Multi-Robot Box-Pushing Using VIKOR Induced Particle Swarm optimization

The paper proposes a unique strategy to formulate the popular multi-robot box-pushing problem in the multi-objective optimization setting. The problem is solved by a novel hybrid particle swarm optimization variant, which we have named as the VIKOR-PSO. This method implements local path planning scheme, and generates an optimal trajectory that minimizes the time and energy requirement of the robots by optimizing the translational and rotational traversal parameters, online through VIKOR-PSO. In contrary to traditional evolutionary algorithms, the proposed approach allows fast convergence of solutions on the Pareto Front, thus making it ideal for solving real-world multi-objective optimization problems, opening a new direction to successful multi-agent coordination.