石油炼制与化工 ›› 2024, Vol. 55 ›› Issue (10): 148-156.

• 控制与优化 • 上一篇    下一篇

基于模型预测控制的柴油加氢装置分馏单元产品质量控制优化

博士1,孙延吉2,潘艳秋1,王成宇1,刘沛鑫1   

  1. 1. 大连理工大学化工学院
    2. 北京溥络数智科技有限责任公司
  • 收稿日期:2024-02-29 修回日期:2024-05-30 出版日期:2024-10-12 发布日期:2024-09-26
  • 通讯作者: 潘艳秋 E-mail:yqpan@dlut.edu.cn

OPTIMIZED CONTROL OF PRODUCT QUALITY IN FRACTIONATOR OF DIESEL HYDROREFINING UNIT BASED ON MPC

#br#   

  • Received:2024-02-29 Revised:2024-05-30 Online:2024-10-12 Published:2024-09-26

摘要: 针对传统比例积分(PI)控制对过程复杂、系统强耦合的柴油加氢装置分馏单元的稳定时间和控制效果不佳的问题,基于某石化企业柴油加氢装置分馏单元原有PI控制方案,提出模型预测控制(MPC)协同PI控制方案,并探讨了该方案控制分馏单元的可行性及控制效果。结果表明:MPC协同PI的控制方案能够提高装置的动态性能;在汽提塔进料量和汽提蒸汽量波动的情况下,PI控制和MPC方案的震荡次数相近,但MPC方案的稳定时间和控制最大偏差均更小;对于±10%进料量的扰动,PI控制和MPC方案的最大温度相对偏差分别为3.48%和0.07%,稳定时间分别为3.5h和1.5h,绝对误差积分(IAE)评价值分别为7.30和0.23,说明MPC方案的控制效果比PI控制更优;同时,验证了MPC协同PI控制的可行性,表明MPC协同PI控制更适用于柴油加氢装置分馏单元的先进控制。

关键词: 模拟, 产品质量, 比例积分, 模型预测控制, 分馏单元

Abstract: The stability time and control effect of traditional proportional integral (PI) control for fractionator of diesel hydrorefining unit with complex process and strong coupling of the system are not good. Based on the original PI scheme of fractionator of diesel hydrorefining unit in a domestic petrochemical enterprise, a model predictive control (MPC) co-PI control scheme was proposed. The results showed that the control scheme of MPC and PI could improve the dynamic performance of the device, and the oscillation frequency of PI control was similar to that of MPC control scheme when the feed rate and steam rate of stripper fluctuate, but the stability time and the control maximum deviation of MPC were smaller. For disturbances of ±10% feed rate, the maximum temperature relative deviation of PI control and MPC scheme was 3.48% and 0.07%, the stability time was 3.5 h and 1.5 h, and the value of absolute error integral (IAE) was 7.30 and 0.23, respectively, and the control effect of MPC was better than that of PI control. At the same time, the feasibility of MPC and PI control was verified, which showed that MPC combined PI control was more suitable for advanced control of fractionator of diesel hydrorefining unit.

Key words: simulation, product quality, PI, model predictive control, fractionation unit