Research on Path Planning of Automatic Parking System
A non-parallel initial state path planning algorithm for the automatic parking system is presented in this paper. Automatic parking system is a hot research point in intelligent vehicle application fields, and the trajectory generation is one of its key technologies. In the actual process of parking, the initial state of vehicle is always non-parallel to the parking place, so this situation should be researched to serve automatic parking system better. According to the analysis of vehicle kinematics model and the parking process, the trajectory can be considered as the combination of some sections of tangent arc. In this article, the trajectory is composed of several period of circular arc and it is improved by combining with the actual circumstance, then a strong adaptable parking trajectory for the automatic parking system is designed. The simulation analysis is done to verify the geometry trajectory, and the method is used on our intelligent vehicle to check out the feasibility. Compared with traditional trajectory generation method, experimental results show the adaptability and success rates of the method designed in this paper is much more better, which the parking success rates is about 90%.
C. H. Piao et al., "Research on Path Planning of Automatic Parking System", Advanced Materials Research, Vol. 590, pp. 416-420, 2012