Planning the Optimum Path for a Mobile Robot using Genetic Algorithm

Abstract

One aspect of interest in robotics in planning the optimum path for a mobile robot or the optimum trajectory for link movements of a stationary robot in order to increase their efficiency. The objective of this paper is to identify the sequence of steps and processes needed for construction off-line path planning system using genetic algorithm (as we coined GPPS). In off-line path planning, the robot is given a map with the location of all obstacles in a given world. The goal is construct the shortest possible path between a pre-defined start and goal positions and then follow this path without running into the obstacles. In addition to the three basic genetic operators, a new operator is proposed here which is coined as repair operator. Repair operator eliminates infeasible path segments and removes path points from nearby obstacles. However, the shortest possible path resulted from applying genetic operators and repair operator may contain overlapping and redundant segments. Hence, to eliminate these drawbacks, a new operator is proposed which is coined as enhancement operator. Eighty experiments are tested on GPPS with different cases. These cases are taken from different perspectives: number and distribution of obstacles, size of obstacles, and number of experiments per a workspace. All experiments with these different cases give, as possible, an acceptable feasible path.