Semi-Autonomous Control of A Multi-Agent Robotic System for Multi-Target Operations

Since multi-targets often occur in most applications, it is required that multi-robots are grouped to work on multi-targets simultaneously. Therefore, this paper proposes a control method for a single-master multi-slave (SMMS) teleoperator to control cooperative mobile multi-robots for a multi-target mission. The major components of the proposed control method are the robot-target pairing method and modified potential field based leader-follower formation. The robot-target paring method is derived from the proven auction algorithm for a single target and is extended for multi-robot multi-target cases, which optimizes effect-based robot-target pairing based on heuristic and sensory data. The multi-robot multi-target pairing method can produce a weighted attack guidance table (WAGT), which contains benefits of different robottarget pairs. The robot-target pairing converges rapidly as is the case for auction algorithms with integer benefits. Besides, as long as optimal robot-target pairs are obtained, a team is split into subteams formed by paired robots regarding types and numbers of the robot-target pairs with the robot-target pairing method. The subteams approach and then capture their own paired targets in the modified potential field based leader-follower formation while avoiding sensed obstacles. Simulation studies illustrate system efficacy with the proposed control method for multi-target operations. Moreover, the paper is concluded with observations of enhanced system performance.