Automatic parking path optimization based on Bezier curve fitting

Abstract

The autonomous parking system is an intelligent technology to park a car into a small space. More and more studies have addressed this issue. However, the planned path in these studies is not smooth enough. Based on Ackermann steering geometry, the car kinematic model is established in this paper, the collision possibility between car body and obstacles in the parking space are analyzed, and then the bound of start point and collision-free space are presented. Based on this, the parking trajectory is planned and Bezier curve is used to smooth the trajectory which is generated on multi arcs of circumference, finally, the generation of continuous curvature path is realized. In the end, we build the environment of the simulation in MATLAB and the Fuzzy PID control method is used to track the trajectory which is planned in advance. The simulation results show that the car can not only avoid the obstacles effectively, but also move smoothly on turning points for using the Bezier curve. The design can meet the continuity of the parking requirements very well.