Vector Field Based Formation Control of Multi-Robot System

Multi-robot system as formation can perform better than a single robot in many situations like collective movement of objects, search and rescue operations etc. In this paper a vector field framework is used for achieving path planning for a single robot, and a nonlinear controller is designed for tracking this path. The tracking controller is later extended for achieving formation control of multi-robot system. Asymptotic stability of the controller for any reference path is ensured by using Lyapunov theory. The robots may encounter obstacles in their path during tracking or may collide with each other during the switching of formation patterns. Obstacle avoidance and collision avoidance have been achieved by modified vector field. If the terrain is constrained then it needs to switch from one formation to another according to the terrain requirements. Thus the capability of switching of formation or reconfiguration is an important aspect in the control of multiple mobile robots. If one of the member robots is lost, then the other robots continue with the remaining formation structure and thus ensures robustness against loss of team member. From the extensive simulations using Matlab environment it is seen that any formation of mobile robots is achieved in addition to reconfiguration and obstacle avoidance. The formation control algorithm has been experimentally demonstrated by a series of experiments conducted using LEGO NXT Mindstorm robot kit.