Predictive control and CSTR application based on step response model

Abstract

Predictive control is a kind of time-domain control method which can display and deal with the control problems of constrained nonlinear systems. It is a new type of computer control algorithm developed in recent years. In the investigation report issued by the International Federation of automatic control in April 19, PID control, system identification, estimation and predictive control after filtering are listed as the most important control technologies, and are considered as the most influential control methods in the future. The inherent robustness of predictive control solves the problem that has puzzled the control theory field for nearly a decade. As soon as it appears, it has been paid attention by the engineering circles at home and abroad, and has been widely and successfully applied in many industrial departments. The current research interests are predictive control, lupon control and the application of predictive control and robust control in electromechanical systems. In the actual industrial production, there is no need for a particularly fine mathematical model, which requires a method that can achieve high quality control effect while the model accuracy is not high First of all, mathematical modelling is needed. The function of prediction model is to predict and control the future output of the system by using the historical information of the controlled object and the assumed future input. Then the optimization algorithm is used for rolling optimization, and the predictive control adopts the finite time domain optimization, which is a kind of repeated online optimization. Last, By detecting the real-time state or output of the current system, the feedback information is used to make the next prediction and optimization closer to the reality before optimizing the control. To avoid the disturbance and system mismatch and other uncertainties, to compensate for the impact of this uncertainty.