A BIDIRECTIONAL PATH PLANNING ALGORITHM FOR AUTONOMOUS PARALLEL CAR PARKING WITH SMOOTH MOVEMENT CONSIDERATION

Abstract

We consider in this paper the problem of path planning for autonomous parallel car parking in partially observable environments. In this scenario, a car is suppose to have sensors to detect obstacles around within a limited distance. We propose an algorithm based on bidirectional search for computing the path with respect to kinematic constraints to drive the car from a current position to the specified parking lot. We aim at computing a path plan for the car so that the movement is smooth because changing the control profile (turning angle, move direction) too much may result in longer execution time and more energy used. We conduct simulation experiments on different positions of the car. Simulation results show that our algorithm can find smooth itineraries to drive the car to the parking lot and the itineraries are natural as those conducted by human experiences.