关键词: 自主泊车, 路径规划, 分段高斯伪谱法, 最优控制

Abstract: A piecewise Gauss pseudospectral method is proposed for autonomous parking route planning in different parking areas in this paper. By establishing a vehicle kinematics model and combining dynamic constraints, end-point constraints and path obstacle avoidance constraints, the autonomous parking route planning problem is transformed into an optimal control problem. In different parking areas, the optimal control problem is discretized into a nonlinear problem by Gauss piecewise pseudospectral method, and the optimal cost function is to minimize the parking time, The interior point method is used to solve the nonlinear problem to obtain the optimal parking path. A simulation model is built in Matlab/Simulink, and five different parking slot lengths are designed. The length of the parking slot is 1.1-1.8 times of the body length. The simulation results show that the proposed method can solve the problem of parking path planning uniformly and effectively. Compared with the traditional pseudospectral method, the proposed method has good performance, faster convergence speed. The vehicle test further verifies that the path information obtained by this method can effectively guide the autonomous parking.

Key words: autonomous parking；path parking；piecewise Gauss pseudospectral method, optimal control