基于切削电流指数新异检测的长壁采煤机监测方法研究  

Novelty Detection of Longwall Shearer Based on Cutting Current Index

作　　者：秦国军[1] 胡茑庆[1] 官志强[2] Hal Gurgenci 

机构地区：[1]国防科技大学机电工程与自动化学院,湖南长沙410073 [2]澳大利亚昆士兰大学采矿联合研究中心

出　　处：《湖南科技大学学报（自然科学版）》2006年第2期55-58,共4页Journal of Hunan University of Science And Technology：Natural Science Edition

基　　金：国家自然科学基金资助项目(50375153)

摘　　要：在定义采煤机切削电流指数的基础上,采用统计分析和支持向量机方法,对采煤机健康状态进行了新异故障检测方法的研究,通过大量正常数据,构建和训练了新异类检测器.通过对澳大利亚某煤矿1号采煤机的实测数据应用表明,该检测器可有效检测意外发生的未知故障.图7,表3,参8.Shearer is one of the key machines in longwall mining system and the production largely depends on it. To avoid downtime of the machine during routine maintenance and ensure better availability, many parameters weve measured during coal cutting. In order to integrate the information in these parameters and monitor the statues of the machine and detect novel faults in time, a useful cutting current index was de.fined firstly in this paper. Then, the methods for novelty detection based on statistical analysis as well as support vector machine （SVM） were studied for health condition monitoring. Based on the application of a large amount of normal data, two types of novelty detectors were built and trained. Finally, a real case of fault detection for the data from a certain shearer of Australia was given and discussed, which showed that both of the detectors can valid detect the unknown fault. 7figs., 3tabs., 8refs.

关 键 词：长壁采煤机 切削电流指数 新异类检测 统计分析 支持向量机 

分 类 号：TD76[矿业工程—矿井通风与安全]