Optimal motion control for multi-robot system by multilayer network operator

Abstract

We are addressing the problem of designing an autonomous feedback control for a group of robots moving in the environment with obstacles and avoiding collisions. A robotic team with full information is considered. We propose a two-stage numerical approach for control synthesis. On the first stage a stabilizing control about some point in the state space is produced. On the second stage optimal trajectories are constructed as a set of stabilizing points. We apply a multilayer network operator for numerical search of control functions. Computational experiment is performed for control synthesis for a group of three mobile robots parking in the constrained space.