To address the inadequacy of the linear quadratic regulator (LQR) algorithm for conventional passenger vehicles, which fails to fully consider the entire trailer system and leads to tracking offsets in the full trailer, an adaptive LQR control method for the off-axle full-trailer system is proposed to achieve accurate trajectory tracking. First, the kinematic model of the full-trailer vehicle and the error model in the road coordinate system are established and linearized. Next, the variation in error is accounted for, and an adaptive function that conforms to the full-trailer system is designed to adjust the parameters of the weight matrix adaptively. Finally, the parameters are continuously updated based on the cost function and adaptive index function iteration to calculate the control variable, ensuring that the system can effectively track the path. Experimental results demonstrate that this method significantly improves tracking performance during tracking tasks. The system improved tracking error accuracy by 25.04% for straight lines and 67.38% for circles, compared to the conventional LQR.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

An Off-Axle Full-Trailer Vehicle Control Method Based on Adaptive LQR

  • Guochen Niu,
  • Hui Xia,
  • Dandan Hu

摘要

To address the inadequacy of the linear quadratic regulator (LQR) algorithm for conventional passenger vehicles, which fails to fully consider the entire trailer system and leads to tracking offsets in the full trailer, an adaptive LQR control method for the off-axle full-trailer system is proposed to achieve accurate trajectory tracking. First, the kinematic model of the full-trailer vehicle and the error model in the road coordinate system are established and linearized. Next, the variation in error is accounted for, and an adaptive function that conforms to the full-trailer system is designed to adjust the parameters of the weight matrix adaptively. Finally, the parameters are continuously updated based on the cost function and adaptive index function iteration to calculate the control variable, ensuring that the system can effectively track the path. Experimental results demonstrate that this method significantly improves tracking performance during tracking tasks. The system improved tracking error accuracy by 25.04% for straight lines and 67.38% for circles, compared to the conventional LQR.