Dynamic Path Planning for Robot Navigation Using Sonor Mapping and Neural Networks

Abstract

This paper presents a new path planning algorithm for safe navigation of a mobile robot in dynamic as well as static environments. The certainty grid concept is adopted to represent the robot’s surroundings and a simple sensor model is developed for fast acquisition of environmental information. The proposed system integrates global and local path planning and has been implemented in a partially known structured environment without loss of generality for an indoor mobile robot. The global planner finds the initial path based on Dijkstra’s algorithm, while the local planning scheme uses three neural networks to follow the initial global path and avoid colliding with static and moving obstacles. Effectiveness of these algorithms is illustrated through simulation and experiment using a real robot. The results show that the proposed algorithm can be efficiently implemented in a time varying environment.