Uncertain Desired Trajectory Tracking Control Of Nonlinear Uncertain Systems

The vast majority of the existing tracking control methods are based on the assumption that the desired target trajectory is precisely known/available a priori.However,in many important applications(such as missile interception,satellite tracking,power system planning and dispatching,etc.),the desired trajectory to be tracked might not be available precisely,and those all involve the whole tracking control of the uncertain desired trajectory.Therefore,the problem of enabling dynamic systems to track a moving target with uncertain trajectory is of theoretical and practical importance.In this project,we study the adaptive tracking control theory and schemes without the need for a priori precise knowledge of the desired target trajectory as follows:Firstly,the uncertain target tracking control problem of uncertain MIMO nonaffine nonlinear systems is investigated.To circumvent the difficulty that the uncertain desired trajectory can not directly be utilized for feedback control,a mathematical model based on the extended Kalman filter(EKF),neural network(NN)and memory reasoning(MR)for reconstructing the uncertain desired trajectory are proposed.Converting the original nonaffine system into an affine one by using the mean value theory.By carrying out the eigenvalue analysis for the two situations of the control gain as the square matrix and the non-square matrix and extracting the core function in order to solve the nonlinear problem,the robust adaptive controller is designed to achieve close target tracking and maintain system stability.Simulation is conducted to validate the effectiveness of the proposed control scheme.Secondly,the problem of uncertain desired trajectory tracking control for uncertain perturbed nonlinear systems with actuator failures is investigated.Primarily,For subsequent controller design,a mathematical model based on the extended Kalman filter(EKF),neural network(NN)and memory reasoning(MR)for reconstructing the uncertain desired trajectory are proposed.Then,the “virtual matrix” is innovatively introduced to deal with the unknown control gain and “health factor” of the actuator,and by using the method of extracting the core function to solve the nonlinear uncertainty of the system.Finally,the robust adaptive controller is designed.It is proved that the proposed control scheme guarantees satisfactory tracking performance,with all the close-loop signals being bounded.Numerical simulation results verify the feasibility of the developed method.Lastly,the uncertain target tracking control problem of uncertain perturbed MIMO systems with unknown actuation characteristics is investigated.Primarily,a mathematical model based on the extended Kalman filter(EKF),neural network(NN)and memory reasoning(MR)for reconstructing the uncertain desired trajectory are proposed.And to facilitate the controller construction,a smooth function is used to approximate the saturation function.Then,the RBF neural network is used to approximate resultant lumped nonlinearities and uncertainties in the system and by introducing the method of matrix decomposition to process the control gain of the system.Finally,the neuro-adaptive controller is designed,which ensures the uniformly ultimately boundedness of all the signals of the closed-loop system.Numerical simulation results verify the feasibility of the developed method.