Research On Autonomous Valet Parking Path Planning And Tracking Control Method

As cars become more and more popular,there are fewer and fewer parking spaces in the parking lot,which makes it difficult for people to park,and the increase in the number of cars in the parking lot also makes safety accidents frequent.Autonomous parking system is realized by intelligent network technology,which not only enables vehicles to find parking space and park automatically,but also can solve people’s trouble in finding parking space and reduce the rate of traffic accidents in the parking lot.Therefore,it is of great practical significance to research on autonomous parking system.In this thesis,the path planning and tracking control methods of autonomous parking system are studied,and the path planning module based on fusion algorithm and the tracking control module based on MPC algorithm are designed.The main contents are as follows:(1)Establishing the vehicle model and designing the autonomous parking scenario.The vehicle outline is simplified using rectangles,a vehicle kinematic model is established,and the validity of the model is verified using Car Sim/Simulink.By studying the architecture of the autonomous parking system,an autonomous parking scenario is established,and a parking map used in path planning is designed according to national standards.(2)Research on autonomous parking path planning method.For the obstacle avoidance problem in the path planning process,the obstacles are inflated;considering the disadvantages of using both single A* algorithm and RRT* algorithm in path planning,the path planning algorithm of fusion of both is designed,and the results of simulation experiments show that the search efficiency of the fusion algorithm is higher than that of the A* algorithm,and the randomness and local search problems of the RRT* algorithm search are improved;for the global path is not completely suitable for vehicle driving,considering the kinematic characteristics of the vehicle,the path is optimized by using the cubic non-uniform B-spline curve;for the typical vertical parking space,the path planning method based on the hybrid A* algorithm is designed.(3)Research on autonomous parking tracking control method.In order to enable the vehicle to track the path accurately,the model prediction controller is designed.Firstly,a threedegree-of-freedom vehicle dynamics model is established,then the linear prediction equation,objective function and constraints of MPC algorithm are designed based on the dynamics model,and the quadratic programming problem is solved,and finally the double shift line working condition is set to verify the tracking results of the vehicle at low and high speed states,and the results show that the tracking control method in this paper has high tracking accuracy and stability.(4)Path planning and tracking control module simulation verification.Based on the joint simulation platform designed by Car Sim,Simulink and Prescan,the autonomous parking simulation environment designed in this paper is built,and the simulation platform is used to integrate the path planning module and tracking control module,and finally,the effectiveness and accuracy of the path planning and tracking control methods studied in this paper are verified by conducting simulation experiments of the parking space finding and vertical parking conditions.