摘要
目前乙烯装置追求高附加值产率等最佳效益,然而由于乙烯裂解原料来源较复杂、各种油品变化较为频繁,仅实现裂解炉平均出口温度定值控制是不能充分实现乙烯裂解装置的最佳效益的,因此开发深度串级控制是各乙烯装置的目标。介绍了通过一种基于裂解深度神经网络预测模型的Smith预估控制方案建立的裂解深度软测量模型,以及该裂解深度软测量模型与APC控制相集成的方法,以实现对裂解炉裂解深度的平稳控制。
Ethylene plant aims to achieve the best benefits.However,due to the complicated sources of ethylene cracking raw material and frequent oil changes,the best benefits of the ethylene plant cannot be fully achieved just by realizing the fixed set point control of average outlet temperature of cracking furnace. So depth cascade control technology is to be developed for ethylene plant. This paper introduces a soft-sensing model of cracking depth,which is established using Smith prediction control scheme based on neural network prediction model for cracking depth,and a method for integrating the soft-sensing model with APC control,so as to achieve the stable control of the cracking depth.
出处
《乙烯工业》
CAS
2013年第3期57-61,7,共5页
Ethylene Industry
关键词
裂解炉
深度控制
稳定
维护
Cracking furnace
Depth control
Stability
Maintenance
