Transformer fault diagnosis analysis method based on bayesian network

Abstract

The invention discloses a transformer fault diagnosis analysis method based on a bayesian network. The method includes the specific steps: step 1, determining transformer oil chromatography property variables Y = { Y1, Y2, Y3... Yn } and fault type variables D = { D1, D2, D3... Dm }, wherein yi is a value of Y1, dm is a value of Dm, using characteristic gas of a transformer oil chromatography as the transformer oil chromatography property variables, and using transformer fault types as the fault type variables; step 2, determining classification models, structural parameters and probability parameters of the bayesian network according to the transformer oil chromatography property variables and the fault type variables; and step 3, determining the transformer fault types by utilization of a connection tree algorithm.

Description

technical field [0001] The invention relates to a fault diagnosis and analysis method, in particular to a Bayesian network-based transformer fault diagnosis and analysis method. Background technique [0002] With the development of society, electric power has increasingly become an important part of the national economy, and the rapid development of modern industry and agriculture has put forward higher requirements for power transmission and transformation. In 2009, the State Grid Corporation proposed to build a smart grid with UHV as the backbone grid and coordinated development of power grids at all levels, and the six links of the strategic framework for smart grid development highlighted the importance of power transmission and transformation. The safe and reliable operation of the intelligent substation is one of the main conditions for the stable operation of the entire smart grid, and the intelligent power transformer is an important part of the intelligent substatio...

AI Technical Summary

Problems solved by technology

[0005] The application number is: 201210196906, and the title of the invention is a power transformer system state analysis and maintenance decision-making method. In this application, it is necessary to calculate the probability distribution function h0(y) of the historical monitoring data Y of the same type of transformer in the normal state and the probability of the fault state Distribution function h1(y), especially with the improvement of transformer fault diagnosis technology, the probability of transformer failure is also greatly reduced. Insufficient fault information in this regard will easily lead to a decrease in the accuracy of the final transformer diagnosis result

[0006] The application number is: 201210358681, and the name of the invention is an intelligent fault diagnosis method based on rough Bayesian network classifier. Most of the application actually introduces the application of rough set principle, and the introduction of Bayesian network is only for seeking probability and applying it to , does not mention the advantages of Bayesian network classifier for transformer fault classification, which expands the amount of calculation

Images