基于模型预测控制的反应再生工艺控制实现

摘要

The model predictive control technology in advanced control system was introduced. The dynamic matrix control algorithms in model predictive control technology including predictive model, rolling optimization, feedback correcting, were described in detail. Analyzed the muhivariable, multi-coupling characteristics of the reac- tion-regeneration process control and pointed out the difficuhy of control by taking the reaction-regeneration process of catalytic： cracking unit in oil refinery as an example. Finally, the optimized control was realized on the tempera- ture in the first stage regeneration and the second stage regeneration of the reaction-regeneration process by taking the dynamic matrix control algorithm, eventually, the control target was realized.%介绍了先进控制系统中的模型预测控制技术。具体阐述了模型预测控制中的动态矩阵控制算法（DMC），包括预测模型、滚动优化、反馈校正。以炼油厂催化裂化装氍的反应再生工艺为例，分析了反应再生工艺控制的多变量、多耦合特性，指出了控制的难点。最后，采用动态矩阵控制算法对反应再生工艺的一再、二再温度进行了优化控制，最终实现了控制目标。