Transformer fault detection method based on nuclear capsule neuron coverage

Abstract

The invention discloses a transformer fault detection method based on nuclear capsule neuron coverage. Fault detection of a transformer in a power system is taken as a background, and a problem that transformer fault detection is low in recognition efficiency and even cannot be correctly recognized due to the fact that transformer oil chromatographic data obtained through actual monitoring is missing, redundant, uncertain in dimension and the like is solved. The kernel thought is introduced to carry out effective feature extraction of the oil chromatography data, then improvement is carried out on the basis of a super-sausage neuron construction method, concepts such as incidence relation and expansion and contraction rate are introduced in the creation process of manifold coverage neurons, and an optimized kernel capsule coverage algorithm is used for identifying oil chromatography characteristic data, so higher precision of transformer fault detection is achieved.

Description

technical field [0001] The invention belongs to the technical field of power grid fault detection methods, and in particular relates to a transformer fault detection method based on nuclear capsule neuron coverage. Background technique [0002] The healthy and long-term development of electric power enterprises is inseparable from the normal operation of the power system, and the normal operation of the power system is inseparable from the normal operation of the transformer. Transformer is one of the important equipment in the power system. When it fails during operation, the transformer insulating oil will undergo a chemical reaction to change the color of the insulating oil and the concentration of low-molecular gases. Therefore, the color and characteristics of the transformer insulating oil All gases have a certain correlation with transformer failure. [0003] The traditional chromatographic analysis method can be used to identify the cause of transformer faults in or...

AI Technical Summary

Problems solved by technology

[0004] In view of the above-mentioned deficiencies in the prior art, the transformer fault detection method based on nuclear capsule neuron coverage provided by the present invention takes the fault detection of transformers in power systems as the background, and aims at the absence, redundancy and Indefinite dimensions and other situations solve the problem of low recognition efficiency or even failure to correctly identify transformer fault detection due to the above situation

Images