| Batch reactor is a critical unit for batch industry process, good control performance plays an important role for products quality and system security. Especially today when batch production industries are still widely used, the research of advanced control strategies for batch process is significant.The aim of advanced batch control strategies is to track the pre-set trajectory in a fast and exact way under the limit of operation condition and restriction. Because of inherent nonlinear and time-variant property in batch reaction process, its control system design is always a challenge work and many advanced control strategies was produced ,in which the predictive control was the most popular one. In most of all the predictive controllers, the nonlinear mechanism model was used to construct the predictive model, while the mechanism model can hardly be obtained accurately and little effective method can be found to identify the parameter in this model, which restrict the control performance.This paper describe a kalman filter based model predictive control for batch reactor temperature set-point tracking in the frame of linear MPC. Instead of nonlinear mechanism model, linear one was used as predictive model. Under the condition of the unknown effect on the output from the heat generation, which is the most important uncertainty, we make use of system model with input disturbance to model the relationship between the temperature of the jacket and the reactor. Then, augment the system state with the disturbance vector and linear Kalman filter was used to estimate the disturbance and its effect on the system output. Finally, a new DMC algorithm with compensation for the disturbance was obtained and make sure of the offset-free control. Otherwise, in order to provide a effective simulation platform for control algorithm, this paper make the design of a real-time industry process simulation platform as another content. |
PDF Full Text Request