Fast Consensus Analysis Of Networked Multi-agent System

The consensus protocol of multi-agent system(MAS) is the key point of the performance of MAS. The convergence speed and the packet loss by data transmission are analyzed in MAS. This thesis proposes the fast consensus protocol, optimizing the convergence speed. After that the thesis introduces the influence of data packet loss for consensus protocol and the rescue car experiment platform in MAS.The first part of this thesis studies the consensus convergence speed of multi-agent system(MAS) from two aspects including communication topology and the state of agents. Two-hop network is a novel idea considered in the communication topology. A novel consensus protocol that includes the information of states motions and their integrals is introduced. Combining the two aspects, this thesis proposes the fast consensus protocol. And the protocol has much faster convergence speed by choosing some appropriate weight values. The protocol can be applied to distributed control and large-scale systems. A numerical example is presented to illustrate the effectiveness and superiority of the proposed method.The second part of this thesis considers consensus of multi-agent system with random packet dropout based on Kalman filter. The impact of data packet dropout is a potential source of consensus ability and poor performance of MAS. The Kalman filter is used to estimate current state of agents due to unmeasureble output or state data. The relationship between packet dropout rate and the consensus ability of MAS is obtained. And the critical rate of packet dropout is gained. One step prediction method which can decrease the rate of packet dropout is presented. Numerical examples illustrate the effectiveness of the proposed results.The last part of this thesis introduces the rescue car experiment platform on MAS. In emergency and special environment, using the rescue car experiment platform is very useful. This part mainly introduces the strategy for software program and test results of rescue car experiment platform.