基于多元回归的原液着色纤维混色规律研究
|
摘要
研究原液着色纤维混色规律。设计了不同比例的三种原液着色纤维进行混色,采用多元回归的方法建立了明度、彩度和色相与单色纤维比例含量的数学模型。同时对该模型进行回归显著性检验以及预测集数据预测。结果表明:建立的数学模型,其决定系数R~2几乎均大于0.9,方程F显著性检验和回归系数t检验结果均为极显著;利用建立的多元回归方程对预测集样本颜色进行预测,64%的样本满足色差要求。认为:采用该数学模型,多数样本基本可以实现一次配色成功,少数样本可以通过进一步修色来达到色差要求。
Color mixture principle of mass-dyed fiber based on multiple regression was studied.Three kinds of mass-dyed fibers blended with different ratios were designed.The method of multiple regression was adopted to build the mathematical model among the lightness,saturation and hue with the fiber proportion content.Meanwhile,the regression significance of the model was tested and the prediction set data.The results showed that the determination coefficient of the mathematical model R~2 was almost all larger than 0.9.The significance test of equation F and regression coefficient t test results were both outstanding.By using multiple regression equation,the color of prediction set sample was predicted,the color difference requirement of the 64% samples could be met.It is considered that color matching of most samples can be achieved in one time by adopting the mathematical model.Minor samples can meet the requirement of color difference by further color repairing.
Color mixture principle of mass-dyed fiber based on multiple regression was studied.Three kinds of mass-dyed fibers blended with different ratios were designed.The method of multiple regression was adopted to build the mathematical model among the lightness,saturation and hue with the fiber proportion content.Meanwhile,the regression significance of the model was tested and the prediction set data.The results showed that the determination coefficient of the mathematical model R~2 was almost all larger than 0.9.The significance test of equation F and regression coefficient t test results were both outstanding.By using multiple regression equation,the color of prediction set sample was predicted,the color difference requirement of the 64% samples could be met.It is considered that color matching of most samples can be achieved in one time by adopting the mathematical model.Minor samples can meet the requirement of color difference by further color repairing.
引文
[1] 金剑,李鑫,王颖,等.原液着色纤维产业“十三五”发展路径[J].中国纺织,2016(12):64-65.
[2] 何成栋,万晓霞,黄新国,等.用于计算机配色的反射光谱数据质量评估方法研究[J].光谱学与光谱分析,2016,36(11):3805-3809.
[3] 袁理,代乔民,付顺林,等.基于原色纤维混配色织物的呈色特性与影响因素分析[J].纺织学报,2018,39(10):38-43.
[4] 马崇启,程璐,金晓,等.Stearns-Noechel模型的全光谱纱线配色算法[J].光谱学与光谱分析,2018,38(8):2488-2492.
[5] 沈加加.基于Friele模型的色纺纱光谱配色研究[J].针织工业,2013(5):32-36.
[6] 陈典红,李海明,徐新胜,等.基于共轭梯度法的纱线染色配方预测优化算法[J].纺织学报,2011,32(3):82-85.
[7] 董红安.用于染色的神经网络计算机配色算法[J].印染助剂,2011,28(5):19-22.
[8] 李楠,张秉森.多元非线性回归在织物染色计算机配色中的研究[J].青岛大学学报(工程技术版),2013,28(4):62-67.
[9] 王功明,刘志勇.基于光谱表示和独立成分分析的混合颜料成分分析方法[J].光谱学与光谱分析,2015,35(6):1682-1689.
[10] 王华清.计算机测色配色技术[M].上海:东华大学出版社,2012:54-56.
[2] 何成栋,万晓霞,黄新国,等.用于计算机配色的反射光谱数据质量评估方法研究[J].光谱学与光谱分析,2016,36(11):3805-3809.
[3] 袁理,代乔民,付顺林,等.基于原色纤维混配色织物的呈色特性与影响因素分析[J].纺织学报,2018,39(10):38-43.
[4] 马崇启,程璐,金晓,等.Stearns-Noechel模型的全光谱纱线配色算法[J].光谱学与光谱分析,2018,38(8):2488-2492.
[5] 沈加加.基于Friele模型的色纺纱光谱配色研究[J].针织工业,2013(5):32-36.
[6] 陈典红,李海明,徐新胜,等.基于共轭梯度法的纱线染色配方预测优化算法[J].纺织学报,2011,32(3):82-85.
[7] 董红安.用于染色的神经网络计算机配色算法[J].印染助剂,2011,28(5):19-22.
[8] 李楠,张秉森.多元非线性回归在织物染色计算机配色中的研究[J].青岛大学学报(工程技术版),2013,28(4):62-67.
[9] 王功明,刘志勇.基于光谱表示和独立成分分析的混合颜料成分分析方法[J].光谱学与光谱分析,2015,35(6):1682-1689.
[10] 王华清.计算机测色配色技术[M].上海:东华大学出版社,2012:54-56.