Mobile Robot Path Planning in 2D Using Network of Equidistant Path

Abstract

This article proposes mobile robot path planning in the presence of static obstacles by arbitrary shapes. The realized path is optimal in a global sense (shortest). The path searching process is divided in three steps: network of equidistant path design, optimal path selection and simulation along the optimal path. The obtained freeways model is small, therefore the solution time is short. During the motion, the robot is absolutely safe from obstacles and never makes sudden rotations. The presented algorithm has been successfully tested on many examples in different situations and difficulties. It is able to solve even maze-type problems. Solution times are comparable with human reactions faced with the same problem. The main contribution of this paper is increasing the efficiently of equidistant path searching and introduction the widest path algorithm.