基于转向/横摆纳什博弈的智能车辆路径跟踪协调控制

Intelligent Vehicle Path Tracking Coordinated Control Based on Steering/Yaw Nash Game

针对仅采用转向控制对车辆路径跟踪性能改善有限的问题,提出一种基于转向/横摆纳什博弈的智能车辆路径跟踪协调控制策略。首先,利用容积卡尔曼滤波算法进行轮胎侧向受力实时估计以实现侧偏刚度自修正;基于车辆单轨动力学模型,采用纳什博弈算法系统建模,以描述转向与横摆力矩同时施加于被控车辆时的交互关系,并分别赋予两者以跟踪精度和横向稳定性为各自的控制目标;通过建立预测模型、设定代价函数以及求解纳什均衡等步骤,得到前轮转角和横摆力矩最优控制量。为合理应对不同的工况需求,基于β-ω相平面获取考虑纵向车速与路面附着系数影响的动态稳定包络边界,定义稳定裕度指标划分车辆状态稳定类别,并设计不同的控制模式;控制模式的选择及控制量大小的调节通过控制权重R自适应调整实现,横摆力矩控制则采用单轮差动制动完成。进行联合仿真和硬件在环试验,结果表明,纳什博弈较模型预测控制具有更佳的路径跟踪综合协调控制性能;控制策略中稳定性判别的考虑提升了差动制动的介入精准度;权重R的自适应调整则加强了控制量输出对工况变化的适应能力。

Aiming at the limited improvement of vehicle path tracking performance only with steering control,a coordinated control strategy for intelligent vehicle path tracking based on steering/yaw Nash game is proposed.Firstly,the cubature Kalman filter algorithm is utilized to estimate the tire lateral force in real time to realize the tire cornering stiffness self-correction.Then,based on the 2-DOF vehicle dynamics model,the Nash game theory is adopted for system modelling to describe the interaction of steering and yaw moment simultaneously applied to the vehicle,and the tracking accuracy and lateral stability are given as the separate control objective;by establishing prediction model,setting cost functions and solving Nash equilibrium,the optimal control variables of front wheel angle and yaw moment are obtained.Next,in order to reasonably meet different driving conditions,the vehicle stability envelope boundary considering the influence of longitudinal velocity and road adhesion coefficient is obtained based onb-wphase plane,and the stability margin index is defined to classify the vehicle stability category for designing different control modes.The control modes switching and control variables’values adjustment are realized by the adaptive adjustment of the control weight R,and the yaw moment is completed by differential braking strategy of single wheel.Finally,simulations and hardware in loop experiments are carried out,and the results show that Nash game has better control performance than model predictive control in path tracking coordination,the consideration of stability discrimination in control strategy improves the intervention accuracy of differential braking,and the weight adaptive adjustment enhances the control output’s adaptability to the change of driving conditions.