CN-121982001-A - Insulation state evaluation method and system based on space charge characteristics

CN121982001ACN 121982001 ACN121982001 ACN 121982001ACN-121982001-A

Abstract

The invention relates to an insulation state evaluation method and system based on space charge characteristics, wherein the method comprises the steps of obtaining an original image of space charge density distribution, extracting a curve outline and identifying a curve category; and based on the local average change rate and the peak value characteristics, carrying out insulation state assessment in a weighted form by combining a constructed sample database, and obtaining an insulation state assessment result. According to the invention, by introducing an image recognition technology, key characteristic quantities in a space charge distribution diagram can be extracted for insulating material samples in different aging states, so that objective analysis and quantitative evaluation of space charge information are realized.

Inventors

LI HAI
WANG ZHIXIAO
SHI KAI
CUI JIABAO
WU JIANDONG
WANG SHUOYU
XIAO JUNYU
HE JIAWEI
ZHANG TIANYI
SUN WEISHA
YANG TIANYU
ZHANG XIN
LIU NING

Assignees

国网上海市电力公司
上海交通大学

Dates

Publication Date: 20260505
Application Date: 20260127

Claims (10)

1. An insulation state evaluation method based on space charge characteristics is characterized by comprising the following steps: Acquiring an original image of space charge density distribution, extracting a curve outline and identifying a curve category; Extracting local average change rate and peak value characteristics of the curve based on the curve profile and the curve category; and based on the local average change rate and the peak value characteristics, carrying out insulation state evaluation in a weighted form by combining the constructed sample database to obtain an insulation state evaluation result.
2. The space charge characteristic-based insulation state evaluation method according to claim 1, wherein the step of extracting a curve profile comprises: graying the original image with the space charge density distribution to obtain a gray image; denoising the gray level image to obtain a denoised gray level image; based on the denoised gray image, binarizing to obtain a binarized image; And extracting a curve outline from the binarized image.
3. The space charge feature based insulation state estimation method according to claim 1, wherein the step of extracting local average change rate and peak feature comprises: Dividing the extracted curve profile into different groups according to curve categories, and extracting pixel coordinates from each group; and determining the initial and final coordinate positions of each curve according to the pixel coordinates of each curve, and then performing curve fitting and numerical processing to obtain local average change rate and peak value characteristics.
4. A space charge characteristic based insulation state assessment method according to claim 3, wherein said curve fitting and numerical processing steps comprise: For each curve, adopting a gradient descent method to detect extreme points to obtain a maximum point and a minimum point which respectively correspond to a positive charge density peak value and a negative charge density peak value and serve as peak characteristics; Dividing the whole curve into n sections of local intervals, and calculating the local average change rate in each section.
5. The space charge characteristic-based insulation state evaluation method according to claim 4, wherein the positive charge density peak value and the negative charge density peak value acquiring steps: initializing, namely selecting a starting point x 0 and a learning rate eta, and bringing the following iterative formulas: , In the formula, For the calculated value of the next data point, Is a gradient; When gradient is formed Near 0, a minimum point P min is obtained as a negative charge density peak, where the negative charge density peak is: , Wherein x is the corresponding position and D (x) is a space charge density function; For the maximum point P max , as a positive charge density peak value, the iterative formula is: , When gradient is formed Obtaining a maximum point P max , wherein the positive charge density peak value is: , In the formula, As a function of space charge density.
6. The space charge characteristic based insulation state evaluation method according to claim 5, wherein the gradient Approximation by the center difference method: , where h is a small step size.
7. The space charge characteristic-based insulation state evaluation method according to claim 4, wherein the calculating of the local average change rate includes: Recording the charge density values and the corresponding geometric positions of the minimum value point P min and the maximum value point P max , and calculating the difference value of the charge density values and the corresponding geometric positions to be used as the thickness of a sample; And calculating the local average change rate of the n sections of local intervals according to the thickness and the resolution of the sample.
8. The insulation state evaluation method based on space charge characteristics according to claim 1, wherein the calculation expression of the local average change rate is: , In the formula, As a local average rate of change of the values, For the number of pixel points contained in each section, Is the ordinate value of the pixel point, namely the space charge density value.
9. The insulation state evaluation method based on space charge characteristics according to claim 1, wherein the operation expression for performing the weighted form of insulation state evaluation is: , In the formula, E 0,1, represents an insulating state, 、 For the weight, w 1 +w 2 =1, To the extent of the change in the charge density distribution, 、 As the characteristic quantity of the original sample, In order to calculate the average value of the slope, As a peak value of the density of positive charges, As a peak value of the negative charge density, For the number of data segments divided based on the measured data, Is the local average rate of change.
10. An insulation state evaluation system based on space charge characteristics, comprising: The curve extraction module is used for obtaining an original image of space charge density distribution, extracting a curve outline and identifying a curve category; The characteristic extraction module is used for extracting local average change rate and peak characteristic of the curve based on the curve profile and the curve category; And the evaluation module is used for carrying out insulation state evaluation in a weighted form based on the local average change rate and the peak value characteristics and combining the constructed sample database to obtain an insulation state evaluation result.

Description

Insulation state evaluation method and system based on space charge characteristics Technical Field The invention relates to the technical field of electric insulation state evaluation, in particular to an insulation state evaluation method and system based on space charge characteristics. Background The space charge detection technology is widely applied to insulating property analysis and state evaluation as an important means for evaluating the internal state of the insulating material. By monitoring the accumulation and distribution of space charges in the insulator, the aging degree of the material, the distortion condition of the electric field and potential insulation defects can be effectively reflected. However, the interpretation of the existing space charge detection results depends on experience and judgment of professionals, so that the subjectivity of the evaluation process is high, and the technical threshold is high. The non-professional can only obtain electric field distribution through space charge numerical integration, calculate electric field distortion rate (the difference value of the integrated electric field and the theoretical electric field accounts for the percentage of the theoretical electric field) and evaluate the quality of the insulating material. The method has the problem of single evaluation system, and ignores a large amount of insulation information hidden by space charge distribution data. In contrast, the partial discharge detection technology has gradually introduced image recognition and intelligent analysis methods in the field of insulating state evaluation, and by means of data-driven pattern recognition and feature extraction, the automation level and objectivity of state evaluation are effectively improved. However, the image recognition research for the space charge detection result is relatively deficient, and a systematic application method and an evaluation means are not yet available, so that the deep utilization and intelligent analysis of space charge image information are still in a starting stage, and innovative technical means are needed to scientifically and accurately evaluate space charge distribution data. In the prior art, patent application CN114019329a discloses a multidimensional evaluation system and method for early degradation of XLPE cable insulation based on machine learning, in which space charge is mentioned as a characteristic parameter for insulation degradation evaluation, but the measured space charge device is an integrating capacitor, only a single value of space accumulated charge is obtained, and space charge distribution of one-dimensional space inside the material is not obtained. Therefore, the evaluation index is single, the extracted information amount is small, and the correlation between the space charge distribution and the insulation state is difficult to deeply mine. Disclosure of Invention The invention aims to provide an insulation state evaluation method and system based on space charge characteristics, which improve the insulation state evaluation accuracy. The aim of the invention can be achieved by the following technical scheme: An insulation state evaluation method based on space charge characteristics comprises the following steps: Acquiring an original image of space charge density distribution, extracting a curve outline and identifying a curve category; Extracting local average change rate and peak value characteristics of the curve based on the curve profile and the curve category; and based on the local average change rate and the peak value characteristics, carrying out insulation state evaluation in a weighted form by combining the constructed sample database to obtain an insulation state evaluation result. Further, the step of extracting the curve profile includes: graying the original image with the space charge density distribution to obtain a gray image; denoising the gray level image to obtain a denoised gray level image; based on the denoised gray image, binarizing to obtain a binarized image; And extracting a curve outline from the binarized image. Further, the step of extracting the local average change rate and the peak feature includes: Dividing the extracted curve profile into different groups according to curve categories, and extracting pixel coordinates from each group; and determining the initial and final coordinate positions of each curve according to the pixel coordinates of each curve, and then performing curve fitting and numerical processing to obtain local average change rate and peak value characteristics. Further, the curve fitting and numerical processing steps include: For each curve, adopting a gradient descent method to detect extreme points to obtain a maximum point and a minimum point which respectively correspond to a positive charge density peak value and a negative charge density peak value and serve as peak characteristics; Dividing the whole curve into n sections of