Abstract:

Based on dynamic model identification, soft measurement and online optimization, model predictive algorithm was applied to the process of the coal-based syngas to ethylene glycol. First, the advanced control structure and model of the process were given in detail. Then some key indexes of the distillation column, such as temperature, pressure, process quality were predicted online, monitored and closed-loop controlled. Therefore, the operation stability of the production process of synthesis gas to glycol was improved, the conversion rate of raw materials and product yield were also increased, which lead to the energy consumption of the process decreased largely. The industry applications show that the manipulated stability of the distillation was largely improved after the application of the advanced control technique, standard deviation of some main manipulated control variables was decreased about 20%, steam was decreased 3.67% compared the past and labor intensity was also lightened than past.

Key words: coal-based syngas to ethylene glycol, dynamic model identification, model predictive algorithm, advanced control

摘要：

为提升合成气制乙二醇生产过程的操作平稳性、增加反应原料转化率和产品收率，提高装置节能降耗水平，本文针对合成气制乙二醇装置的特点，基于多变量预测控制技术、动态模型辨识、软测量、在线优化技术，搭建过程生产的先进控制结构及模型，实现了对各精馏塔关键指标——温度、压力、产品质量的实时监测、预测及闭环优化控制。工业装置应用表明，先进控制实施后，合成气制乙二醇生产过程的精馏单元操作平稳性大幅提升，主要被控变量的波动标准方差降低20%以上，过程操作强度有效降低，蒸汽消耗降低3.67%。

关键词: 合成气制乙二醇, 动态模型辨识, 多变量预测控制, 先进控制