基于Spark流处理的轴承剩余使用寿命预测  

Prediction of Remaining Service Life of Bearing Based on Spark Flow Processing

作　　者：马荣杰 吴文江[2] 胡毅[2] MA Rong-jie;WU Wen-jiang;HU Yi(University of Chinese Academy of Sciences,Beijing 100049,China;Shenyang Institute of Computing Technology,Chinese Academy of Science,Shenyang 110168,China)

机构地区：[1]中国科学院大学,北京100049 [2]中国科学院沈阳计算技术研究所,沈阳110168

出　　处：《组合机床与自动化加工技术》2021年第11期19-22,共4页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家科技重大专项(2018ZX04002001-009);辽宁省“兴辽英才计划”科技创新领军人才项目(XLYC1902091)。

摘　　要：对数控机床关键部件的剩余使用寿命进行了研究,利用Kafka消息中间件对接收到的数据进行缓存,解决数控机床发送数据和Spark服务器接收数据的速度差异问题。在分析了轴承水平振动加速度和剩余使用寿命的关系后提出了第一种基于线性模型的剩余使用寿命预测方法,在分析了轴承水平加速度和垂直加速度平方和后提出了第二种基于相似度的剩余使用寿命预测方法。通过对比发现基于相似度的方法准确率较高,并将其结果写入MySQL数据库进行存储,通过web页面进行展示,方便工作人员进行维护。The residual service life of key components of computer numerical control machine tools is studied.Kafka message middleware is used to cache the received data to make up for the speed difference between sending data from CNC machine tools and receiving data from spark server.After analyzing the relationship between horizontal vibration acceleration and residual service life of bearing,the first residual service life prediction method based on linear model is proposed.After analyzing the square sum of horizontal acceleration and vertical acceleration of bearing,the second residual service life prediction method based on similarity is proposed.Through comparison,it is found that the accuracy of the method based on similarity is high,and the results are written into MySQL database for storage,and displayed through web page,which is convenient for staff to maintain.

关 键 词：数控机床 大数据 流处理 相似度 

分 类 号：TH16[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]