Flocking Control Of Multiple Agents Based On Artificial Potential Field

Recently, the development of science and technology, the problem of flocking control of multiple mobile agents has attracted a lot of attention among researchers in the control field. Flocking is a kind of behavior in the process of biological evolution, it could be seen from a series of phenomenon, such as the migratory of birds, foraging of fish school, avoiding predators of deer etc. In the practical research, researchers in various fields abstracted the biological population as multiple mobile agents system. The wide attention and research on flocking behavior of multi-agent will get a good applications in many fields, such as network technology, social behaviors, physics, formation control of unmanned aerial vehicle and Intelligent robot control etc.This paper have deeply insight of the methodology and current research status of multi-agent flocking control problem, which agents have the second order nonlinear dynamics. Consider the perspective of individual artificial potential, a new artificial potential function related to agent location information is proposed. Then, using hysteresis in adding new links and applying new potential function method, a flocking control algorithm with the rules of maintain network connectivity is proposed.The algorithm guarantee fragmentation of the network can be avoided, under which all agents approach to a common velocity vector, asymptotically converge to a fixed value of inter-agent distances and collisions between agents can be avoided throughout the motion. The rationality of the algorithm is checked by LaSalle invariance principle and algebraic graph theory. Finally, using the Matlab platform to illustrate the rationality of this algorithm.Furthermore, this paper extend the flocking algorithm to solve the flocking situation of the group with a virtual leader agent. Combined with the new artificial potential function, This paper proposed flocking control algorithm with the rules of avoid collision, the algorithm can make all agents approach to a common velocity vector, asymptotically converge to the stabilization of inter-agent distances and collisions can be avoided between agents in the motion evolution. Using LaSalle invariance principle and algebraic graph theory to prove the algorithm rationality. Finally, using the Matlab platform to illustrate the rationality of this algorithm.