Numerical Potential Field and Ant Colony Optimization Based Path Planning in Dynamic Environment

Abstract

A new path planning method is developed in this paper for mobile robot navigation. A novel method called numerical potential field is employed to model the environment the robot resides in. The locomotion of the obstacles is taken into considerations by varying the local potential values. The concepts of reachable place and reachable field are defined. Ant colony optimization method is then applied to perform the path search. The trap position problem and its solution is also discussed. Experiments showed the effectiveness of the algorithm.