Transformer fault diagnosis method based on support vector description and K periapse neighbor

Abstract

The invention relates to a transformer fault diagnosis method based on support vector description and K periapse neighbor. The method comprises the following steps: acquiring content data of characteristic gas H2, CH4, C2H6, C2H4 and C2H2 dissolved in transformer oil by virtue of a technology for analyzing dissolved gas in the transformer oil, preprocessing and normalizing the data, forming a transformer fault diagnosis sample set, establishing a multi-classification suprasphere model based on a single-classification support vector description method, diagnosing a fault of a transformer, further classifying an aliasing domain sample in a support vector description multi-classification process by using a K periapse neighbor classification algorithm, so that the fault diagnosis accuracy of the transformer is increased. By combining the multi-classification support vector description method and the K periapse neighbor classification method, small samples and imbalance sample sets can be accurately classified, the fault diagnosis precision of the transformer can be remarkably improved, and powerful support can be provided for making a state maintenance decision of the transformer.

Description

technical field [0001] The invention belongs to the technical field of transformer fault on-line monitoring, and in particular relates to a transformer fault diagnosis method based on support vector description and K near-centroid neighbors. Background technique [0002] With the development of society, electricity has increasingly become an important part of the national economy and people's lives. Whether it is industrial and agricultural production or people's daily life, a stable and reliable power supply is required. The power transformer is mainly responsible for the transformation, distribution and transmission of electric energy in the power grid. As one of the key equipment of the power system, its operation status directly determines whether the power system can provide safe, stable and reliable power supply. The internal structure of the power transformer is complex and the operating environment is harsh. During operation, it not only has to withstand uneven elect...

AI Technical Summary

Problems solved by technology

However, these pattern recognition methods also have some problems: the application of artificial neural networks is often limited by conditions, the search space and calculation amount are large, the convergence speed is slow, and it is easy to fall into local mean and overfitting; support vector machine as a binary Classification algorithm, for the multi-classification problem of transformer fault diagnosis, it is often necessary to transform it through "one-to-many", "one-to-one" or "binary tree" methods to achieve multi-classification, so there will inevitably be error accumulation problems; The Yeesian network needs a large amount of sample data in the classification process, and needs to convert the state variables into discrete variables. In the discrete process, the state information of the transformer may be lost.

In addition, artificial neural network, support vector machine and Bayesian network have good classification performance under the condition that the sample data set is roughly balanced. Therefore, the data of the transformer fault state that can be obtained is far less than the data of the normal state. For the classification problem of unbalanced data sets with few fault samples and many normal samples, it is often difficult for these three pattern recognition methods to obtain satisfactory results. Effect

Images