Hierarchical clustering cooperation flocking based on feedback mechanism

Abstract

The mathematical theory of flocking underlies several omnipresent multi-agent phenomena in biology, ecology, sensor networks, and economics, and also plays an important role in the study of swarm intelligence. Motivated by the results of recent research on flocking estimates, we study a class of hierarchical clustering cooperative models based on feedback mechanism. First, we present hierarchical clustering cooperative models with feedback. By means of new differential inequality techniques and Lyapunov functional method, sufficient criteria on the asymptotic convergence of these models are obtained. Meanwhile, the appropriate initial data that allows collision avoidance between agents are presented. Further, similar results are extended to the case where some pairs of agents do not have direct communication, but the whole system is connected. Moreover, the free-will agents are added to the clustering cooperative models, and unconditional convergence is proved. Combined clustering cooperative flocking with virtual structure, a scenario is simulated with unmanned aerial vehicle (UAV) swarms accomplishing coordinated control with multiple formations. The introduction of the cluster-structure and feedback mechanism makes the flocking models more suitable for practical applications.