A Cluster Analysis Method for Transformer Status Evaluation Based on Data Imbalance Metrics

Abstract

The invention discloses a transformer state evaluation cluster analysis method based on data unbalanced measurement. parameters, and use the method of proportional normalization to process the index parameters; randomly select two groups of data in the index parameters as the initial cluster centers, and set the cluster analysis parameters; calculate the difference between the various types of fault index parameters and the initial cluster centers Euclidean distance, according to the Euclidean distance, the data in the unbalanced monitoring data is divided into the lower approximation set or boundary area of ​​the cluster, and the unbalanced degree between the clusters is calculated; the unbalanced degree of the clusters is combined to measure the membership degree of the monitoring data According to the cluster data distribution, iterative calculation is performed on the cluster center; finally, the state evaluation of the power transformer is carried out according to the clustering result; the invention effectively improves the state evaluation accuracy of the power transformer.

Description

technical field [0001] The invention belongs to the technical field of power transformer state evaluation, and in particular relates to a transformer state evaluation cluster analysis method based on data imbalance measurement. Background technique [0002] As an important power transmission and transformation equipment, power transformers are located in the key nodes of the entire power transmission and transformation network, and their safe and stable operation is essential to ensure power supply. Once a failure occurs, it will not only cause significant economic losses, but also affect social stability. Therefore, it is of great significance to evaluate the operation status of power transformers by analyzing the monitoring data of power transformers to maintain the safe and stable operation of the entire power system. By analyzing the fault data recorded by the fault recorder, the traditional state assessment method can better diagnose the fault type that has occurred in...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problem that the existing clustering algorithm has poor effect on processing the power transformer unbalanced monitoring data set, the present invention provides a transformer state evaluation cluster analysis method based on data unbalanced measurement, which is based on common fault indicators of power transformers system, from the unbalanced monitoring data, select the index parameters corresponding to different types of fault analysis of power transformers, and use the method of proportional normalization to process the index parameters; finally, use the rough fuzzy k-means clustering method of unbalanced measurement The power transformer state assessment method performs cluster analysis on unbalanced monitoring data, and obtains the corresponding power transformer state assessment results; the specific technical implementation plan is as follows:

Images