Mobile Agents-Based Collaborative Perception for Pollution Source Localization in Complex Environments

Abstract

In this letter, we present a multiagent collaborative perception scheme for pollution source localization in complex environments. To ensure reliable connectivity, a crowd-sensing model is designed to adaptively adjust the sensing range of the swarm of mobile agents. While by introducing the residual energy of agents, a modified mayfly algorithm (MA) is adopted to guide mobile agents to move toward the pollutant source, which can give attention to both energy efficiency and localization accuracy. Besides, artificial potential field is utilized to construct the repulsive force field of each mobile agent to avoid collisions in the process of moving toward the pollution source. Experimental simulations validate that the proposed method has better performance in localization accuracy, collision avoidance, and energy efficiency.