PSO-based cooperative control of multiple mobile robots in parameter-tuned formations

This paper addresses the coordination of multiple mobile robots in navigation formations. Difficulties encountered in conventional leader-follower approaches, such as bounds on control commands and formation constraints are revealed. A generic control structure is then proposed, based on the leader-follower strategy and virtual robot tracking framework, to parameterize the formation configuration for cooperatively deploying the robots into desired patterns. In order to achieve both tracking accuracy and control feasibility, the design is further cast as a constrained optimization problem to obtain formation configuration parameters and controller parameters simultaneously. The particle swarm optimization (PSO) algorithm is used, owing to its computationally-efficient capability of handling multi-objective criteria. Satisfactory results obtained through simulation are included to illustrate the effectiveness of the proposed technique for a number of benchmark patterns.