首页 | 本学科首页   官方微博 | 高级检索  
     检索      

合成醋酸乙烯的空时得率和催化剂选择性软测量
引用本文:徐余法,俞金寿.合成醋酸乙烯的空时得率和催化剂选择性软测量[J].上海交通大学学报,2007,41(8):1339-1342.
作者姓名:徐余法  俞金寿
作者单位:1. 上海电机学院,上海,200240
2. 华东理工大学,自动化研究所,上海,200237
摘    要:以某石化厂乙烯气相法合成醋酸乙烯反应过程的空时得率和催化剂选择性软测量建模为研究对象,基于现场采集数据及机理分析,确定了辅助变量.在对现场数据进行处理的基础上,建立了基于径向基函数神经网络的多输入多输出神经网络软测量模型,取得了较好的效果.仿真结果表明,模型精度达到了工艺要求,可用于指导生产,为实现先进控制和优化控制创造了条件.

关 键 词:软测量  空时得率  催化剂选择性  多输入多输出  神经网络模型
文章编号:1006-2467(2007)08-1339-04
修稿时间:2006-10-10

Soft-Sensing on the Rate of Non-Occupied Time and Catalyzer Selection of Synthetic Acid Ethene
XU Yu-fa,YU Jin-shou.Soft-Sensing on the Rate of Non-Occupied Time and Catalyzer Selection of Synthetic Acid Ethene[J].Journal of Shanghai Jiaotong University,2007,41(8):1339-1342.
Authors:XU Yu-fa  YU Jin-shou
Institution:1 Shanghai DianJi Univ. , Shanghai 200240, China; 2. Research Inst. of Automation, East China Univ. of Science and Technology, Shanghai 200237, China
Abstract:A soft-sensing on the rate of non-occupied time and catalyzer selection of process of synthetic acid ethene in a refinery was studied.The auxiliary variables are determined according to the actual data and the process mechanism analysis.On the basis of processing actual data,the soft-sensing models based on multiple-input and multiple-output(MIMO) neural network are established.The simulation results show that the models are accurate and the method achieves good performance,which can provide favorable conditions for refinery to realize advanced control and optimal control.
Keywords:soft-sensing  rate of non-occupied time  catalyzer selection  multiple-input and multiple-output(MIMO)  neural network model
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