Robust Cooperative Control Of Multiple High-order-integrator Systems

Cooperative control of multi-agent systems has been broadly applied in both military and civil fields.Research on consensus algorithms is usually regarded as the core problem of cooperative control,which can be applied in formation control,cooperative reconnaissance and cooperative search.As a result,extensive attention to consensus algorithms has been received from researchers in different countries and insightful outcomes were proposed,including consensus algorithms of first-order systems and second-order systems and cooperative tracking consensus algorithms.However,there is few work about robust consensus algorithm,whereas robustness plays a critical role in multi-agent systems dealing with the disturbance in the input-output channel as well as the perturbation on the states.In this thesis,ZD method was adopted to design the robust consensus algorithm which will be used for formation control of multi-UAV systems.This thesis includes theoretical part and application part.First of all,consensus algorithm was designed for multiple singleintegrator systems and multiple double-integrator systems using ZD method.Then a general consensus algorithm of multiple high-order integrator systems is proposed with theoretical proof.Next,the robustness characteristic of the proposed algorithm was studied and analyzed.Using convolution theorem,multi-agent systems with different types of disturbance were studied.Simulation were conducted in MATLAB to verify the proposed algorithm with various cases.In the end,the robustness of the proposed algorithm was verified as well and the algorithm can be used for cooperative tracking of moving targets with bounded disturbance.In addition,the proposed algorithm was applied in multi-UAV systems for fixed formation control and time-varying formation control.Fixed-wing UAV and quadrotor UAV mathematical models were set up and linearized through feedback linearization method and small angle linearization method.In terms of fixed-wing UAV,highly nonlinear model was linearized into second-order linear model and the proposed algorithm was adopted in fixed pattern formation of fixed-wing UAV.As for quadrotor UAV,forth-order integrator model was established using small angle linearization method and the proposed algorithm was adopted for time-varying formation while robustly tracking moving targets.For both cases above,simulation was conducted in MATLAB to validate the proposed robust formation control law.The UAVs' control inputs and state variables shown in the simulation results are within the prescribed constraints.Besides,different type of disturbances were injected into the multiUAV systems and simulation results of UAVs' formation flying were shown to validate the effectiveness of the specific formation control law and time-varying formation law as well as the robustness of the consensus algorithm.