Research of path planning for mobile robot based on improved ant colony optimization algorithm

The paper proposes an improved ant colony optimization algorithm. This method first designs two fuzzy controllers to optimize three parameters α, β, ρ. Then it establishes a dynamic searching window for ants and chaos information are added when near-neighbour city table is constituted in order to increase research speed in initial stages of algorithm. In addition, the concept of active degree of city node is presented as future information to supervise ants to construct solution and update pheromone. Finally a new evaluation criterion is produced to distinguish where paths are excellent or not. So the strategy not only conquers the weakness of easily running into local optimization while making route optimization, but also enhances efficient convergence of ant colony optimization algorithm. Results of large numbers of computer simulations demonstrate that this novel algorithm can plan optimal path rapidly in intricate three dimension (3-D) environment.

I. Introduction

Path planning problem of mobile robot is an important content for study area of mobile robot^[1]. It indicates that mobile robots search an optimal or secondary optimal and collision- free route in terms of some evaluation functions (such as shortest traveling path or best time etc.) in obstacle environment t^[2]. Ant colony optimization (ACO) put forward by M Dorigo in 1991 is a new intelligent optimization algorithm; it has advantages of strong robustness, preferable global optimization performance and easy to blend with other algorithms. Its biologic mechanism is that ant colony seeks a shortest and feasible path between formicary and food fountainhead, so it is fitly consistent with that of path planning of mobile robot. Inartificial relations of interior mechanism between them provide powerful gist for research of path planning based on ant colony optimization algorithm.