Abnormal data detection for multivariate alarm systems based on correlation directions

详细信息    查看全文

• 作者：Yan Yu^a ; ^{yuyan_coe@pku.edu.cn} ; Di Zhu^a ; ^b ; ^{zhudi@pku.edu.cn} ; Jiandong Wang^c ; ^{jiandong@sdust.edu.cn} ; Yan Zhao^d ; ^{zhaoyan@yjy.sd.sgcc.com.cn}
• 关键词：Industrial alarm systems ; Abnormal data segments ; Correlation direction ; Piecewise linear representation
• 刊名：Journal of Loss Prevention in the Process Industries
• 出版年：2017
• 出版时间：January 2017
• 年：2017
• 卷：45
• 期：Complete
• 页码：43-55
• 全文大小：1677 K
• 卷排序：45

文摘

A method is proposed to detect abnormal data segments for multivariate time series. Correlation directions are exploited as the features to detect abnormal conditions. Key turning points are determined by a piecewise linear representation of segments. Spearman's rank correlation coefficients are calculated for correlation directions. Numerical and industrial examples are provided as illustrations.