Since that a single controller for intelligent vehicle path tracking is difficult to adapt to complex working conditions and achieve good system control performance, inspired by the concept of control redundancy, an adaptive weight based dual-steering control cooperative game coordinated optimization with fault-tolerant control method for intelligent vehicle path tracking is proposed in this paper. Firstly, the three-degree-of-freedom of lateral, yaw and longitudinal directions is considered to establish the vehicle single-track dynamics model, and two model predictive path tracking steering controllers with different control objective emphases in tracking accuracy and yaw stability are designed. Then, the cooperative game theory is introduced to model the steering interaction of dual controllers, and its Pareto equilibrium optimal solution is solved by the distributed model predictive control framework. According to the influence trend of two game players’ relative participation weights on the path tracking results, an adaptive weight adjustment strategy based on predicted output deviation is designed, to achieve the coordinated optimization of dual steering control, and the steering fault-tolerant control is also realized by cooperative game in some cases of steering partial failure. Carsim/Simulink co-simulation and hardware-in-loop test under the variable curvature path and control failure conditions prove that by the proposed method intelligent vehicle path tracking dual steering control can be coordinated optimized and optimal path tracking performance is obtained, and demonstrate the effectiveness of the adaptive weight adjustment method which has better fault-tolerant control compared to fixed weight allocation under different degrees of steering output failures.

中文翻译：

---

基于具有容错功能的双转向协同博弈的智能车辆路径跟踪协同优化


针对智能车辆路径跟踪的单一控制器难以适应复杂的工况并取得良好的系统控制性能，受控制冗余理念的启发，该文提出一种基于自适应权重的双转向控制协同博弈协同优化算法，用于智能车辆路径跟踪。首先，考虑横向、偏航和纵向三自由度建立车辆单轨动力学模型，设计了两个在跟踪精度和偏航稳定性方面具有不同控制目标侧重的模型预测路径跟踪转向控制器。然后，引入合作博弈论对双控制器的转向交互进行建模，并通过分布式模型预测控制框架求解其帕累托均衡最优解。根据两个博弈者相对参与权重对路径跟踪结果的影响趋势，设计了基于预测输出偏差的自适应权重调整策略，实现双转向控制的协同优化，在部分转向部分失效的情况下，也通过合作博弈实现转向容错控制。变曲率路径和控制失效条件下的 Carsim/Simulink 协同仿真和硬件在环测试证明，所提出的方法能够协调优化智能车辆路径跟踪双转向控制并获得最优路径跟踪性能，并证明了自适应权重调整方法的有效性，该方法在不同程度的转向输出故障下与固定权重分配相比具有更好的容错控制能力。