余月琳 顾学锋 吴乐元 陈慰来

摘 要：為了提高添纱衬垫织物平方米质量的预测精度，在生产过程中更便捷地管控品质和核算成本，根据平方米质量的经验计算公式选择5种不同组织工艺参数编织30种涤棉混纺添纱衬垫织物，测试平方米质量值后利用Matlab进行单因素和多元回归分析，探讨面纱线密度、面纱含棉量、衬垫纱线密度、衬垫纱含棉量、面纱与地纱长度比这5种组织工艺参数对平方米质量的影响规律;使用SPSS多层感知器分析各影响因子的权重;构建平方米质量预测模型。结果表明：其他参数固定时，5种组织工艺参数均与1隔1添纱衬垫织物平方米质量呈正相关;影响因子权值从大到小分别为衬垫纱线密度、面纱线密度、面纱与地纱长度比、面纱含棉量、衬垫纱含棉量;构建的多元回归预测模型相对误差在±2%以内，预测精度较好，可为添纱衬垫织物平方米质量的预测提供理论与实践指导。

关键词：添纱衬垫织物;平方米质量;涤棉混纺;多元回归;预测模型

中图分类号： TS184.1

文献标志码：A

文章编号：1009-265X（2021）06-0078-06

收稿日期：2020-06-29 网络首发日期：2021-01-25

基金项目：浙江理工大学研究生培养基金项目

作者简介：余月琳（1997-），女，浙江金华人，硕士研究生，主要从事针织产品与技术方面的研究。

通信作者：陈慰来，E-mail： wlchen193@163.com

Research on the Weight Per Square Meter of Invisible Fleecy Fabricand the Construction of a Prediction Model

YU Yuelin1，GU Xuefeng2，WU Leyuan3， CHEN Weilai1

（1.College of Textile Science and Engineering （International Institute of Silk）， Zhejiang Sci-TechUniversity， Hangzhou 310018， China;

2.Ningbo Myfitt Textile Technology Co.， Ltd.， Ningbo 315700，China;

3.Zhejiang Huzhou Meiyue Knitting Co.， Ltd.， Huzhou 313009， China）

Abstract： In order to improve the prediction accuracy of the weight per square meter of invisible fleecy fabric and control the quality and calculate the cost in the production process more easily， we chose 5 different process parameters to knit 30 kinds of polyester/cotton invisible fleecy fabric according to the empirical formula. After testing， Matlab was used to perform univariate and multivariate regression analysis to explore the influence law of 5 process parameters， i.e.， the linear density of face yarn， cotton content of face yarn， linear density of laid-in thread， cotton content of laid-in thread， and the length ratio of face yarn to ground yarn on the weight per square meter. An SPSS multilayer perceptron was adopted to analyze the weight of each influence factor， and a prediction model for weight per square meter was constructed. The research results show that： when other parameters are fixed， 5 kinds of knitting parameters are positively correlated with the weight per square meter of the fabric with 1 separator and 1 invisible fleecy; from large to small， the order of the weights of the influence factors are the linear density of laid-in thread， linear density of face yarn， the length ratio of face yarn to ground yarn and cotton content of laid-in thread. The relative error of the constructed multiple regression prediction model is within ±2%， and the prediction accuracy is good， which can provide theoretical and practical guidance for the prediction of weight per square meter of invisible fleecy fabric.

Key words： invisible fleecy fabric; weight per square meter; polyester/cotton; multiple regression; prediction model

平方米质量作为织物内在质量的重要物理指标，是成品布和服装贸易中需考核的重要参数。在国内、国际市场交易过程中，客户为了减少收货风险，常申请品质证书，其主要的考核指标往往是织物的平方米质量。……