文献类型：期刊文献

中文题名：连续催化重整装置辛烷值软测量研究

英文题名：Research on Soft-Sensing of Octane Number in the Process of Continuous Catalytic Reforming

作者：孙自强[1];顾幸生[1];党晓恒[1];俞金寿[1]

机构：[1]华东理工大学自动化研究所,上海200237

年份：2001

卷号：13

期号：Z1

起止页码：171

中文期刊名：系统仿真学报

外文期刊名：Journal of System Simulation

收录：CSTPCD;;Scopus;CSCD:【CSCD2011_2012】；

基金：受教育部高等学校骨干教师资助计划项目;上海市曙光计划项目资助

语种：中文

中文关键词：辛烷值;软测量;神经网络;BP;PLS-BP;建模

外文关键词：RON; soft-sensing; neural networks; BP; PLS-BP; modeling

摘要：:研究连续催化重整生产过程中重整产品辛烷值软测量问题。首先结合工艺机理分析,找出影响辛烷值的主要参数,确定为辅助变量,建立起基于BP神经网络的辛烷值软测量模型。其次再采用PLS-BP算法改进模型结构,建立起基于PLS-BP神经网络的辛烷值软测量模型,可以看出PLS-BP所建模型结构相对简单,更适用于实际过程。最后对模型校正后在现场使用,结果表明测量结果令人满意。
In this paper, soft-sensing of research octane number (RON) about reforming production in the process of continuous catalytic reforming is researched. At first according to the analysis of technics principles several major parameters affecting RON are found out and used as assistant variables. Then a model of measuring RON is established by use of neural networks with BP algorithm. In order to modify the model structure, another model of measuring RON is established by use of neural networks with PLS-BP algorithm. It is obvious that the second model structure is simpler correspondingly and more suitable for practice. Finally the model is applied in practice after revised. The result of measurement is satisfactory.