Multi-robot self-organizing cooperative pursuit method based on probabilistic graphical model
HUANG Yi-xin
XIANG Xiao-jia
ZHOU Han
YAN Chao
CHANG Yuan
SUN Yi-hao
Abstract：Multi-robot cooperative pursuit is a typical application of collective intelligence in adversarial environments.In real environments where perception is limited,environmental structure is unknown,and the target status is uncertain,multi-robot cooperative pursuit faces many challenges such as the environmental adaptability and the task scalability.To address this problem,a self-organized cooperative pursuit method based on the probabilistic graphical models is proposed.First,the kinematic models of the pursuit robots and target are established,and the mathematical description of the pursuit problem is given.On this basis,a scalable cooperative pursuit"perception-decision"probabilistic graphical model struc-ture is constructed,and a probability distribution parameter estimation method is designed for the states of each node in the model.Then,inspired by hunting behaviors of wolves,a staged pursuit strategy is designed to improve the capture effi-ciency.Finally,the numerical simulation and software-in-the-loop experiments are conducted to verify the node scalability,environmental adaptability,system risk resistance,and model transferability of the proposed method.
Keywords：probabilistic graphical modelsself-organizationcooperative pursuitmulti-robotunknown environments
Publication Date：2023-12-28
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：11( 2225-2235 )