摘要
针对丁二烯生产装置精馏塔塔顶控制回路不能实时测量丁二烯-1,3的问题,提出了基于软测量技术的先进控制方案。利用从集散控制系统采集的大量现场数据,运用基于多元线性回归方法的软测量建模技术,建立了塔顶丁二烯-1,3产品纯度的软测量模型,并基于此对控制回路进行改造,设计了推断控制方案。通过DeltaVDCS系统自带的CL语言编写软仪表程序,并在控制策略组态环境Control Stud io中实现控制回路的改造。现场运行结果表明,控制系统能够有效地解决精馏塔产品质量不能在线实时检测和直接质量闭环控制的问题,实现了丁二烯-1,3产品纯度的闭环控制,显著改善了控制效果。
In view of the control problems of the distillation columns in the butadiene plant, the advanced control strategy based on soft-sensing technique is proposed. Using voluminous field data collected from DCS and the soft measurement modeling technology based on multiple linear regression method, the soft-sensing model of butadiene content in the distillation column is built. At the same time, the inferential control strategy is designed to rebuild the control loops. The program of the soft-sensing instrument is written with CL in DeltaV DCS. The control loops are reconstructed in Control Studio. The application result shows that the control system can resolve effectively the problems that the product quality cannot be measured on-line and be close-loop controlled directly, and that the effect of control is improved greatly.
出处
《化工自动化及仪表》
EI
CAS
2005年第4期11-14,共4页
Control and Instruments in Chemical Industry
关键词
丁二烯
软测量
推断控制
DELTAV系统
精馏塔
产品纯度
butadiene
soft-sensing
inferential control
DeltaV system
distillation column
product purity
