Design And Optimization Of Flocking Control Algorithm

Flocking is the group phenomenon in which a large number of agents influenced each other with a common goal, organized into a collective behavior. In biological, the phenomenon is commonly found, such as fish, birds, deer, etc. Flocking control with limited environmental information and simple rules in multi-agent systems has attracted more and more attentions from various fields of science. With the further development of the science, multi-agent system has been applied in system cooperation control, such as controlling unmanned aircraft group, controlling the autonomy of submarine group, and robot control with no body, etc. Recently, there has been a surge of interests among control theorists in flocking problems, partly due to the wide applications of flocking in many control areas including cooperative control of mobile robots and design of mobile sensor networks. This paper surveys the recent advances in the flocking algorithms, and emphasizes on the design and stability analysis of flocking algorithms. Main work can be summarized as follows:Firstly, the background and significance of the flocking control are introduced, followed by several typical flocking models like Reynolds model, Couzin model, Olfati algorithm and so on.Secondly, based on Reynolds and Couzin models, fuzzy multi-agent flocking control is studied. Though analysis, the system is verified Lagrange stability. The convergence of the model could still be illustrated effect though the simulations, the convergence probability and the relative size are as well.Thirdly, flocking algorithm based on multi-agent with a virtual leader is investigated. Usually the virtual leader with variable speed, the agents couldn't track the speed of the virtual leader. Therefore the speed of tracking optimization algorithm based on Olfati is designed with the acceleration feedback. The redesigned system is Lyapunov stable. A better performance could be shown through the simulations.Finally, flocking algorithm with multiple virtual leaders is analyzed. Traditionally the multi-agent flocking control tracking algorithm often involves one virtual leader. A variable speed with multiple virtual leader of multi-agent flocking control algorithm is presented in this paper. The algorithm can make all the agents tracking virtual leader accurately. All the agents that can track the target to a stable state, as the same purpose of the group leaders; agents could divided into several groups, as the different goal for leaders. The simulation results show that multi-agent system can track the leader as required with the optimization of the Olfati algorithm2.