Analysis And Design On Multi-Collaboration Of Complex Multi-Agent Networks

Recently, the coordination control of multi-agent network widely used in economy, civilian and military fields has received extensive attention of the researchers from different fields. Among which, the multi-collaborative control has been one of the research hotspots and frontier research issues. And the research of the multi-collaborative control is not only to reveal the inherent law of multi-consensus, multi-flocking, multi-aggregation, multi-synchronization of biological existing in the nature but also to have a better understanding of its complicated system so as to serve for the human social life. The multi-coordination refers to the fact that there is more than one target state for every intelligent being in the nature or manual system after coordination. Based on previous researches, this dissertation deeply studies the multi-consensus and containment control of the dynamic multi-intelligent net in the complex circumstances by using the theory of control theory, stability theory, algebraic graph theory and matrix theory. The main contents are given as follows:In this dissertation, the multi-consensus of multi-agent systems with input delay is studied. The dynamical multi-consensus, the static multi-consensus, and the periodic multi-consensus of the system have been investigated firstly. Different type of multi-consensus can be achieved by means of choosing the control parameters. The second, the multi-consensus of multi-agent network with input delays has been discussed due to the fact that it is inevitable to have delay during the course of information-delivery between intelligent beings in the multi-intelligent internet. The most admissible delay has been found by the using of the theory of Hopf bifurcation. The result reveals that the inner links between the multi-consensus and network topology and control parameters. This dissertation studies a more practical significance problem because of the existence of delays.The multi-consensus for second-order discrete time multi-agent network with position information has been researched in this dissertation. Compared with continuous time systems, multi-agent systems based on sampling control gain an edge on the aspects of control accuracy, control speed, and control performance. The multi-consensus and multi-tracking for second-order discrete time multi-agent systems with position information have been discussed in this dissertation based on the current and the past sampling position information without the availability of velocity information for the second-order discrete time multi-agent systems. By applying the characteristic root analysis, the necessary and sufficient conditions of multi-consensus have been achieved.The multi-consensus of second-order nonlinear multi-agent network has been studied based on intermittent control. In this dissertation, the global exponential stability conditions of nonlinear multi-agent systems have been explored based on periodic intermittent control. In addition, the multi-consensus can be achieved by applying stability of Lyapunov. Finally, the feasibility and validity of these methods are verified through numerical examples. The continuous control method is limited by its high cost. However, our research avoids this disadvantage.The multi-consensus of fractional-order uncertain multi-agent network is considered in this dissertation. The ability of effectively handling the memory and the history independence of the fractional calculus makes it fit into dynamical model fractional-order multi-agent systems. In this dissertation, the multi-consensus problem of uncertain multi-agent systems has been discussed by applying the linear matrix inequality for fractional-order systems. The conclusion of the dissertation is compatible with the first-order multi-agent network. Therefore, the dissertation studies more generally.The containment control problem of fractional-order uncertain multi-agent network has been studied in this dissertation. Different distributed protocols are designed respectively based on state feedback and state observer in the dissertation. The conditions of feedback matrix are also obtained by the theory of stability of fractional-order systems and the linear matrix inequality.Finally, this dissertation summarizes the work and the innovative points of the full text and points out some issues to be examined in the future.