Robots path learning using Neuro-evolution when changing formation in cooperative transport with multiple robots

Abstract

The cooperative transportation using multiple robots has been expected to be used in factories and construction sites. Cooperative transportation, which using multiple robots, can support various situations compared to transportation using one robot. In terms of form closure, more than four robots are required on two-dimensional plane to constrain the baggage geometrically. However, it is desirable to use as few robots as possible due to cost and other issues. In this paper, a cooperative transportation system using three robots for formation change is proposed. Because of the robots’s path when the formation change is complex, neuro-evolution is used to learn the robots’s path. Finally, the effectiveness of proposed system is experimented by using simulation.