Flocking Control Of Multi-agent Systems And Its Applications

Recently, distributed coordination control of networked swarm systems has attractedmore and more attentions of many researchers from various felds. The flocking of multipleagents which is one of the phenomena of swarm, not only can describe and explain most ofthe coordinated behaviors and self-organization phenomena of biological swarms, but alsohas applications in engineering, such as multiple robots, unmanned air vehicles, wirelessmobile sensors networks. Thus, it is signifcation to study the theory an applications ofswarm systems. The main work of this dissertation can be generalized as follows:1. The flocking algorithm with connectivity preserving is considered. Firstly, thestrategy of information feedback from followers to the virtual leader is introduced, whichcan decrease the possible of fragmentation. The theoretical analysis and simulation resultsare provided to demonstrate the validity of the proposed algorithm. Secondly, on the basisof the algebraic graph theory and the Lyapunov theorem, the multiple agents flockingwith external perturbations is investigated. Using velocity consensus combining boundedartifcial potential function, the presented algorithm can drive the group of agents to astable flocking motion under the extern perturbations if the initial network is connected.2. Flocking with static or moving obstacle avoidance in complex environments isdiscussed. Based on the strategy of combining artifcial repulsive function with obstaclecollision prediction, flocking control model which has obstacle avoidance capabilities isbuilt. The stability analysis for the system are given, and the sufcient condition ofobstacle avoidance is obtained.3. On the basis of Razumikhin theorem, the stability of flocking with fxed time delayis studied, and the criteria for the stability is presented. Moreover, the upper bound oftime delay and the control input parameter are obtained. Furthermore, the stabilityfor the switched network with multiple time-varying delays is studied, and the sufcientcondition of the systems stability is obtained through the Lyapunov theorem.4. By means of Kalman-consensus fltering algorithm combining flocking controlstrategy with connectivity preserving, the single target tracking algorithm is discussed inthe mobile sensor networks. Some sensors part of the networks implement the Kalman-consensus fltering algorithm and guide the whole networks to the target with networkconnectivity preserving. Moreover, when the number of targets is more than the sensors,the virtual leader which is made by the Fisher information matrix steer the mobile sen-sors motion in order to get more information. Numerical simulations are presented toillustrate the efectiveness of the proposed results.