Analysis Of Mixed Model Of Ma - Ray Graphic Fiber

The colored fiber blend was also known as colored spun yarn, which contained two or more than two kinds of different colored fiber. The yarn has unique color mixing effect compared to ordinary white yarns.But there were also problems in the production,for example color matching efficiency and accuracy was low.At present, most of the domestic enterprises, especially small and medium enterprises are using traditional artificial color match method, and have no color match equipment, Which greatly reduces the production efficiency of enterprises.To solve the problem of colored spun yarn matching difficult,the analysis of fiber hybrid model in gray spun yarn would be studied and carried out from different angles.The raw material of gray spun yarn is the mixture of black viscose fiber and white viscose fiber, also the mixture of black viscose fiber and white polyester fiber,and comparie the model differences between the two conditions.On one side,simulating the match of dyes on fabrics,the software of Data color MATCH would give a formula of certain yarn sample,and the color matching results were good.On the other side,based on color mixture model of Kubelka-Munk two-constant theory,we can calculate the values of colored fiber absorption coefficient and scattering coefficient s, and further get the colored fiber formula.The conclusion is that the theory of K-M two-constant theory has a certain degree of applicability in color matching.In addition,Using the nonlinear fitting function of Origin8.0,the paper analyzed the relationship between the sum of reflectance under different wavelengths and the colored fiber quality ratio.The paper choose three exponential functions in the Origin software to fit the curve of the sum of reflectance under different wavelengths and the colored fiber quality ratio with the three curve fitting coefficients more than 0.95, and the final choice of the function was y=y0+/A1e-x/t+A2e-x/t2.In view of the three layers BP neural network having simulation arbitrarily complex nonlinear mapping ability in the case of enough hidden nodes,the neural network was applied to analyse the relationship between the reflectance value of the colored spun yarn and the colored fiber quality ratio,and the color matching error was less than 5%.But the BP neural network also showed the disadvantages, such as slow operation time, sometimes more than a few minutes.Therefore,we combined genetic algorithm with BP neural network to improve the problem, and the improved neural network was greatly improved in the computation speed, also had a certain degree of improvement in the aspect of precision.The research gave a contrast among the five methods application in color matching, the results showed that, whether one component gray spun yarn or two-component gray spun yarn, color results using Ga-BP neural network had the minimum color matching error, then followed by the Origin of nonlinear fitting method, which had poor performance in small-proportion colored spun yarn but good average error. K-M two-constant theory, BP neural network, and computer color matching system had the similar color matching effect.On the basis of theory, this paper mainly discussed the application of the theory in the actual color matching, and provided an effective theoretical basis for the hybrid model of the colored spun yarn and the earlier realization of the computer color matching.