Hybrid Stochastic Exploration Using Grey Wolf Optimizer and Coordinated Multi-Robot Exploration Algorithms

Abstract

Multi-robot exploration is a search of uncertainty in restricted space seeking to build a finite map by a group of robots. It has the main task to distribute the search assignments among robots in real time. In this paper, we proposed a stochastic optimization for multi-robot exploration that mimics the coordinated predatory behavior of grey wolves via simulation. Here, the robot movement is computed by the combined deterministic and metaheuristic techniques. It uses the Coordinated Multi-Robot Exploration and GreyWolf Optimizer algorithms as a new method called the hybrid stochastic exploration. Initially, the deterministic cost and utility determine the precedence of adjacent cells around a robot. Then, the stochastic optimization improves the overall solution. It implies that the robots evaluate the environment by the deterministic approach and move on using the metaheuristic algorithm. The proposed hybrid method was implemented on simple and complex maps and compared with the Coordinated Multi-Robot Exploration algorithm. The simulation results show that the stochastic optimization enhances the deterministic approach to completely explore and map out the areas.