The Research On Robust Model Predictive Control For Singular Systems With Uncertainty

Model Predictive Control(MPC) is a kind of computer optimization algorithm to predict the future behaviors of the controlled object. The control mechanism of MPC includes model prediction, receding optimization and feedback correction. Model predictive control(MPC) is one of most popular optimization strategies which has been widely adopted in industry process as an effective mean of dealing with multivariable and constrained control problems. Singular system describes a larger class of systems than normal linear systems and Model predictive control for the systems has been widely concerned. The paper investigates the robust MPC for a class of time delay singular systems with linear fractional uncertainty via state-feedback, the robust MPC for a class of time delay singular systems with uncertainty via output-feedback and the adaptive MPC for a class of nonlinear systems. The main contents of the thesis are as follows:1)The paper addresses robust model predictive control(MPC) for a class of time delay singular systems with linear fractional uncertainty and input constrains. The systems are transferred to the piecewise continuous descriptor systems and a piecewise constant control sequence is calculated by minimizing the worst-case quadratic objective function. At each sampling internal, by means of Lyapunov theory and optimization theory, the optimal problem with infinite horizon objective function is reduced to a convex optimization problem involving linear matrix inequalities. The sufficient conditions on the existence of the state feedback control are derived and expressed as linear matrix inequalities. Further, an iterative model predictive control algorithm is proposed for the on-line synthesis of state feedback controllers with the conditions guaranteeing that the closed-loop singular systems are regular, impulse-free and robust stable. Finally, a numerical example is presented to shows the efficiency of the proposed approach.2)The paper presents robust MPC scheme of output feedback for time delay singular systems with bounded disturbance. At each sampling internal, by Lyapunov theory, the optimal problem with infinite horizon objective function is reduced to a convex optimization problem minimizing the worst-case linear quadratic objective function and a control sequence is calculated. In terms of linear matrix inequalities and relaxation matrices, the sufficient conditions on the existence of output feedback control are derived and expressed as linear matrix inequalities. Further, an iterative MPC algorithm is proposed for the on-line synthesis of the output feedback controllers with the conditions guaranteeing the robust stability of the closed-loop systems. Finally, a simulation example is presented to show the effectiveness of the proposed MPC scheme.3)The problem of MPC is discussed for a class of nonlinear systems via parameter adaptation in this paper. A new parameter adaptive law and a receding horizon control law are proposed respectively. In the process of parameter adaptation, the value of estimated parameter is gradually closed to the real value, and the set of parameter became smaller on the basis of always contains the real value with the estimation of parameter. The parameter adaptation considered the influence of the parameter error estimation for the system. The research incorporates the robust features to guarantee closed loop stability and constraints satisfaction. The last simulation results are exploited to illustrate the efficiency of the proposed method.