基于线性回归模型的经编送经量预测研究  

Research on Warp Run-in Prediction Based on Linear Regression Model

作　　者：李言 蒋高明[1,2] Li Yan;Jiang Gaoming(College of Textile and Clothing,Jiangnan University,Wuxi,Jiangsu 214122,China;Engineering Research Center for Knitting Technology,Ministry of Education,Jiangnan University,Wuxi,Jiangsu 214122,China)

机构地区：[1]江南大学纺织服装学院,江苏无锡214122 [2]江南大学针织技术教育部工程研究中心,江苏无锡214122

出　　处：《针织工业》2020年第10期1-4,共4页Knitting Industries

基　　金：国家自然科学基金(61772238);泰山产业领军人才(tscy20180224)。

摘　　要：针对经编生产过程中送经量难以确定、调试费时费力且易造成资源浪费的问题,提出一种基于线性回归模型的送经量预测算法。以经编产品数据库中的大量产品工艺数据为基础,详细介绍经编工艺设计参数的筛选、模型的训练及工作原理,同时使用真实生产数据进行预测试验,并进行模型评估,给出模型改进方案。测试结果表明,模型的预测误差仅为10.57%,具有较好的拟合预测效果,同时可以通过增加特征数量、优化数据结构等方法使模型预测性能进一步提升。Aiming at the problem that during the warp knitting manufacture process,the warp run-in being hard to determine,the debugging is time-consuming and laborious,and the resource wasting is big,a warp let off volume prediction algorithm based on linear regression model is proposed.Based on a large number of product process data in warp knitting product database,the selection of warp knitting process design parameters,model training and working principle are introduced in detail.At the same time,the real production data is used for prediction test,the model evaluation is carried out,and the model improvement scheme is given.The results show that the prediction error of the model is only 10.57%,which has a good fitting and prediction effect.At the same time,the prediction performance of the model can be further improved by increasing the number of features and optimizing the data structure.

关 键 词：经编 送经量 线性回归模型 数据挖掘 

分 类 号：TS184.3[轻工技术与工程—纺织材料与纺织品设计]