Path Planning of Mobile Robot Using Traveling Salesman Problem Modeled for Genetic Algorithm

Abstract

One of the major requirement of automation is to reduce or eliminate human hands, even in extreme situations like rescue of humans from disaster where robots are employed in large numbers. The robots in such instances need to be proficient in identifying the targets and tiding through hazardous environment without collection. Literature has several techniques of path planning of a robot to reach the destination without collision. In this paper, path planning of a robot is attempted using genetic algorithm which is modelled as a traveling salesman problem (TSP). The basics of the TSP is that when the robot deviates from the designed path, the TSP algorithm is executed for several iterations to reach the destination optimally. The algorithm is tested on a robot hardware called iRobot (ICreate) which is interfaced with ARM Cortex M3 controller. The iRobot was able to track the destination successfully by avoiding obstacles and by making rotation of 90°. Improvement are being worked out to implement more optimal rotations.