Robust Model Predictive Control For A Class Of Systems With Actuator Saturation

Actuator saturation is a very common constraint problem in industrial control.In industrial control system design,if we didn't considered it may lead to major accidents easily.The problem of actuator saturation do not usually exists in isolantion,the other control problems are also accompany with it.It is difficult to obtain the most appropriate control method if we only considering the system actuator saturation problem.However,few studies have been done on the control of such systems with coexisting conditions.In this thesis,mainly studies the control method of a kind of system with actuator saturation characteristic.The main research objects include:multiple time-delay systems with actuator saturation characteristics and actuator saturation systems with bounded nonlinear disturbances.In this thesis,robust model predictive control for systems with actuator saturation characteristics is studied in two aspects:(1)For systems with both actuator saturation and multiple time delays,a design method of robust model predictive controller is presented,which is based on the predictive control principle of receding optimization,the quadratic performance index of the infinite time domain is optimized by means of Lyapunov stable theory and linear matrix inequalities(LMIs)technique.First,the multiple time-delays of the system model will be processed,converted into a general linear saturated system model,and then the saturation term is processed by convex hull representation to obtain the system model.At each sampling moment,the complex min-max optimization problem is transformed into a convex optimization problem with LMI constraints for solving,and the effectiveness of the robust model predictivecontroller is proved by simulation.(2)For a class of actuator saturation systems with bounded nonlinear disturbances,a robust model predictive controller design method is presented,which is based on the predictive control principle of receding optimization,the quadratic performance index of the infinite time domain is optimized by means of Lyapunov stable theory and linear matrix inequalities(LMIs)technique.In this thesis,an innovative saturation processing method is used to deal with the saturation term of the system.This method can effectively avoid solving complex attraction domain problem and reduce the difficulty of solving the controller.First,the quadratic performance index optimization problem in infinite time domain is solved approximately.Then,the maximum upper bound of the nonlinear disturbance term is optimized.The disturbance suppression problem of robust predictive control in norm-bounded sense is studied quantitatively,and the sufficient conditions for the existence of robust predictive controller are given.Finally,the feasibility and superiority of the proposed method are verified by simulation.