal LQR controller and the optimized GA+LQR controller have good tracking accuracies under different lane-change conditions.
万方数据
维普
