A Constructive Density Function Path Leading to Global Coverage Strategy for A Gaussian Random Field

The coverage of a given environment is an important issue in detection and rescue problems. This article presents a novel Lloyd algorithm for multi-agent coverage of a Gaussian random field characterized by a density function. The core issue is to construct a density function path (DFP) which ensures the distribution function of agents to converge to the one of targets iteratively. By proposing a triggering condition and a series of Gaussian distributionn functions, local optimum of traditional Lloyd algorithm can be avoided. The proposed algorithm make the coverage function reach a global solution with a small enough iteration interval. Simulation result is provided to illustrate the effectiveness of the proposed approach.