Research On Anti-disturbance Cooperative Controller Design Of Linear Multi-agent Systems

The research on cooperative control of multi-agent systems has been promoted for several decades,with many results reported on consensus control,formation control and containment control.The majority of these results are derived based on the assumption that the agent is disturbance-free.However,in practical cases,various disturbances inevitably exist in multi-agent systems due to environment interference,measurement noise and load variation.The negligence on disturbances in controller design may cause performance degradation or failure of cooperative control.To achieve high-precision and reliable cooperative control,the dissertation investigates the problem of anti-disturbance cooperative control.Based on a wide comparison study of existing results on anti-disturbance cooperative control,a disturbance observer-based anti-disturbance cooperative control strategy is considered,and some special problems of the anti-disturbance cooperative control design are discussed for linear multi-agent systems.The main results of the dissertation include:1)The anti-disturbance consensus control problem with exosystem disturbances is investigated,and a distributed anti-disturbance consensus algorithm based on relative outputs of neighboring agents is proposed.The algorithm is designed based on the theory of unknown input observer.By defining an error variable with respect to the graph Laplacian,a consensus analysis method is established based on the stability analysis of an augmented system.By using Lyapunov's direct method,the asymptotic consensuability of the closed-loop multi-agent system is rigorously proved.Finally,two simulation examples are provided to illustrate the effectiveness of the algorithm.2)The anti-disturbance consensus control problem with exosystem disturbances and input delays is investigated,and a distributed anti-disturbance consensus algorithm based on relative outputs of neighboring agents is proposed.The algorithm is designed based on the theories of unknown input observer and predictor feedback.By defining an error variable with respect to the graph Laplacian,a consensus analysis method is established based on the stability analysis of an augmented system with input delay.By using Lyapunov-Krasolvskii approach,the asymptotic consensuability of the closed-loop multi-agent system is rigorously proved.Moreover,the related results are successfully extended to the case of leader-follower consensus.Finally,two simulation examples are provided to illustrate the effectiveness of the algorithm.3)The anti-disturbance consensus control problem with relative state-dependent uncertainties is investigated,and a distributed anti-disturbance consensus algorithm based on relative outputs of neighboring agents is proposed.The algorithm is designed based on the theories of special coordinate basis(SCB)and extended high-gain observer.By defining an error variable with respect to the graph Laplacian,a consensus analysis method is established based on the practical stability analysis of an augmented system with input constraint.Based on the results of input saturation and singular perturbation theory,the practical consensuability of the closed-loop multi-agent system is rigorously proved,with the tuning method of the adjustable controller parameters given.Finally,a simulation example is provided to illustrate the effectiveness of the algorithm.4)The anti-disturbance consensus control problem with mismatched uncertainties and heterogenous agents is investigated,and a distributed anti-disturbance consensus algorithm based on relative outputs of neighboring agents is proposed.The algorithm is designed based on the theories of output regulation and generalized extended state observer.By defining an error variable with respect to the graph Laplacian and solving a series of regulator equations,a consensus analysis method is established based on the input-output stability analysis of an augmented system.By using related results on output regulation and input-to-state stability of linear systems,the asymptotic stability of the closedloop regulated outputs is rigorously proved,with a decay rate based cooperative controller design method given.Finally,a simulation example and comparison studies are provided to illustrate the effectiveness of the algorithm.These results can be regarded as extensions of anti-disturbance control,as well as complement materials for existing results on anti-disturbance cooperative control.