A Novel 10kV XLPE Cable Failure Prediction Method

摘要

The purpose of this study was to predict 10kV cable failures based on deficiency data, and provide data basis for maintenance adjustment. Deficiency Data Analysis (DTA) tree is built on history data. Then probability distribution of deficiency and failure data is constructed considering the family characteristics. Kullback-Leibler distance is used to analyze the correlation between deficiency and failure, which is used to predict failure based on monitoring and testing data. Shenzhen Nanshan power supply company adopted the algorithm analysis achievement, as a result, the cable failure number drop 18% comparing to last year. It is proved that this algorithm can be used to predict malfunctions of cables, then according maintenance could be carried out and enhance system reliability.