摘要
本文采用具有对角递归神经网络结构及相应计算方法，依据生产过程的实测数据，初步研究了磷酸羟胺法（ＨＰＯ法）中环己酮肟化投酮量的在线预估并得到了较为满意的结果，为进一步探索环己酮肟生产的优化操作做了一些尝试。
This paper presents a diagonal recurrence neural network and its corresponding algorithm dealing with the hyam phosptate oximation method to be studied for the on line preestimation of cyclohexanone according to the actually measured data in the cyclohexanone oxime production process.A fairly satisfied result has been obtained,thus making it possible to optimize the operation further in the cyclohexanone exime production process.
出处
《工业仪表与自动化装置》
1999年第3期3-5,共3页
Industrial Instrumentation & Automation
基金
国家自然科学基金
关键词
HPO法
神经网络
肟化
投酮量
在线预估
己内酰胺
Hyam phosphate oximation(HPO) Cyclohexanone Diagonal recurrence neural network(DRNN) Modelling pre estimation
