A*-Based Path Planning Algorithm for Swarm Robotics

Currently path planning for a swarm of mobile robots is a relevant problem in the domain of robotics. Various approaches to its solution exist. One of such approaches comprises different methods of informed sampling, which boost the search process through direction of search frontier towards the target. This paper presents combination of such method with the simplified representation of the operational environment, specifically, cellular decomposition. The proposed method ensures transition of a robot swarm into the predefined formation in such manner, that during robot motion along the planned paths no collisions occurred. The experimentation was performed with groups of 5, 10, 15, 20, 25 and 30 robots in three different scenes. Upon experimentation it was revealed, that for a swarm of 30 robots in a complex scene the path computing time does not exceed 7 s.