Probabilistic strategies to coordinate multiple robotic pursuers in pursuit-evasion games

Abstract

This paper examines the problem of coordinating multiple robotic pursuers in locating and tracking a non-adversarial mobile evader in a dynamic environment. We have proposed a probabilistic and distributed algorithm in this paper. Our method considers the uncertain state information of the pursuers and the evader using the probabilistic formulation of the pursuit-evasion problem, in order to minimize the expected time of capture rather than guaranteeing capture. Unlike many other approaches to this problem, our method provides a distributed technique to enhance the search efficiency of multiple pursuers as a result of avoiding overlapping search. Finally, we present simulated results to demonstrate the performance of our algorithm in a complex dynamic environment. The results show that our algorithm successfully reduces the expected capture time of the evader with a limited number of pursuers.