关键词: 智能汽车, 横向跟踪, ＬＱＲ控制, 蚁狮算法

Abstract: In order to solve the problem of poor vehicle tracking accuracy and poor stability caused by poor adaptability of a linear quadratic regulator (LQR) to the large curvature reference path under the traditional fixed weight coefficient, this paper designs an adaptive weight coefficient LQR controller with preview feedforward angle compensation to track the path laterally based on ant lion optimization (ALO) algorithm. Firstly, a classical LQR controller is designed based on the lateral tracking error model of two-degree-of-freedom vehicle dynamics. Secondly, the preview feedforward control is used to eliminate the steady state error caused by error model simplification. Then, an adaptive LQR weight coefficient correction strategy based on ALO is proposed, which takes lateral distance deviation, heading angle deviation and output front wheel angle as the evaluation functions. Finally, through the real vehicle test, the control effect of the controller in the real vehicle environment is verified. The results show that the designed controller can adapt to the large curvature reference path, take into account of the path tracking accuracy and driving stability, and perform well in robustness at different vehicle speeds.