摘要
纺织热轧机的故障具有多发性、隐秘性等特点,故障排查大多依赖于人工经验,如何快速定位产品故障位置,是及时维修的关键。对此,针对纺织热轧机展开了故障分类算法研究,首先基于优化的N-Gram模型对纺织热轧机故障文本进行分词处理;然后利用基于词频-逆文档频率的方法提取故障文本特征向量;最后采用遗传算法优化的支持向量机(Support Vector Machine,SVM)对故障文本建立分类模型,经实例验证了遗传算法对支持向量机分类模型优化的有效性,其分类预测精确性较高,具有良好的使用价值。
The faults of the textile hot rolling mill have had the characteristics of frequent occurrence and concealment.The troubleshooting has mostly depended on the experience of the workers.How to quickly locate the fault location of the product has been the key to timely maintenance.In this regard,the fault classification algorithm research has been carried out for the textile hot rolling mill.First,the fault text of the textile hot rolling mill has been segmented based on the optimized N-Gram model;then the fault text feature vector has been extracted by the method based on the word frequency-inverse document frequency;Finally,the support vector machine optimized by genetic algorithm has been used to establish a classification model for the fault description text.The example has verified the effectiveness of the genetic algorithm on the optimization of the support vector machine classification model.Its classification prediction accuracy has been high and model has had good use value.
作者
王少伟
徐锋
晁海涛
刘宇
左敦稳
WANG Shaowei;XU Feng;CHAO Haitao;LIU Yu;ZUO Dunwen(School of Electrical and Mechanical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)
出处
《现代制造工程》
CSCD
北大核心
2021年第6期116-121,共6页
Modern Manufacturing Engineering
基金
教育信息化2.0时代高校智慧校园建设模式及实践探索的研究项目(NW202003)。
关键词
故障诊断
文本分类
特征提取
支持向量机
遗传算法
fault diagnosis
text classification
feature extraction
Support Vector Machine(SVM)
Genetic Algorithm(GA)