首页 | 本学科首页   官方微博 | 高级检索  
     

常压塔柴油凝点动态软测量模型的研究
引用本文:毛帅,熊智华,徐用懋,庄爱霞,黄海龙,王立群. 常压塔柴油凝点动态软测量模型的研究[J]. 控制工程, 2005, 12(4): 342-345
作者姓名:毛帅  熊智华  徐用懋  庄爱霞  黄海龙  王立群
作者单位:清华大学,自动化系,北京,100084;中石化,济南分公司,山东,济南,250101
基金项目:国家自然科学基金资助项目(60404012)
摘    要:研究了某炼油厂常压塔三线柴油凝点的软测量建模问题。分析了影响柴油凝点的多种因素,并充分利用仪表分析值提供的被测变量历史信息,建立了一种神经网络和kvinson预测器相结合的动态软测量模型,该模型消除了分析值存在纯滞后的影响。针对某炼油厂常压塔三线柴油凝点的软测量,对该模型进行了验证。仿真研究表明,该模型的预报准确性要优于静态软测量模型,取得了较好的预测效果。

关 键 词:神经网络  Levinson预测器  动态软测量模型  柴油凝点
文章编号:1671-7848(2005)04-0342-04
修稿时间:2005-02-28

Dynamic Soft-sensor Model of Diesel Oil Solidifying Point on a Crude Distillation Unit
MAO Shuai,XIONG Zhi-hua,XU Yong-mao,ZHUANG Ai-xia,HUANG Hai-long,WANG Li-qun. Dynamic Soft-sensor Model of Diesel Oil Solidifying Point on a Crude Distillation Unit[J]. Control Engineering of China, 2005, 12(4): 342-345
Authors:MAO Shuai  XIONG Zhi-hua  XU Yong-mao  ZHUANG Ai-xia  HUANG Hai-long  WANG Li-qun
Affiliation:MAO Shuai~1,XIONG Zhi-hua~1,XU Yong-mao~1,ZHUANG Ai-xia~2,HUANG Hai-long~2,WANG Li-qun~2
Abstract:Neural network based soft sensor models of diesel oil solidifying point(DOSP) on a crude distillation unit are discussed.To fully utilize historial data set of output variable obtained by online analysis device,a dynamic soft sensor model of DOSP is developed by a feed-forward neural network that is combined with a Levinson predictor.The Levinson predictor provides multi-step-ahead predictions of output recursively,so the model can diminish the output delays of online analysis device significantly.The simulation results show that the proposed model of DOSP is better than a general static soft sensor model.
Keywords:neural network  Levinson predictor  dynamic soft sensor model  diesel oil solidifying point
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号