Research on Fault Diagnosis of Power Equipment Based on Big Data

Abstract

Aiming at the big data generated by the on-line monitoring system of power equipment under the smart grid, a technical scheme for fault diagnosis of power equipment is proposed. The wavelet packet analysis technique, Renyi entropy theory and Teager energy operator are combined to extract and analyze the fault feature vector from a large number of data monitored. Based on the feature vector of the fault information, the state is divided into four states: normal operation, abnormal operation, early warning and warning. Training with historical data can form a neural network to achieve the assessment and decision-making of the operation status of power equipment. The results can be applied to the whole life-cycle management system of power equipment.