The Consensus Of A Complex Dynamic Network Of Multi-agent System

Recently, coordinated control of the multiple agents has attracted a great deal of researchers from different disciplines including physics, computer sciences, biology and control engineering-This is because that it can be widely utilized in these fields. In complex dynamic network of multi-agent systems,"consensus"recalls to reach a same value agreement regarding a certain quantity of interest that depends on the state of all agents. As the basis of the coordinated control problem for multi-agent systems, it is one of the key problems that need to solve in coordinated control of the multiple agents. It has very important theoretic value and practical meaning, which attract a large number of researchers. Now it is the research focus of control community.This paper mainly theoretical research, using algebraic graph theory, matrix analysis, stability theory, the linear system theory and complex network theory. Sufficient conditions are obtained to ensure multi-agent complex dynamic network consistency through the establishment of multi-agent complex dynamic network model (graph model) and linear matrix inequalities (LMI) techniques, application method of multi-agent complex dynamic network consistency sufficient conditions, use of software tools for computer simulation. The main contributions of this dissertation are summarized as follows:The first chapter outlines the multi-agent complex network consistency of application background, research value in theory and application, as well as the development and research status;The second chapter describes the theory of complex network and algebraic graph theory, including network model, the related statistical parameters and its application;The third chapter discusses the consencus of multi-agent and Robust control theory of continuous time system;The fourth chapter is concerned with the reference time varying state of the multi-agent network with delays Robust Consistency problem. By defining a proper control output, the problem is transformed into a robust control, proposes a reference time varying state control protocol. By using the linear matrix inequality method, has been with the reference time varying state of the multi-agent network with delays Robust Consistency convergence sufficient conditions. Simulations are presented to validate effectiveness of the theoretical results. At the end of this paper the research contents are summarized, and pointed out the future research can be further explored in the research direction.