煮糖结晶过程的多模型动态矩阵控制研究

A nonlinear multi-model predictive controller of application to crystallizing process

由于工业结晶过程存在非线性和滞后现象,因此采用常规的控制策略已经达不到工业生产的要求。针对这种情况,通过对煮糖结晶过程的机理分析建立了反映实际煮糖过程的动态模型,在此基础上采用一种先进的预测控制算法即多模型非线性动态矩阵控制(DMC)算法进行控制。仿真和对比研究有力地证明了所建立的模型是有效的,同时也证明该算法可以应用到工业结晶过程弗能够取得较好的控制品质。

The nonlinear system with time-delay usually exists in the industry crystallizing process, so traditional control strategy can not meet the industry＇s requirement. Considering the complexity of crystallizing process, firstly author builds a model of boiling crystallization process based on mechanism analysis and an improved predictive control on nonlinear multi-dynamic matrix control algorithm （DMC） was adopted. Both simulation research and comparison results have powerfully proved that the model of boiling crystallization process is valid and nonlinear mnlti-DMC can be applied to crystallizing process control and get better control quality.