Research On Path Planning And Tracking Method Of Automatic Parallel Parking

As an important part of the field of intelligent driving research,automatic parking technology has become more prominent as the number of cars continues to grow.It helps to solve the problem of parking difficulties,reduce the driving pressure of the driver,and at the same time provides an important reference for the practical application of automatic driving technology.The automatic parking technology in parallel parking scenarios was studied.Firstly,the target vehicle was simplified in the model,then based on the Ackerman steering principle,in view of the low-speed movement characteristics of the target vehicle in parallel parking conditions,a two-degree-of-freedom kinematics model of the vehicle was established.MATLAB/Simulink was used as software to analyze its motion trajectory and its rationality was verified through Car Sim combined with Simulink,which laid a model foundation for the follow-up path tracking.Secondly,the detection method of automatic parallel parking was studied.According to the position of the obstacle vehicle relative to the target parking space,the parallel parking conditions were divided into three conditions,i.e.I,II and III,and methods for judging different types of parking spaces were proposed.The parking space was calculated,and the detection method of the parking space size was tested through the Arduino smart car.Then,the path planning of the two parking conditions was carried out respectively.In class I condition,the three segment curve of arc-straight-arc was used to design the parking path,and the path was optimized by combining Bezier curve and RBF Neural Network.The path with continuous curvature and maximum value of 0.2284 m^-1 was obtained.In class II condition,the quintic polynomial curve was used for path planning.According to the path constraints,the objective function of path parameters was designed and the Particle Swarm Optimization algorithm was used to solve the problem.The optimal path with continuous curvature and meeting the requirements of obstacle avoidance was obtained.Finally,based on the Pure Pursuit algorithm and Model Predictive Control algorithm,the parking path tracking controller was designed to track the planned path under two parking conditions and the simulation experiment was carried out through MATLAB.The results show that the vehicle can better track the planned path under the action of the two control algorithms,the path tracking error based on Pure Pursuit algorithm is less than 0.0562m and the path tracking error based on Model Predictive Control algorithm is less than 0.0077m,which proves the feasibility of the algorithm used in parking control.At the same time,during the process of path tracking,the vehicle did not collide with surrounding obstacles and the front wheel angle of the vehicle always changed continuously within the allowable range,which proved the safety and traceability of the planned path.