基于BP神经网络的秸秆发酵过程乙醇浓度软测量方法  

Soft measurement method of ethanol concentration based on BP neural network

作　　者：马凤英 于文志 MA Fengying;YU Wenzhi(School of Electrical Engineering and Automation,Qilu University of Technology,Jinan 250000,China)

机构地区：[1]齐鲁工业大学电气工程与自动化学院,山东济南250000

出　　处：《传感器与微系统》2021年第7期148-150,160,共4页Transducer and Microsystem Technologies

基　　金：济南市高校自主创新项目(201401210);山东省住房城乡建设科学技术计划项目(RK017)。

摘　　要：针对秸秆发酵制备燃料乙醇过程中,乙醇浓度实时在线检测困难、离线检测滞后性大等难题,设计了一种考虑时滞的神经网络软测量模型。首先对部分过程变量进行分析与处理;其次引入时滞参数T减小离线测量带来的滞后性影响;最后使用反向传播(BP)神经网络建立软测量模型。仿真结果表明:加入时滞参数后,辅助变量与关键变量之间具有更高的相关性,能够更精确实现对乙醇浓度的实时软测量。Aiming at the problem that in the process of preparing fuel ethanol by straw fermentation,real-time online detection of ethanol concentration is difficult,and the lag of off-line detection is large,a neural network soft measurement model considering time lag is designed.Firstly,some process variables are analyzed and processed.Secondly,the time-delay parameter T is introduced to reduce the hysteresis effect caused by offline measurement.Finally,back propagation(BP)neural network is used to establish the soft-measurement model.The simulation results show that after adding the time-delay parameter,the auxiliary variable has a higher correlation with the key variable,which can realize the real-time soft measurement of ethanol concentration more accurately.

关 键 词：秸秆发酵 乙醇浓度 软测量 时滞参数 反向传播(BP)神经网络 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置] TQ015.9[自动化与计算机技术—控制科学与工程]