Application Of Predictive Control Algorithm In Hysteresis System

In actual industrial process control,systems often exhibit varying degrees of uncertainty,nonlinearity and other characteristics that are influenced by many objective factors.Strictly speaking,most controlled objects have different degrees of time lag.The existence of lag-time makes it impossible for disturbances to be detected immediately,which will undoubtedly increase the adjustment time and produce a large overshoot,resulting in a serious decline in the stability of the control system.Therefore,it is very important to study the control algorithm of the time-delay system.Effects on systems caused by time lag can be overcome by traditional control algorithms theoretically,but it requires precise mathematical models.However,in the actual industrial process,it is difficult to establish an accurate mathematical model due to the presence of uncertainty,nonlinearity,and other characteristics.At this time,an advanced control algorithm is needed to overcome the above problems.The algorithm not only needs to satisfy the requirement that the mathematical model does not strictly depend on the controlled object,but also has the ability of time delay compensation and anti-interference.Model Predictive Control exactly meets the above requirements.The algorithm optimizes the input at the next sampling time based on the predicted output of the system at each sampling time and it is only relevant to the initial input of the system.With the advantages of strong adaptability,good control performance and strong robustness,it is considered to be one of the most suitable control algorithms for the study of time-delay systems.In this paper,the time-delay system is taken as the control object,and different kinds of model predictive controllers are designed according to different time-delay objects.Simulation results are used to verify the control effect of the designed predictive controllers.The main contents are as follows:(1)A dynamic matrix control algorithm for a typical first-order time-delay system of an electric heating furnace is proposed.A stable closed-loop system can be obtained by adjusting the control parameters.By extracting the eigenvalues of the characteristic matrix,it can be verified whether the designed state estimator and controller are stable.By introducing disturbance variables,anti-jamming capability of the system can be illustrated.Finally,the effect of parameter changes on the control effect of the entire system such as control weight matrix and error weight matrix is studied through simulation experiments.(2)Considering uncertain systems with multi-state and input delays,distributed model predictive control algorithm is proposed.Firstly,for the model predictive control problem in multi-state time delay systems,independent controllers for each subsystems are designed.By solving a set of linear matrix inequality optimization problems,the state feedback control law of subsystems and the upper bound of the performance index of the closed-loop system are obtained.Then,the results are extended to uncertain systems with multi-state and input delays,and the conservativeness of designed controllers are reduced by introducing multi-state feedback control laws.Finally,the effectiveness of the proposed scheme is verified by simulation experiments.