Dielectric function gradient insulating dual-mode non-destructive testing method

Abstract

The invention discloses a dielectric function gradient insulating dual-mode non-destructive testing method which comprises the steps of performing segmentation according to the density change conditions at different positions in a three-dimensional insulator for obtaining the boundary of different density areas, then using geometric information as input, obtaining a training sample according to measurement of capacitance change on a boundary pole plate, performing training after normalization processing; introducing a boundary capacitance matrix of a functional gradient insulator into a trained model, obtaining the magnitude of a dielectric constant number in a corresponding material subarea in the insulator, and finishing non-destructive testing. According to the method, advantages of high spatial resolution of ICT and direct dielectric inversion of ECT are combined, thereby realizing a requirement for non-destructive testing in dielectric function gradient insulating shape and parameter fixation, improving inversion speed through a machine learning algorithm and settling problems of nonlinear and morbid state in a parameter inversion process.

Description

technical field [0001] The invention belongs to the technical field of non-destructive testing of electric equipment, and in particular relates to a dual-mode non-destructive testing method of dielectric functional gradient insulation. Background technique [0002] In my country, flashover / breakdown faults of cast epoxy insulators in GIS occur frequently, and the problem is prominent. In recent years, the emerging "Functionally Graded Material" (Functionally Graded Material, FGM) can significantly improve the insulator flashover by adjusting the spatial distribution of dielectric parameters (relative permittivity ε and conductivity γ) to improve the uniformity of the electric field. Voltage. [0003] At present, the techniques for preparing d-FGM mainly include centrifugal method, lamination method and 3D printing. Due to the integrated molding characteristics of the manufacturing process, it is necessary to judge whether the dielectric parameter distribution inside the in...

AI Technical Summary

Problems solved by technology

Obviously, the above method is difficult to operate and cannot meet the requirements of non-destructive identification

It should be emphasized that when the spatial distribution of inhomogeneous media is more complex, destructive detection methods no longer meet the test requirements

[0004] The current non-destructive testing methods mainly include ultrasonic testing, ray testing, magnetic particle testing, acoustic emission testing, microwave non-destructive testing and other technologies. Among them, the object of ultrasonic testing is acoustic impedance, which is

Images