摘要
该文针对化工生产过程设计开发软测量建模软件。考虑到实际工业过程复杂多变,软件采用了线性回归建模和非线性神经网络建模两种方法来设计。通过对化工生产过程苯乙烯装置乙苯含量和淀粉含量的建模软测量应用,结果表明软件能够实现软测量的趋势预测功能。
This paper designs a soft-sensing modeling software for chemical production process,Considering the complexity in the practical industry process,the software applies the linear regression modeling approach and the nonlinear neural network modeling approach to design the measurement software.Practice have been carried on the production process of Ethyl benzene and Starch content prediction,and the results show that the software can fulfill the function of trend prediction.
作者
侯岩松
颉刚
张敏
刘亚儒
HOU Yan-song,XIE Gang,ZHANG Min,LIU Ya-ru （Automation Research Institute of Lanzhou Petrochemical Company Petrochina,Lanzhou 730060,China）
出处
《电脑知识与技术》
2011年第2期753-756,共4页
Computer Knowledge and Technology
关键词
软测量
神经网络
软件设计
soft-sensor
neural network
software development
