Power transformer state prediction method based on machine learning and neural network

Abstract

The invention discloses a power transformer state prediction method based on machine learning and a neural network. The method comprises the following steps: (1) training an RP neural network and a deep belief network; (2) outputting a future concentration value and a future concentration ratio of a to-be-identified transformer characteristic gas by utilizing the trained RP neural network and thedeep belief network; and (3) predicting the future operation state of the transformer based on the future gas concentration ratio. The future operation state of the transformer can be effectively predicted; the operation state of the transformer is predicted, potential threats of the transformer can be perceived in time, and the fault development trend of the transformer is mastered; and great significance is achieved for improving the equipment operation safety and reliability.

Description

technical field [0001] The invention relates to a prediction method, in particular to a state prediction method for a transformer. Background technique [0002] Although the transformer itself has a variety of protective measures, such as lightning arresters, gas, differential and grounding protection, the accident rate is still very high due to the complex insulation structure inside the transformer, uneven distribution of internal electric and thermal fields, and long-term uninterrupted work. high. The online monitoring technology of transformer oil chromatography is the basis of fault diagnosis and online monitoring. However, the technical cost and time cost of online monitoring technology are high, the sampling data often has problems such as singular values ​​and missing values, and the detection cycle of dissolved gas in oil is too long Long, unable to achieve online monitoring. Transformer faults are mostly irreversible faults. Even if the transformer is disconnecte...

AI Technical Summary

Problems solved by technology

[0002] Although the transformer itself has a variety of protective measures, such as lightning arresters, gas, differential and grounding protection, the accident rate is still very high due to the complex insulation structure inside the transformer, uneven distribution of internal electric and thermal fields, and long-term uninterrupted work. high

The online monitoring technology of transformer oil chromatography is the basis of fault diagnosis and online monitoring. However, the technical cost and time cost of online monitoring technology are high, the sampling data often has problems such as singular values ​​and missing values, and the detection cycle of dissolved gas in oil is too long long, unable to achieve online monitoring

Images