基于BP神经网络的液化石油气芳构化反应体系的预测模型  

Prediction Model for Aromatization Reacting System of LPG Based on Back-Propagation Neural Network

作　　者：黎小辉[1] 朱建华[1] 武本成[1] 郝代军[2] 

机构地区：[1]中国石油大学化学工程学院,北京102249 [2]中国石化集团洛阳石油化工工程公司,河南洛阳471003

出　　处：《化学工业与工程》2012年第1期39-42,72,共5页Chemical Industry and Engineering

摘　　要：将BP神经网络应用于液化石油气芳构化反应体系,建立了该体系在指定反应条件下的神经网络模型,网络的训练结果达到了预设要求,且网络输出结果与试验值能够较好的吻合,所建模型能较准确地预测指定反应条件下该体系的产物分布状况。分析结果可为LPG芳构化技术的进一步开发、生产过程中的产物分布预测和工业过程的模拟提供一些指导。In this paper, back-propagation neural network was applied to the system of LPG aromatization. The neural network model of this reacting system was established under given reacting conditions. The network＇s training results have achieved present requirements and the network output could match the experimental data well. This model could predict products distribution of the LPG aromatization system under given conditions. The results could provide some reference for further development on LPG aromatization, prediction of product distribution and simulation of the industrial processes.

关 键 词：LPG芳构化 BP神经网络模型 预测 

分 类 号：TE65[石油与天然气工程—油气加工工程]