Research On Path Planning And Tracking Algorithms Of Automatic Parking System Based On Four-wheel Steering

With the rapid development of the automobile industry,the increasing number of cars,scarce urban parking spaces and crowded parking environment have brought a series of difficulties and problems to parking.Automatic parking assistance system can effectively reduce the burden of drivers and improve parking safety.On the basis of front wheel steering,the four-wheel steering system can reduce the parking space,effectively improve the utilization rate of parking space and the comfort in the parking process.At the same time,it can better cope with the dynamic obstacles in the parking process,so it has better adaptability to the parking environment.In this paper,the path planning and tracking algorithms of four-wheel steering automatic parking system is studied.The parking planning coordinate system is established according to the characteristics of parallel parking and vertical parking.Taking the center point of the rear axle as the reference point,the kinematics model of the four-wheel steering vehicle is derived,and compares the model with the dynamics model provided by Car Sim software at different speed conditions to verify the accuracy of the kinematics model at low speed.The vehicle collision detection model is simplified,and a collision detection method based on a rapid repulsion-straddle experiment is proposed for static obstacles in a given state.Aiming at parallel and vertical parking conditions with static and dynamic obstacles,a Sobol-RRT*Reeds-Shepp parking trajectory planning algorithm combined with numerical optimal control is proposed.Firstly,Sobol sequences with uniform difference characteristics are used to replace the pseudo-random sequences of RRT(Rapidly-exploring Random Tree)algorithm.Meanwhile,the optimization strategy of sampling pool is established to optimize the selection of random sampling points and improve the sampling efficiency of the algorithm.Then,the parking trajectory planning problem is constructed,considering the system dynamic equation constraint based on the four-wheel steering kinematics model and the dynamic obstacle constraint based on the "triangular area method".The parking trajectory planning of the four-wheel steering system is realized by solving the optimal control problem,and the feasibility of the planning algorithm is verified by simulation analysis.In order to realize the tracking control of the planned trajectory,the longitudinal velocity controller based on the Proportional-Integral-Derivative control algorithm and the lateral motion controller based on the model predictive control algorithm are designed respectively,and the effect of the trajectory tracking controller is verified by the co-simulation of MATLAB/Simulink and Car Sim.The trajectory planning algorithm and tracking control strategy proposed in this paper are co-simulated and verified.Using Car Sim to build parallel and vertical parking scenarios,simulation results show that the trajectory planning algorithm proposed in this paper can plan a reasonable parking trajectory,and the parking trajectory tracking controller designed in this paper has good tracking effect and can control vehicle to complete parking operations safely.