A novel multi-robot exploration approach based on Particle Swarm Optimization algorithms

Abstract

This paper proposes two extensions of Particle Swarm Optimization (PSO) and Darwinian Particle Swarm Optimization (DPSO), respectively named as RPSO (Robotic PSO) and RDPSO (Robotic DPSO), so as to adapt these promising biological-inspired techniques to the domain of multi-robot systems, by taking into account obstacle avoidance. These novel algorithms are demonstrated for groups of simulated robots performing a distributed exploration task. The concepts of social exclusion and social inclusion are used in the RDPSO algorithm as a “punish-reward” mechanism enhancing the ability to escape from local optima. Experimental results obtained in a simulated environment show that biological and sociological inspiration can be useful to meet the challenges of robotic applications that can be described as optimization problems (e.g. search and rescue).