Research On Cooperative Formation Control Algorithm For Multi-uav Based On Data Driven

In recent years,due to its wide application value,quadrotor UAV formation has become an important research direction in the field of control.The quadrotor dynamic model is nonlinear and highly coupled.In addition,there are many uncertainties in the dynamic model obtained in practical applications.Traditional model-based control methods cannot achieve the desired flight performance.Therefore,the quadrotor formation control problem with uncertainties and nonlinear coupled dynamics remains to be resolved.This paper mainly considers the perturbation of the internal parameters of the UAV and the external uncertainty perturbation.Using a data-driven method,combined with the theory of distributed consistency,the cooperative control of the quadrotor UAV formation is studied.Corresponding research on UAV formation is mainly conducted in three aspects.The research mainly includes the research of the attitude control algorithm of the single UAV based on data-driven;the research of the optimal formation control algorithm and the research of the robust formation control algorithm.The specific content is as follows:First of all,this paper studies a single UAV nonlinear attitude control strategy.When obtaining the attitude controller,this paper proposes a data-driven iterative method that does not require system model parameters based on the traditional modelbased strategy iterative algorithm to solve the HJB equation.This method collects UAV states data online,uses the idea of polynomial approximation,and obtains the optimal controller by means of least squares.And this article gives a rigorous proof of stability and convergence.Secondly,based on the virtual leader-follower model,the data-driven optimal coordinated control algorithm of UAV formation is studied.The system is decoupled into two fourth-order linear subsystems and two second-order linear subsystems containing attitude and position information to design an adaptive optimal controller with perturbation of internal parameters.When the dimensions of the state and input are given and a priori state feedback gain matrix that can make the system progressively stable,the optimal feedback control strategy can be obtained by using the system data collected online instead of the traditional solution of the Riccati equation(ARE).And the convergence and stability of the formation control algorithm are given.Finally,considering the situation of unknown disturbances in the model,two kinds of disturbances are mainly analyzed,which are the input disturbance during the transmission of information between UAVs and the uncertain wind disturbance outside the UAVs.Aiming at uncertain wind disturbance,a data-driven robust formation control algorithm is proposed to suppress the influence of disturbance.In this paper,the perturbed formation system control problem is transformed into a zerosum game stable equilibrium problem,and a data-driven iterative learning algorithm is designed to iteratively approximate the value function,control strategy,and disturbance strategy.The control and disturbance strategy is the stable equilibrium solution of the zero-sum game.