Transformer state identification method based on hybrid sampling and ensemble learning

Abstract

The invention discloses a transformer state recognition method based on mixed sampling and ensemble learning. The method specifically comprises the steps that 1, collected data of dissolved gas in transformer oil is divided into two data sets; 2, SMOTE oversampling is conducted on the training set obtained in the step 1, and a data set obtained after SMOTE oversampling is conducted is recorded asa new fault training data set; 3, combining the obtained new normal training data set with the new fault training data set obtained in the step 2 to generate a new balanced data set; 4, taking a leastsquare support vector machine as a base classifier, and training q base classifiers by utilizing the q groups of balanced sub-data sets generated in the step 3; 5, integrating the q base classifiersobtained by training in the step 4 to obtain a strong classifier, and performing state recognition on the transformer; wherein the strong classifier obtained through combination is the transformer state identification optimal model, and testing the model. The method can be used for accurately identifying the state of the transformer.

Description

technical field [0001] The invention belongs to the field of transformer online monitoring and fault diagnosis, and in particular relates to a transformer state recognition method based on mixed sampling and integrated learning. Background technique [0002] As the key equipment of the power grid, the safe and stable operation of the transformer is the basis for ensuring the normal supply of electricity and the safety of the power system. Once the state of the transformer occurs, it will have a huge impact on the surrounding economy and life. Therefore, the state recognition of transformers has become a hot research topic for scholars at home and abroad. [0003] With the rapid development of artificial intelligence technology, traditional DGA-based methods such as three-ratio method, David's triangle, Rogers ratio method and other methods can no longer meet people's current requirements for transformer state recognition accuracy. Therefore, a series of intelligent identifi...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a transformer state recognition method based on mixed sampling and integrated learning, which can solve the problem of data imbalance and improve the accuracy of transformer state recognition

Images