Method for establishing frequency domain dielectric response characteristic fingerprint database based on depth fitting

Abstract

The invention relates to the technical field of high voltage and insulation, and particularly discloses a method for establishing a frequency domain dielectric response characteristic fingerprint database based on deep fitting, which comprises the following steps: acquiring a frequency domain dielectric spectrum of transformer solid insulation; respectively extracting frequency bands of frequencydomain dielectric response characteristics sensitive to transformer solid insulation aging and damp states; respectively carrying out integral operation on the extracted different frequency bands to obtain integral values of the FDS curve; obtaining a dielectric response characteristic fingerprint according to the integral value; enabling the dielectric response characteristic fingerprints to be in one-to-one correspondence with the corresponding insulation states, placing the dielectric response characteristic fingerprints and the corresponding insulation states in the same spatial dimensionto obtain a dielectric response characteristic fingerprint change rule, and performing deep fitting to establish a universal model capable of calculating the dielectric response characteristic fingerprints corresponding to any insulation state; and establishing a frequency domain dielectric response characteristic fingerprint database by utilizing the universal model. According to the method, thefrequency domain dielectric response characteristic fingerprint can be calculated by utilizing the model on the premise of not preparing any sample.

Description

technical field [0001] The invention belongs to the technical field of high voltage and insulation, in particular to a method for establishing a frequency-domain dielectric response characteristic fingerprint database based on depth fitting. Background technique [0002] As a large-scale key equipment in the power system, the quality of its solid insulation performance is related to the stable and safe operation of the entire power system. Therefore, it has very important theoretical and practical significance for the aging and damp state diagnosis of transformer solid insulation state. [0003] In recent years, transformer solid insulation state diagnosis method based on frequency domain dielectric response technology has attracted extensive attention because of its advantages of being sensitive to insulation state and more suitable for field testing. The study pointed out that the traditional decoupling analysis method based on frequency-domain dielectric spectrum (FDS) t...

AI Technical Summary

Problems solved by technology

However, the above methods are not only time-consuming and labor-intensive but also difficult to expand, not flexible and universal

And face the following disadvantages that limit its development and promotion

[0004] ① Affected by the operation level of personnel and the accuracy of sample preparation, it is very difficult to prepare a sample that can accurately represent the insulation state of different standards;

Images