摘要
In this paper, the control of a butadiene distillation column is discussed and improved. At first, a neural network soft-sensor instrument of product quality was built based on abundant on-the-spot data collected by DCS and simulated data obtained by a theoretical model. Then, an inferential control scheme based on the soft-sensor was designed. By increasing logic and expert controllers in the inferential control arithmetic, the robustness of the system was enhanced. The practical application showed that the scheme could run smoothly over a long period and realized close-loop control of product quality.
The control of a butadiene distillation column is discussed and improved. At first, a neural network soft-sensor instrument of product quality was built based on abundant on-the-spot data collected by DCS and simulated data were obtained by a theoretical model. Then, an inferential control scheme based on the soft-sensor was designed. By increasing logic and expert controllers in the inferential control arithmetic, the robustness of the system was enhanced. The practical application showed that the scheme could run smoothly over a long period and realized close-loop control of product quality.
出处
《化工学报》
EI
CAS
CSCD
北大核心
2004年第2期331-334,共4页
CIESC Journal
