Abstract: The problem of path following is one of the research hotspots of intelligent connected vehicle, and the quality of design for path following controller has a direct impact on the reliability of intelligent connected vehicle. Comprehensively considering the vehicles path following accuracy and lateral stability, the 2-DOF vehicle handling dynamics model and path following error model were established. The desired yaw rate was determined by inverse method, and the path following controller was designed by using quadratic optimal control method. The parameters c1, c2 and k in the expression for desired yaw rate were optimized by using QPSO. Taking lane-changing behavior and curved road driving as examples, the control effect of the optimized controller parameters and the robustness of the controller under different vehicle speeds and different path curvature frequencies were simulated. The results show that after optimizing the parameters c1, c2 and k, the controller improves the control effect of vehicle centroid sideslip angle and yaw rate. When the vehicle speed and the frequency of path curvature change, the controller has obvious robustness. This study can further improve the vehicles path following accuracy and lateral stability, which provides a certain reference for the research of path following control of intelligent connected vehicle.

Key words: vehicle engineering; intelligent connected vehicle; path following; QPSO algorithm; optimum control; yaw rate; centroid sideslip angle

摘要： 路径跟踪问题是智能网联汽车的研究热点之一，路径跟踪控制器设计的好坏直接影响智能网联汽车跟踪行驶时的可靠性。综合考虑车辆的路径跟踪精度和横向稳定性，建立两自由度汽车操纵动力学模型和路径跟踪位姿误差模型。通过反推法确定期望横摆角速度表达式，运用二次型最优控制方法设计路径跟踪控制器。采用量子粒子群算法(quantum-behaved particle swarm optimization, QPSO)，对期望横摆角速度表达式中的参数c1、c2和k进行优化。以车辆换道和弯曲道路行驶为例，仿真研究控制器参数优化后的控制效果，以及不同车速和不同路径曲率的频率下控制器的鲁棒性。结果表明：参数c1、c2和k优化后，控制器改善了车辆质心侧偏角和横摆角速度的控制效果；当车速和路径曲率的频率变化时，控制器具有较明显的鲁棒性。能够进一步改善车辆的路径跟踪精度和横向稳定性，为智能网联汽车的路径跟踪控制研究提供一定的参考。

关键词: 车辆工程；智能网联汽车；路径跟踪；QPSO算法；最优控制；横摆角速度；质心侧偏角