基于BP神经网络的蒸煮卡伯值预测模型  

Kappa Number Prediction Model of Paper Pulp Cooking Based on BP Neural Network

作　　者：黄俊梅[1] 汤伟[1,2] 许保华 王孟效[1,2] 

机构地区：[1]陕西科技大学电气与信息工程学院,西安710021 [2]陕西西微测控工程有限公司,陕西咸阳712081

出　　处：《化工自动化及仪表》2010年第11期34-37,共4页Control and Instruments in Chemical Industry

摘　　要：卡伯值反映了纸浆的硬度,是稳定纸浆质量的关键。本文利用BP人工神经网络与工厂实际数据建立了一个简单的卡伯值BP神经网络模型。且介绍了以BP网络为基础的卡伯值预测模型在蒸煮终点预报系统中的应用。将该模型的仿真结果与LUO模型进行比较,可知利用BP人工神经网络进行纸浆卡伯值预测是可行的,且具有很大的理论意义和实用价值。Kappa number is the key to stabilize the quality of paper pulp as it reflects the hardness of pulp.A simple kappa number prediction model was proposed based on BP neural network and actual data.On the other hand,endpoint prediction system of pulping cooking process was introduced based on the presented BP model.Compared with the LUO model,the BP model of kappa number is feasible and it is significant in theory and application to develop BP model of Kappa number in cooking process.

关 键 词：卡伯值 BP神经网络(BPNN) 软测量 间歇蒸煮过程 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]