多变量约束控制器及其在反应再生控制中的应用

Improved Multivariable Constrained Controller and Its Application in Reactor-Regenerator System

提出了一种改进的多变量约束控制器 (MCC)及其在反应再生控制系统中的应用 .基于模型预测控制技术的MCC是一种带有约束、多变量和灵活控制任务切换的优化控制器 .在优化算法中 ,使用全局线性规划 (GLP)和DMC滚动优化相结合的方式 ,将经济指标约束和软硬约束有机的结合起来 ;在控制体系中 ,多变量约束控制器使用多模块控制结构设计 ,有效提高了控制器的在大规模控制问题上的运行实时性 .在组态体系上 ,改进后的MCC使用了一个带有C/S体系的分布式数据库作为控制器组态数据载体 ,使得控制器能够同时提供本地和远程的数据访问 ,有效提高了工程师的工作效率 .实际的反应再生控制结果表明 。

This paper represents a novel architecture of advanced process control (APC) - multivariable constraint control (MCC) and its application in reactor regenerator system. Based on model predictive control techniques, MCC is a multivariable constrained optimal controller with easy task switching. In the optimization phase, MCC combines a global linear programming in steady optimization and DMC horizon optimization together. In MCC control strategy, a multi module structure enables MCC to run efficiently in large control task. Using a C/S architecture in its configuration, MCC can be configured by a remote mode, which greatly enhances engineer's working efficiency. A case study in reactor regenerator system shows that, MCC performances well in petrochemistry and other process industry.