HOME JOURNALS CONTACT

Journal of Software Engineering

Year: 2017 | Volume: 11 | Issue: 2 | Page No.: 210-216
DOI: 10.3923/jse.2017.210.216
New Pigeon Flocks Behavior Used for Multiple Robots Autonomous Formation
Shoulin Yin , Hang Li and Jie Liu

Abstract: Background: Generally, robots are developing towards to system application, which has been becoming a hot issue. However, a single robot cannot better accomplish tasks and multiple robots cannot cooperate with each other to accomplish works. How to make robots more better collaboratively complete the task is a problem. Multiple robots autonomous formation has the similarity to pigeon flocks behavior mechanism in biology swarm intelligence algorithms. Materials and Methods: Therefore, this study proposes a multiple robots autonomous formation based on pigeon flocks behavior method in this study. The new method includes three steps. Firstly, it builds pigeon flocks behavior mechanism model by imitating specific level behavior in pigeon flocks. Secondly, topology structure and leadership mechanism in pigeon flocks are modeled by adopting directed graph and artificial potential field theory based on the existed pigeon flock models. Thirdly, after deep analysis on robots autonomous formation, it designs a multiple robots autonomous formation controller based on pigeon flocks behavior method. This controller is designed with pigeon flocks behavior as the core, which contains two auxiliary parts: Control instruction solver and state transformer. Results: Finally, simulation experiments show that multiple robots autonomous formation can form desired formation with this method. Conclusion: Multiple robots autonomous formation based on pigeon flocks behavior is an effective method, which can effectively improve the quality of robots autonomous formation.

Fulltext PDF Fulltext HTML

How to cite this article
Shoulin Yin, Hang Li and Jie Liu, 2017. New Pigeon Flocks Behavior Used for Multiple Robots Autonomous Formation. Journal of Software Engineering, 11: 210-216.

© Science Alert. All Rights Reserved