Groups Of Self-organization Behavior And Group Collaboration

This paper considers the swarm behavior of multi-agent systems and analyzes the stability. Firstly, we investigate first-order integrator continuous time systems. Then we consider an double-integrator discrete time system model. This paper have considered three continuous time systems and a discrete time system.The first kind of models consider M-member "individual-based" continuous time systems in an n-dimensional space. Their environment are respectly quadratic profiles, gaussian profiles and multi-model gaussian-type profiles. There is an attractant-repellent profile in every system. In order to consider the swarm behavior, we definite a weighted virtual center of the swarm members. Graph theory and the Lyapunov approach are helpful in the analysis of the swarm behavior. It is proved that each agent of multi-agent systems will converge to our virtual center and steady in the virtual center's limited scope no longer move out of the scope. We present some simulations results for illustrating the theory obtioned in the previous section.This paper considers the consensus of the leader-following systems with a discrete-time model. The velocity of the active leader is unknown in real time. So we design the control laws and observers based on the neighbors to solve the consensus problem. The Lyapunov approach has played an important role in the leader-following systems. It is shown that all agents asymptotically move with the same velocity and position. Moreover, it is also proved that each follower can track the active leader in a noisy-free environment, and the tracking error is estimated in a noisy environment.