Population‐Based Optimization With Decentralized Method

Abstract

ABSTRACT Multirobot exploration includes the investigation of uncertain, constrained situations to create a finite map with a team of robots. The principal challenge is the real‐time dynamic allocation of exploratory missions among robots. This work introduces an innovative hybrid stochastic optimization technique for multirobot exploration, drawing inspiration from the coordinated predatory behavior of gray wolves. We combine the Reptile Search Algorithm (RSA) and the Coordinated Multirobot Exploration (CME) algorithm to create a mixed stochastic exploration strategy. This method uses both deterministic and metaheuristic approaches. Initially, deterministic cost and utility measures establish the priority of neighboring cells around each robot. Subsequently, stochastic optimization enhances the total solution, allowing robots to assess the surroundings in deterministic ways and navigate utilizing metaheuristic algorithms. The suggested mixed method was tested on both simple and complex maps, and it was compared to the standard CME algorithm, the CME with Whale Optimization Algorithm (CME‐WO), and the CME with Slap Swarm Algorithm (CME‐SSA). Results from simulations show that adding stochastic optimization to the deterministic method makes it much better. This technique makes exploring and mapping the environment faster and more thorough. This research emphasizes the promise of bio‐inspired metaheuristic algorithms in enhancing multirobot exploration systems.