基于模型预测轮廓控制的自动驾驶车辆路径跟踪控制

Automatic vehicle path tracking control based on model predictive contouring control

为了实现无人驾驶汽车的路径跟踪和避障功能,基于模型预测轮廓控制提出一种路径跟踪控制策略。首先将二自由度单轨车辆模型和Pacejka轮胎模型结合在一起搭建整车模型,遵循模型预测的思路将路径规划和跟踪问题结合为一个非线性优化问题,利用线性化得到线性时变模型,在每个采样时间内以凸二次规划的形式建立非线性优化问题的局部逼近。其次,为了使车辆在跟踪路径的时候维持稳定,结合车辆稳定包络线,施加横向稳定性约束,并通过施加车道线边界约束,实现自动驾驶车辆的避障功能。最后将Matlab/Simulink和Carsim结合进行联合仿真。仿真结果表明,所设计的控制器在较宽路面能自主规划出一条最短路径,实现了跟踪和避障功能,同时维持了车辆在行驶中的稳定性和安全性,在赛道竞速下具有较明显的优势。研究结果可为结构化道路下的汽车自动驾驶提供技术参考,具有一定的应用价值。

In order to realize the function of tracking and obstacle avoidance of unmanned vehicles,a path tracking control strategy based on model predictive contour control was proposed.Firstly,the 2-DOF single vehicle model and the Pacejka tire model were combined to build the vehicle model,the path planning and tracking problems were combined into a nonlinear optimization problem by following the idea of model prediction,and the linear time-varying model was obtained by linearization.The local approximation of the nonlinear optimization problem was established in the form of convex quadratic programming in each sampling time.Then,in order to maintain the stability of the vehicle when tracking the path,the lateral stability constraint was applied in combination with the vehicle stability envelope.By applying lane boundary constraint,the obstacle avoidance function of the autonomous driving vehicle was realized.Finally,Matlab/Simulink and CarSim were combined to conduct co-simulation.The simulation results show that the designed controller can independently plan a shortest path on a wide road surface,realize the function of tracking and obstacle avoidance,and maintain the stability and safety of the vehicle in the running,which has a good advantage in the race track.The results provide technical reference for vehicle autonomous driving on structured roads and has a certain application value.