采煤机运行数据预处理算法研究  

Research on Preprocessing Algorithm of Shearer Operation Data

作　　者：寇元宝 汪崇建[2] 贠瑞光[2] Kou Yuanbao;Wang Chongjian;Yun Ruiguang(China Coal Research Institute,Beijing 100013,China;Shanghai Branch,Tiandi Science and Technology Co.,Ltd.,Shanghai 200030,China)

机构地区：[1]煤炭科学研究总院,北京100013 [2]天地科技股份有限公司上海分公司,上海200030

出　　处：《煤矿机械》2020年第10期41-43,共3页Coal Mine Machinery

摘　　要：针对当前采煤机设备运行数据量大,存在大量异常值、无效值与缺失值,此类数据没有被进行有效预处理和分析等问题,通过Python语言编写预处理算法,算法内容包括数据整合、上传数据库、异常值检测等。通过基于箱线图的异常值检测和DBSCAN聚类算法分别进行了异常值检测,对比2种方法的适用范围,并对预处理后的数据通过统计描述、聚类分析方法进行了探究,为其他采煤机运行数据预处理和分析提供了参考方案和思路。In view of problems that the current amount of operation data on the shearer equipmentis large,there are a large number of abnormal values,invalid values and missing values in the data,such data has not been effectively preprocessed and analyzed,the preprocessing algorithm was written by python language.The content of the algorithm include data integration,uploading to the database,outlier detection.Through outlier detection based on boxplot graph and DBSCAN clustering algorithm,outlier detection was carried out respectively.The scope of application of the two methods was compared,and the preprocessed data was explored by statistical description and the cluster analysis method,which provides reference schemes and ideas for other shearer operation data preprocessing and analysis.

关 键 词：采煤机 运行数据 数据清洗 数据分析 

分 类 号：TD412.6[矿业工程—矿山机电]