Abstract: Aiming at the lateral control of autonomous vehicle, a preview control based on jerk signal is proposed. The error model of 2-DOF vehicle dynamic is established, and the closed-loop lateral control model is composed by linear quadratic regulator (LQR) feedback control and feedforward control. Use CarSim and Simulink to simulate the vehicle entering and tracking a steady state circle. By analyzing the proxy such as lateral position error and yaw angle error, it is shown that the preview control has good results in accuracy of lateral control. Similarly, the tracking speed has also been significantly improved.

Key words: Autonomous vehicle; Preview control; Path tracking; Lateral control; 2-DOF vehicle dynamics model

摘要： 针对自动驾驶车辆的横向控制，提出了一种基于跃度信号的预瞄控制。建立车辆二自 由度动力学误差模型，采用线性二次调节（LQR）反馈控制和前馈控制构成闭环横向控制模 型。以 CarSim 和 Simulink 为仿真平台，模拟车辆进入并跟踪稳态圆。通过分析横向位置误 差和横摆角误差等指标，表明该预瞄算法在横向控制精度、跟踪速度方面具有良好效果。

关键词: 自动驾驶汽车；预瞄控制；路径跟踪；横向控制；二自由度车辆模型