Research For Multi-agent Control Based On Robust Control And Output Regulation Theory

As a hot issue, the coordination control problem of multi-agent systems has attracted considerable attention for the domestic and international researchers recently. As a novel interdisciplinary subject which develop from the artificial intelligence, its significance lies in its ability to resolve the problems which beyond the single-agent system of large, complex practical problems, and is widely used in many fields, such as nature, engineering and technology. It is helpful to the human understand and explore the nature. Therefore, the study on the multi-agent systems has profound meaning both in theory and in practical engineering. In this thesis, on the basis of summary of the corresponding research about the coordinate control of multi-agent systems, we investigate the consensus control and formation control problem for the nonlinear multi-agent systems. The main results of this thesis can be summarized as follows:1Leader-following consensus of second-order multi-agent systems with time-delay and nonlinear dynamicsWe consider leader-following consensus problem for second-order multi-agent system with nonlinear dynamic. At first, when time-delay is zero, for the topologies of the multi-agent systems are both fixed and switched, based on the knowledge of the graph theory, we obtain the sufficient condition which render the followers can track the leader. Then considering the time-delay, a neighbor-based distribute controller is given for each agent. Based on Lyapunov-Razumikhin stability theorem, get some sufficient conditions which make the system reach consensus.2Robust H∞formation control for nonlinear multi-agent system with time-delayThe robust H∞formation control problem for second multi-agent system with nonlinear dynamic and time-delay is addressed. Firstly assuming there exist no external disturbance and uncertainty, for the topologies of the multi-agent systems are both fixed and switched, applying linear matrix inequality method, two sufficient conditions obtained, respectively, which can solve the formation control problem. Moreover, when there exist the external disturbance and uncertainty in the multi-agent systems. Considering the robust H∞, formation control for the multi-agent system, addressing the sufficient conditions which solve the problem i.e. using the designed control scheme, the multi-agent system not only form the required shape, but also satisfy the H∞, performance.3Output regulation distributed formation control for affine nonlinear multi-agent systemsDistributed formation control problem for multi-agent systems with affine nonlinear dynamic is considered. In the multi-agent system, the active leaders or the environmental disturbance can be seen as the exo-system in the output regulation problem. In order to solve the problem, we generalize the output regulation problem of nonlinear systems to multi-agent systems and design a distributed control scheme. Based on the internal model principle, the distributed output regulation formation control problem can be solved by solving the regulator equations.4Formation control for nonlinear multi-agent systems based on robust output regulationWe study with the robust output regulation distributed formation control for nonlinear multi-agent systems with uncertainty parameters. As not each agent can obtain the external signal, we design the distributed dynamic feedback control law based on the virtual regulator error. Applying the center flow theorem and internal model principle, we get the condition for multi-agent system output regulation formation control is solved.The main innovative points of this thesis are as follows:1) For fixed and switched topologies, we consider the leader-following consensus problem for second-order multi-agent systems with nonlinear dynamic. In both cases with and without time-delay, by graph theory and stability theory, the condition which the followers can track the leader is given.2) The formation control for second-order multi-agent systems with nonlinear dynamic is studied. When considering the disturbance and uncertainty of the multi-agent system, using the linear matrix inequality method, some delay-dependent sufficient conditions of formation control for the multi-agent systems are obtained, which guarantee that the considered multi-agent systems can achieve desired formation and satisfy the H∞performance index.3) We consider the formation control problem for general nonlinear multi-agent systems using the output regulation theory. Assume that the active leader or the environment disturbance as the external system, we generalize the traditional nonlinear output regulation theory to the multi-agent systems. As the exo-system can not be measurable completely for other agents, a distributed formation control scheme is designed. Based on the internal model principal, the sufficient and necessary condition is obtained, which guaranteed the formation problem can be solved. And when the uncertainty parameter is concerned, we studied the robust formation control problem of the multi-agent systems.