Self-Correlation Analysis Framework with Property Data in Master Data Management — A Case on Power Utility Equipment Retire Analysis

Abstract

Causing Reason Locating (CRL) is a negative going decision making process. It provides specific individual causing reasons to events or abnormal variation so that decision maker can find out clear-cut points for updates or maintenance. Traditional CRL is usually implemented with self-correlation analysis on Value Comparable Data (VCD). Property Data (PD) is not covered in CRL for its nominal evaluations. This paper initiates a new scheme for PD data correlation analysis, including virtual data creation method and correlation balancing method for Kendall correlation computation. A numerical study is implemented for model support on equipment retire analysis.