摘要
为了方便快捷检验棉花颜色级,采用BP神经网络算法构建了基于棉纤维色度参数的棉花颜色级预测模型。模型以棉纤维的色度参数明度值和黄度值为输入变量,以棉花颜色级为输出变量,采用三层网络,网络的拓扑结构为2-12-1。通过对280份棉花试样的颜色级进行测试和BP神经网络模型预测,结果表明:BP神经网络预测值与实测值的平均相对误差为3.203%,所建立的回归方程拟合优度达到99%。认为:所建立BP神经网络模型具有较好的预测效果。
In order to quickly and easily test cotton color scale,cotton color scale prediction model based on cotton fiber chrominance parameter was constructed by BP neural network algorithm.Cotton parameter chromaticity parameter brightness value and yellowness value were used as input variables,cotton color scale was used as output variable.Three-layer network was adopted and network topology was 2-12-1.Through testing color scale and BP neural network model prediction of 280 cotton samples,the results showed that the average relative error between predicted and measured values of BP neural network was 3.203%.Goodness-of-fit of established regression equation was 99%.It is considered that established BP neural network model has a better prediction effect.
作者
李帅
单国华
贾丽霞
钟民
刘瑞
LI Shuai;SHAN Guohua;JIA Lixia;ZHONG Min;LIU Rui(Xinjiang University,Xinjiang Urumqi,830046;Xinjiang Fiber Inspection Bureau,Xinjiang Urumqi,830046)
出处
《棉纺织技术》
CAS
北大核心
2019年第3期68-71,共4页
Cotton Textile Technology
基金
新疆维吾尔自治区质量技术监督局科技计划项目(XJZJ20161230160034)
关键词
棉花颜色级
测色配色仪
色度参数
BP神经网络
神经元个数
预测模型
Cotton Color Scale
Color Measurement and Matching Meter
Chrominance Parameter
BP Neural Network
Number of Neuron
Prediction Model
作者简介
李帅(1995—),男,在读硕士研究生;单国华,通信作者,讲师,sgh_116@126.com.
