Study On The Robust Fuzzy Predictive Control Methods And Their Applications In The Coal-fired Unit

Many actual systems are essentially nonlinear,and the manipulated variables are also constrainted and there exist many kinds of disturbances and uncertainties,this dissertation studies the nonlinear robust fuzzy model predictive control methods and its related problems,and applies them to the coal-fired units to improve the control performance.The main achievements of this dissertation include:(1)This paper presents a gap-based fuzzy modeling method for the nonlinear model.By systematically investigating the nonlinearity of the nonlinear model at a series of equilibrium points,the linearization point of the nonlinear model is determined,which makes the Takagi-Sugeno fuzzy model have appropriate complexity and approximation accuracy to the nonlinear model.Finally,the proposed method is applied to the nonlinear mechanism model of coal-fired units,and its effectiveness is verified by simulation(2)To enhance the feasibility of fuzzy predictive control algorithm,this dissertation develops relaxed stabilization conditions for discrete-time Takagi–Sugeno fuzzy control systems based on the extended nonquadratic Lyapunov function,the nonparallel distributed compensation law and the convexity of the fuzzy blending coefficients.Three new results are proposed to further reduce the conservatism by fully exploring the slack matrix technique and introducing new slack matrices and extra collection matrices.The new stabilization conditions are gradually less and less conservative,and more advantageous than the existing results by the overall consideration of the conservatism and the computational efforts.A well-known numerical case and a practical case are carried out to demonstrate the effectiveness of the proposed stabilization conditions.(3)Considering the unknown disturbance and the constraints on the control,a robust offset-free fuzzy model predictive tracking control strategy for nonlinear systems based on the extended state observer is designed assuming that the disturbance and its first-order difference are bounded and the first-order difference equals zero in steady state,and the disturbance and output dimension are equal.By adding disturbance terms to Takagi-Sugeno model,a general disturbance model is constructed to characterize the dynamic characteristics of the nonlinear system.Based on the accurate estimation of the unknown disturbance by the extended state observer and the proposed relaxed stabilization condition of Takagi-Sugeno fuzzy control system,a robust offsetfree fuzzy model predictive tracking control strategy is designed.With the proposed scheme,the closed-loop system is theoretically proved to be asymptotically stable and offset-free in output tracking,while the input constraints are satisfied by both the free control variable and the future control input in the form of the state feedback law.Finally,the effectiveness of the proposed control method is verified by the simulation results of the coal-fired unit control system.(4)Assuming that the disturbance and its first-order difference are bounded and the first-order difference becomes zero in steady state,and the disturbance dimension is no longer specifically required,a disturbance observer based fuzzy predictive control strategy for nonlinear systems is designed using the traditional nonlinear disturbance observer in this dissertation.It is composed of nominal fuzzy model predictive control law and disturbance compensation control law.The nominal fuzzy model predictive control is obtained through the simplification of the proposed robust offset-free fuzzy model predictive tracking control algorithm,and ensures the nominal control system is asymptotically stable,and the control input satisfies the constraints on both amplitude and change rate.Under the disturbance assumption,based on the accurate estimation of disturbance by traditional nonlinear disturbance observers,the disturbance compensation gain is designed such that the influence of the disturbance is removed from the output channels by the proposed control strategy at the steady state,thus ensuring that the closed-loop control system can asymptotically track the system set values.Finally,the proposed control strategy is applied to the coal-fired unit which improves the robustness of the unit control.(5)Aiming at the bounded high-order complex disturbance that is more generalized than the disturbance considered before,a disturbance observer based fuzzy model predictive control strategy for nonlinear systems is proposed using a novel generalized discrete nonlinear disturbance observer.Since the traditional nonlinear disturbance observer cannot accurately estimate highorder complex disturbances,a generalized discrete-time nonlinear disturbance observer is constructed by systematically extending it.The proposed generalized observer exhibits a series structure of the internal states,and can precisely estimate the disturbance if its order is equal to or greater than that of the disturbance.The nominal fuzzy model predictive tracking controller is designed based on the robust offset-free fuzzy model predictive tracking control algorithm,which guarantees the asymptotic stability of nominal control system and the satisfaction of control input constraints.At the same time,by properly designing the disturbance compensation gain,the disturbance compensation control law eliminates the influence of higher order disturbance on the output of closed-loop system in steady state.Therefore,the proposed control strategy can still guarantee the asymptotic tracking performance of the closed-loop control system under bounded high-order disturbances,and is verified by the simulation of the control system of coal-fired units.