GUIDANCE BASED MOTION PLANNING FOR AUTONOMOUS NONHOLONOMIC VEHICLE.pdf

Abstract

In this thesis a novel, fast and computationally inexpensive method of generating the autonomous trajectory for nonholonomic vehicles is proposed. The proposed algorithm is developed for autonomous parking of nonholonomic (car-like) vehicles. The proposed method called Modified Trajectory Shaping Guidance is based on missile guidance laws. The advantage of using guidance law is that it is an analytical method that requires very less computation since the trajectory is shaped by using only one parameter (lateral acceleration). The application areas of the proposed method include autonomous parallel parking, diagonal parking, docking of mobile robots and autonomous guided vehicles (AGV). The method is robust and works for different combinations of parking space and vehicle initial position. Generated path satisfies the nonholonomic constraints on the motion of vehicle so that the vehicle can actually follow the path generated by the algorithm