Judge Yarn Quality Assurance System Based On The The Hvi Detection System Build

This paper evaluates raw cotton by classification and membership in fuzzy mathematics, predicts the qualification by the curve fitting and combination prediction in order to find the best blending strategy and create the quality security system based on HVI data according to the blending application of enterprise. It aims at transform the way of blending from the artificial blending based on perception detection to instrumentation detection.It mainly expounded the the elimination of exceptional data, comprehensive evaluation of the cotton quality,cotton management and the relation between the cotton and yarn quality for the creation of the quality security system based on HVI data.The computer blending based on HVI request more accuracy data. So there are many different ways to realize the data cleansing.It is found that the UHML, UI, Strength which are the four most important indexes among all the indexes by fuzzy clustering and partial analysis. The technical grade is created to evaluate the quality inside, while the color grade evaluate the quality outside. The locality, technical grade and color grade are considered the three new indexes for the cotton management instead of locality, commercial grade and handle length.The relation between the cotton and yarn quality is analyzed quantitatively. Create the forecast model of yarn quality.C14.8tex as an example. With C14.8 tex for example, realization to of the raw cotton and the yarn quality the quantitative analysis of relation, and start to build up the C14.8 tex yarn quality estimate model.