A Transformer Fault Type Diagnosis Method Based on Semi-supervised dbnc

Abstract

The invention discloses a transformer fault type diagnosis method based on semi-supervised DBNC, which includes: selecting a sample data set; dividing the sample data into a pre-training set without labels, a labeled set, a test set 1 and a test set 2; State encoding of fault types; establishment of a transformer fault diagnosis model based on a deep belief network classifier; initialization of the parameters of each layer of the model; use of contrastive divergence to train each RBM at the bottom layer by layer; Optimize the parameters to make the network classification performance reach the global optimum; save the trained network, and use the sample data of the test set 1 to verify the classification performance of the network; solve the problem of using deep learning network fault data for transformer fault diagnosis Analysis and processing, usually only a small number of complete data samples can be obtained, and it is very difficult to obtain a large number of complete data samples with labels, which requires a lot of manpower and material resources.

Description

technical field [0001] The invention belongs to transformer fault diagnosis technology, in particular to a transformer fault type diagnosis method based on semi-supervised DBNC. Background technique [0002] As an important equipment for voltage conversion and power distribution in the power system, the power transformer is closely related to the safety and reliability of the power system. However, due to manufacturing defects, human factors, and weather effects, the fault diagnosis and development trend prediction of transformers have always been highly concerned. Most of the power transformers in our country are oil-immersed transformers. At the initial stage of transformer failure, the gas formed is dissolved in the oil, and when the fault energy becomes larger, free gas will be formed. Therefore, Dissolved Gas Analysis (DGA) in oil has become the main means of transformer fault diagnosis. [0003] At present, the fault diagnosis methods of power transformers based on ...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is: to provide a transformer fault type diagnosis method based on semi-supervised DBNC, to solve the existing technology for transformer fault diagnosis using deep learning network to analyze and process a large number of transformer fault data, and then diagnose the transformer fault type ; However, deep learning requires accurate and complete samples in order to obtain satisfactory results. Usually, only a small number of complete data samples can be obtained, and it is very difficult to obtain a large number of complete data samples with labels, which requires a lot of manpower and material resources

Images