Maze path planning of mobile robots by gradient map rendering and gradient follow

Abstract

Path planning for a human being is very easy to reach a desired location in a room, by avoiding obstacles on the way, by generating a mental map and uses this map to find the optimal path. This is a difficult task in case of a robot. In order to make the robot adapt, the system is fed with different obstacle arrangement in the same room and allow the robot to finalise the optimal path avoiding the obstacles. To achieve, the gradient map rendering algorithm is proposed with successful simulation results in MATLAB. The new map produced has given a gradient again using the same algorithm. After the rendering process is completed, the robot climbs up or down the gradient using the maximum or minimum local gradient technique respectively, to find its way to the destination cell. Gradient surface plots are obtained for a variety of mazes to give a visualisation of how the gradient is being formed. Results are obtained after maze simulation successfully shows the most optimised path in any kind of maze.