A transformer state prediction method and system

Abstract

The invention discloses a transformer state prediction method, which comprises the following steps: (1) respectively training a long-term and short-term memory neural network and a deep belief neuralnetwork; (2) inputting the current concentration value of the characteristic gas in the transformer oil chromatographic data to be identified into the trained long-short-term memory neural network toobtain the future concentration value of the characteristic gas outputted from the trained long-short-term memory neural network; Obtaining a future concentration ratio of the selected gas based on the future concentration value of the characteristic gas; feeding The selected gas future concentration ratio into the trained deep belief neural network to obtain the transformer state prediction results. In addition, the invention also discloses a transformer state prediction system, which comprises a long-term and short-term memory neural network module and a deep belief neural network module. The transformer state prediction method can predict the transformer state efficiently and accurately, and has guiding significance for the operation and maintenance of the transformer.

Description

technical field [0001] The invention relates to a prediction method and system, in particular to a prediction method and system for transformers. Background technique [0002] During the operation and use of the transformer, due to the influence of the external environment, aging, electrical, thermal and other faults will occur, and a small amount of gas will be dissolved in the insulating oil. The volume fraction of various components of gas in oil and the difference in the proportion relationship between different components will be closely related to the operating state of the transformer. [0003] In the current prior art, although the dissolved gas in the transformer oil can be analyzed to a certain extent through historical data analysis and data models, there are still deficiencies. For example: the traditional research method is mainly to diagnose the fault at the current moment, lacking the evaluation of the future state of the transformer and the analysis of the f...

AI Technical Summary

Problems solved by technology

[0003] In the current existing technology, although the dissolved gas in the transformer oil can be analyzed to a certain extent through historical data analysis and data models, there are still deficiencies.

For example: the traditional research method is mainly to diagnose the fault at the current moment, lacking the evaluation of the future state of the transformer and the analysis of the fault early warning

In addition, when the state assessment and fault classification of transformers are carried out in the existing technology, the gas concentration ratio code is mainly used as the input of the model, and there are problems such as incomplete code and too absolute limit; and the prediction method usually only adopts a single or Less parameters to speculate on future development trends

Images