A State Evaluation Method of Metal Oxide Arrester Based on EEMD-FCM

This paper proposes a new method for predicting the preliminary aging of MOA (metal oxide arrester). First, the pre-processing of the collected MOA-related data is performed to standardize the data, and the data set is divided into a training set and a test set. Then the EEMD (Ensemble Empirical Mode Decomposition) analysis and processing is performed on the key feature of leakage current to extract key frequency domain features. Finally, FCM is performed on multiple features in the time and frequency domains to obtain two central points corresponding to the normal state and preliminary aging state of the MOA, respectively. By comparing the distance between the test set data and the two center points to determine whether the MOA has entered a preliminary aging. Through verification and comparison experiments, it is finally determined that the accuracy of the evaluation algorithm can reach more than 95% when the ambiguity is set to 1.5, and the prediction effect is excellent.