基于多关联参数特征子空间的纺纱质量波动预测  

Spinning Quality Fluctuation Prediction Based on Feature Subspace of Mul-Correlation Parameters

作　　者：李哲 胡胜 张守京[1] 李文 LI Zhe;HU Sheng;ZHANG Shoujing;LI Wen(School of Meehanical and Electrical Engineering,Xi'an Polytechnic University,Xi'an 710048,China)

机构地区：[1]西安工程大学机电工程学院,陕西西安710600

出　　处：《轻工机械》2022年第5期22-28,共7页Light Industry Machinery

基　　金：中国纺织工业联合会指导计划项目(2020112);陕西省自然科学基金(2022JQ-721);西安工程大学博士启动基金项目(BS201834)。

摘　　要：针对纺纱生产过程影响因素多、监测维度广导致的过程波动难以分析和纱线质量难以预测的难题,课题组提出一种基于多关联参数特征子空间的纺纱质量波动预测方法。首先分析影响纱线质量的关联参数之间关系,构造能够表征纱线质量波动的特征子空间;然后构建面向特征子空间的纱线质量深度学习预测模型,实现纱线质量智能预测。通过实例进行分析,结果显示提出的方法能够有效分析纱线质量的多关联参数波动规律,并能准确对纱线质量进行预测。Aiming at the problems of the analysis process fluctuation and the prediction of yarn quality caused by multiple influencing factors and wide monitoring dimensions in the spinning production process, a method for predicting spinning quality fluctuations based on feature subspaces of multiple correlation parameters was proposed. Firstly, the relationship between the correlated parameters that affects yarn quality was analyzed, and the feature subspace that characterizes yarn quality fluctuations was constructed. Then the feature subspace-oriented deep learning prediction model of yarn quality was constructed to realize the intelligent prediction of yarn quality. Finally, through the analysis of examples, the results show that the proposed method can effectively analyze the fluctuation law of multi-correlation parameters of yarn quality and accurately predict yarn quality.

关 键 词：纺纱过程 质量波动预测 多关联参数 深度信念网络 特征子空间 受限玻尔兹曼机 

分 类 号：TH165.4[机械工程—机械制造及自动化] TS111.9[轻工技术与工程—纺织材料与纺织品设计]