Analysis of Random Walk Models in Swarm Robots for Area Exploration

摘要

The purpose of area exploration is to cover an area effectively and swarm robots are used widely for this type of area exploration because of their robustness, flexibility, and scalability. Meanwhile, due to the limited individual abilities of swarm robots, the random walk methods have been the generally used area-exploration strategy. Although random walk methods possess better performance in area exploration, there still exist some problems. For one thing, little work has been devoted to the theoretical analysis of the random walk models. For another, no effective measures are used to evaluate the searching efficiency of random walk methods and the searching efficiency is verified mainly by simulation experiments. Therefore, in order to make up for the deficiency, this paper presents the mathematical description of the random walk models by drawing on the experience of related researches in biology. The proposed mathematical theory can not only be used to aid our understanding of random walks but also be easy to analyze and control random motion of the robots. Also, the mean squared displacement (MSD) is introduced as the performance measure to evaluate the effectiveness of the random walk methods. In order to proof the effectiveness of the performance measure of MSD, the area-exploration missions of swarm robots are carried out in the simulation experiments and the the coverage rate is used to evaluate the searching efficiency. The experimental results prove that the MSD is an effective performance measure to evaluate the searching efficiency.