Research On Modelling And Control Of Swarm Systems

In nature, swarming behavior can be found in many organisms ranging from simple bacteria to more advanced mammals. Examples of swarms include flocks of birds, schools of fish, herds of animals, and colonies of bacteria. Such a collective behavior has certain advantages such as threatening predators and increasing the chance of finding food. Swarming behavior is a key issue in the current investigation of complexity. It has been studied by many biologists, physicists and computer scientists. A challenging problem in complexity is the mathematical modeling of such swarm behaviors and its application to the control of social behavior in artificial world, such as group of robots, traffic flow, team of flights etc.This paper delves into an anisotropic swarm model with an attraction/repulsion function in Euclidean space and studies its aggregation properties. It is shown that the swarm members will converge to a bounded region around weighted center of swarm in a finite time. Simulation further shows the results very well. The coupling matrix in this paper is more general.The method of sliding-mode control based on reaching law is proposed for swarm systems to eliminate chattering, which makes the agents of swarm getting to an expectant trajectory and tracking it. This paper gives the definite control law by using the upper and lower bounds instead of uncertainties. Simulation further shows the effectiveness very well.An algorithm of"soft control"for swarm systems is proposed. The advantage of this method is keeping the local rules of the existing agents in the system by adding a few controlled intelligent agents. It shows that the center of swarm members will be effectively transferred to an expectant position or trajectory by adding H controlled intelligent agents and controlling their positions. The swarm members will converge to a bounded region around expectant position or trajectory in a finite time. This paper gives the controlled law. The results of simulations show this control method has good control effectiveness.