基于改进型BP神经网络的打浆度软测量  

Soft Sensing for Beating Degree Based on Improved BP Neural-Network

作　　者：郭文强[1] 侯勇严[1] 孙瑜[2] 

机构地区：[1]陕西科技大学电气与电子工程学院,陕西咸阳712081 [2]南京理工大学自控系,南京210094

出　　处：《微计算机信息》2006年第10S期274-276,共3页Control & Automation

基　　金：陕西省教育厅产业化培育基金资助02JC17

摘　　要：将改进型BP神经网络算法应用于打浆度的测量,建立起打浆度的BP网络软测量模型。结果表明,该打浆度的神经网络模型能对打浆度进行较精确的预测,并为后续进行打浆过程控制和优化控制、提高打浆质量提供了良好的基础。The improved BP neural-network algorithm is applied in the measurement for beating degree. The relative BP neural-network soft sensing model of beating degree is established. Experimental results show this model could supply the accurate forecasting and provide the fair basis for following process control in beating and optimization control in order to improve the beating quality.

关 键 词：软测量 打浆度 神经网络 泛化 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]