基于灰色-BP神经网络理论的上染率模型研究  

Research on Dye-uptake Rate Model Based on Gray-BP Neural Network Theory

作　　者：徐文龙[1] 汪澜[1] 张永兴[2] 

机构地区：[1]浙江理工大学先进纺织材料与制备技术教育部重点实验室,杭州310018 [2]浙江理工大学机械与自动控制学院,杭州310018

出　　处：《浙江理工大学学报（自然科学版）》2014年第4期343-347,353,共6页Journal of Zhejiang Sci-Tech University(Natural Sciences)

基　　金：国家自然科学基金(61074154)

摘　　要：以活性黄3RE上染棉织物为例,首先利用灰色系统GM(1,1)和Verhulst建立起上染率-染色工艺单因素模型,再将其输出直接作为神经网络的输入,最终建立灰色-BP神经网络上染率-染色工艺多因素模型,其拟合值的相对误差小于1.3%,并通过实验验证,预测值的误差均在1.0%以内。验证结果表明,该数学模型精确度较高,能较准确地反映棉织物活性染料染色后的上染率,并可以满足预测上染率的需求。This paper takes dyeing of cotton fabric with Reactive Yellow 3RE for example.Firstly,dye-uptake rate and dying single-factor model was established with gray system GM（1,1）and verhulst.Secondly,the output directly served as the input of neural network,and finally dye-uptake rate and dying multi-factor model based on gray-BP neural network was established.The relative error of the fitting values was less than 1.3%.The experimental result shows that the error of the predicted value is within1.0%.The verification results show that the mathematical model has high accuracy and can exactly reflect the dye-uptake rate of reactive dyes in the dyeing process and meet the actual requirement of forecasted dye-uptake rate.

关 键 词：活性染料 棉织物 灰色-BP神经网络 上染率 多因素模型 

分 类 号：TS190.9[轻工技术与工程—纺织化学与染整工程]