Qualitative analysis of using particle swarm optimization for multi robot agents in three dimensional space

Abstract

In the field of multi robot systems, algorithms that control communication and movement of multi robot agents has became an interesting arena for researchers recently. A big challenge in this area is to design an effective algorithm which make multi robots to work as a team of robots to perform their task and reach to their goal. In this article we use a Modified version of Particle Swarm Optimization Algorithm that is called MPSA. This algorithm allow us to use a virtual multi robot search to find optima in a three dimensional function space. The presented model has the advantages of being capable to change parameters and number of robots or agents, in order to improve the functionality of the multi agent system. In order to avoid collision with obstacles, we use the "leader follower" technique which can help to change the direction of swarm movement to avoid collision with obstacles while trying to get closer to their target. Simulation results show that with this algorithm, our team of robots can perform a swarm movement to reach the target while avoiding collision among themselves or with the obstacles that may be in the environment.