Abstract:

Aiming at the compared color matching principle of colred spun yarn and difficult prediction on matched colors, the application characteristics of several prediction models for color matching of colored spun yarn were summarized and analyzed. The calculation precision of Stearns-Noechel model, Kubelka-Munk theory and BP neural model are ideal, and the precision of Friele model is lower. The Stearns-Noechel model and the Friele model,however, need to solve unknown parameters which have a great influence on the prediction accuracy. The Kubelka-Munk theory still has a gap with the ideal conditions and the calculation is cumbersome. The BP neural network requires a large number of training samples to enhance the generalization ability. Finally, it is pointed out that the color matching of color spinning yarns should improve the accuracy of the conventional model, and also improve the method for solving the unknown parameters in the color matching model and seek a new color matching model. Meanwhile, the personalization of the company's parameter settings should be paid moer attention and the coloring fibers of the undiluted liquid should be standardized. The complicated calculation in color matching is simplified, and the computer color matching technology of the color spinning yarn is improved.

Key words: colored spun yarn, color matching technique, Stearns-Noechel model, Friele model, Kubelka-Munk theory, neural network model