A Pursuit-Evasion Problem of Multiple Pursuers from the Biological-inspired Perspective

The pursuit-evasion problem is common in both nature and the artificial intelligence area. We consider the multi-player pursuit-evasion problem of multiple pursuers with communication constraints on an evader. Besides, the velocity of all pursuers is varying, and the initial positions do not need to...

Full description

Saved in:
Bibliographic Details
Published in2021 40th Chinese Control Conference (CCC) pp. 1596 - 1601
Main Authors Wang, Jianan, Li, Guilu, Liang, Li, Wang, Dandan, Wang, Chunyan
Format Conference Proceeding
LanguageEnglish
Published Technical Committee on Control Theory, Chinese Association of Automation 26.07.2021
Subjects
Artificial intelligence
Biological-Inspired
Biology
Collision avoidance
Fitting
Multiple Pursuers
Newtonian Dynamics
Pattern analysis
Pursuit-Evasion
Online AccessGet full text

Cover

More Information
Summary:The pursuit-evasion problem is common in both nature and the artificial intelligence area. We consider the multi-player pursuit-evasion problem of multiple pursuers with communication constraints on an evader. Besides, the velocity of all pursuers is varying, and the initial positions do not need to be on the circle. The conditions for successful capture, including distance conditions and quantity conditions, are defined in this paper. Furthermore, the capture strategy and escape strategy are designed from the biological-inspired perspective. We realize collision avoidance, gathering, and direction alignment between pursuers by introducing repulsion, attraction, and alignment functions in the pursuer strategy. Then, we observe the sharp phase-transition between the capture zone and escape zone and fit the relevant phase-transition lines by varying the repulsion-coefficient and attraction-coefficient. Based on the coherent analysis for pursuit-evasion behaviour, the relationship between the parameters and the outcome can be obtained.
ISSN:2161-2927
DOI:10.23919/CCC52363.2021.9549448