A data monitoring method and system for AC insulation measurement test equipment

Abstract

The present invention relates to a data monitoring method and system for AC insulation measurement test equipment, which belongs to the field of AC insulation measurement technology. The method obtains the sequential real-time temperature chain of each insulation monitoring area block of the target insulation measurement test object, and analyzes the probability of occurrence of a three-dimensional thermal energy volume diagram under different local discharge modes based on the sequential real-time temperature chain; performs a local discharge test on the target insulation measurement test object according to the temperature monitoring result, and determines the current local discharge mode according to the spectrum amplitude distribution characteristics of the discharge signal of each insulation monitoring area block and the time-frequency density distribution characteristics of the discharge signal, generates an insulation defect layout diagram based on the monitoring positioning of the current local discharge mode, and analyzes the insulation health state of the target insulation measurement test object based on the insulation defect layout diagram. The present invention can accurately monitor the test data of the AC insulation measurement test equipment to ensure that more accurate insulation measurement data is obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of alternating current insulation measurement, and in particular to a data monitoring method and system for alternating current insulation measurement test equipment. Background Art

[0002] AC insulation test equipment is usually used to test the insulation performance of equipment or products in an AC environment. It usually includes components such as a high-voltage generator, a test fixture, and an insulation resistance meter, which can test and evaluate the insulation performance of the equipment to ensure that the equipment meets relevant standards and requirements in terms of electrical safety. However, the current AC insulation test equipment has a slow data monitoring acquisition rate during the measurement test of the insulation test object, and the accuracy of data monitoring is low, which is prone to errors, affecting the accurate analysis of the insulation health status, insulation status, and insulation quality, making it difficult to obtain more stable and reliable insulation measurement data; and the recognition rate and sensitivity of the partial discharge mode are low, resulting in errors in the partial discharge mode recognition results. Therefore, it is necessary to develop a method that helps to improve the data monitoring rate and accuracy of AC insulation measurement test equipment to solve the above problems in order to meet the needs of insulation measurement tests. Summary of the invention

[0003] The present invention overcomes the shortcomings of the prior art and provides a data monitoring method and system for AC insulation measurement and testing equipment.

[0004] To achieve the above object, the technical solution adopted by the present invention is:

[0005] A first aspect of the present invention provides a data monitoring method for AC insulation measurement test equipment, comprising the following steps:

[0006] An insulation test is performed on the target insulation measurement test object through an AC insulation measurement test device to obtain a sequential real-time temperature chain of each insulation monitoring area block, and a temperature storage model of each insulation monitoring area block is constructed. Based on the temperature storage model and the sequential real-time temperature chain, the probability of occurrence of a three-dimensional thermal energy volume diagram under different partial discharge modes is analyzed to obtain a temperature anomaly monitoring result;

[0007] Perform a partial discharge test on the target insulation measurement test object according to the temperature monitoring result to obtain the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal in each insulation monitoring area block, determine the current partial discharge mode by analyzing the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal, and generate an insulation defect layout map based on the monitoring and positioning of the current partial discharge mode;

[0008] Based on the insulation defect layout diagram, the insulation defects of the target insulation measurement test object are feature extracted to obtain the edge direction matrix of each insulation defect. The initial insulation defect heat map is drawn based on the distribution of the Manhattan distance between the edge direction matrix of each insulation defect and the edge direction normalized matrix to analyze the insulation health status of the target insulation measurement test object and obtain the insulation health monitoring results.

[0009] According to the insulation monitoring results, a breakdown test is carried out on the target insulation measurement test object, and a tree-like discharge model diagram corresponding to the target insulation measurement test object at each uniform time stamp is constructed. The adjacent tree-like discharge model diagrams are aligned in the model diagram alignment field and the alignment overlap rate is calculated. According to the alignment overlap rate, it is estimated whether the breakdown test produces a dielectric loss factor to generate the insulation quality monitoring result.

[0010] Furthermore, in a preferred embodiment of the present invention, the insulation test is performed on the target insulation measurement test object by the AC insulation measurement test equipment to obtain the time series real-time temperature chain of each insulation monitoring area block, and a temperature storage model of each insulation monitoring area block is constructed. Based on the temperature storage model and the time series real-time temperature chain, the probability of occurrence of the three-dimensional thermal energy volume under different partial discharge modes is analyzed to obtain the temperature anomaly monitoring result, which specifically includes the following steps:

[0011] Obtain model parameters of an AC insulation measurement test device and type information of a target insulation measurement test object, and obtain a safe measurement voltage threshold of the AC insulation measurement test device for the target insulation measurement test object in a big data network based on the model parameters and the type information;

[0012] Obtaining a circuit layout diagram inside a target insulation measurement test object, presetting and calibrating a plurality of insulation monitoring area blocks of the target insulation measurement test object according to the circuit layout diagram, and extracting the circuit layout density and insulation component setting points in each insulation monitoring area block;

[0013] A quadratic polynomial regression algorithm is introduced to calculate the quadratic polynomial coefficients between the circuit arrangement density and the insulation component setting points in each insulation monitoring area block to obtain a plurality of quadratic polynomial coefficients, and a heat concentration index of each insulation monitoring area block is determined according to the plurality of quadratic polynomial coefficients;

[0014] Based on the preset heat accumulation index, a temperature storage model for each insulation monitoring area block is constructed. At this time, the safety measurement voltage threshold is input into the AC insulation measurement test equipment for insulation measurement, and the sequential real-time temperature chain of each insulation monitoring area block is monitored by an infrared temperature detector;

[0015] Import the storage corresponding to each time series real-time temperature chain into the temperature storage model of each insulation monitoring area block to obtain the heat energy storage height of each temperature storage model. If the heat energy storage height is greater than the preset heat storage height, mark the temperature storage model corresponding to the heat storage height greater than the preset heat storage height, and define it as one or more high temperature storage models;

[0016] A naive Bayesian network of different partial discharge modes is constructed, and the thermal energy volume stored in one or more high-temperature storage models is stripped to form a three-dimensional thermal energy volume map. The probability of the three-dimensional thermal energy volume map appearing under different partial discharge modes is calculated in the naive Bayesian network. Based on the occurrence probability, whether the temperature performance of the target insulation measurement test object causes partial discharge is analyzed to obtain the temperature anomaly monitoring result.

[0017] Furthermore, in a preferred embodiment of the present invention, a partial discharge test is performed on the target insulation measurement test object according to the temperature monitoring result to obtain the discharge signal spectrum amplitude distribution characteristics and the discharge signal time-frequency density distribution characteristics of each insulation monitoring area block, and the current partial discharge mode is determined by analyzing the discharge signal spectrum amplitude distribution characteristics and the discharge signal time-frequency density distribution characteristics, and an insulation defect layout map is generated based on the monitoring and positioning of the current partial discharge mode, which specifically includes the following steps:

[0018] Analyze the temperature monitoring result by using AC insulation measurement test equipment, and if the temperature monitoring result shows that there is at least one result situation in which the probability of occurrence of the three-dimensional diagram of the thermal energy volume under different partial discharge modes is greater than a preset probability of occurrence, perform a partial discharge test on the target insulation measurement test object, and preset an AC voltage safety measurement threshold of the partial discharge test according to the probability of occurrence;

[0019] The AC voltage safety measurement threshold is input into the AC insulation measurement test equipment to perform insulation measurement. During the measurement process, the original discharge signal of each insulation monitoring area block is detected by an ultrasonic discharge point detector, and the original discharge signal of each insulation monitoring area block is denoised to be converted into a sinusoidal signal sequence;

[0020] The discrete Fourier transform algorithm is introduced. A rectangular window function is preset in the discrete Fourier transform algorithm based on the sinusoidal signal sequence of each insulation monitoring area block. The positively selected signal sequence is discretely transformed and output according to the rectangular window function to obtain the sinusoidal frequency domain signal of each insulation monitoring area block. The amplitude spectrum and phase of the sinusoidal frequency domain signal are calculated to extract the spectrum amplitude distribution characteristics of the discharge signal of each insulation monitoring area block.

[0021] The Hilbert transform method is introduced to calculate the complex amplitude of each sinusoidal signal sequence, and the real part and imaginary part of each complex amplitude are extracted. The envelope contour of the sinusoidal signal is analyzed based on the real part and the imaginary part, and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area are determined according to the envelope contour;

[0022] A partial discharge identification decision tree and a multi-dimensional partial discharge map are constructed based on the spectral amplitude distribution characteristics of the discharge signal of each insulation monitoring area block and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area. Based on the partial discharge identification decision tree, the multi-dimensional partial discharge map is identified and analyzed to obtain the current partial discharge mode of the target insulation measurement test object.

[0023] The standard spectral density grayscale value of the current partial discharge mode of the target insulation measurement test object is obtained. If the actual spectral density grayscale value displayed by the multi-dimensional partial discharge spectrum exceeds the standard spectral density grayscale value, the defect position of the target insulation measurement test object is monitored and located based on all phase resolutions, and an insulation defect layout map is generated and uploaded to the data acquisition terminal of the AC insulation measurement test equipment.

[0024] Furthermore, in a preferred embodiment of the present invention, the local discharge identification decision tree and the multidimensional local discharge map are constructed through the spectrum amplitude distribution characteristics of the discharge signal of each insulation monitoring area block and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area, and the multidimensional local discharge map is identified and analyzed based on the local discharge identification decision tree to obtain the current local discharge mode of the target insulation measurement test object, which specifically includes the following steps:

[0025] Obtain the information gain of the discharge signal spectrum amplitude distribution characteristics and the discharge signal time-frequency density distribution characteristics of each insulation monitoring area relative to the insulation monitoring area, and determine whether the information gain is greater than the preset information gain. If so, extract the discharge signal spectrum amplitude distribution characteristics and the discharge signal time-frequency density distribution characteristics corresponding to the information gain greater than the preset information gain;

[0026] The spectral amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal corresponding to the preset information gain are defined as segmentation characteristics, and each insulation monitoring area is recursively divided into N sub-identification areas by the segmentation characteristics until a preset recursive frequency is reached to obtain a partial discharge identification decision tree;

[0027] Constructing a multidimensional spectrum framework, importing the spectrum amplitude distribution characteristics of the discharge signal of each insulation monitoring area block and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area into the multidimensional spectrum framework for phase resolution description, and obtaining a multidimensional partial discharge spectrum;

[0028] Based on the multi-dimensional partial discharge spectrum, all phase resolutions and corresponding actual spectrum density grayscale values ​​of each insulation monitoring area block in a partial discharge cycle are obtained, which are defined as identification samples. Feature values ​​are taken according to the phase resolutions and corresponding spectrum density grayscale values ​​to obtain the feature value domain of the identification samples.

[0029] Based on the recognition sample feature value range, the entire partial discharge recognition decision tree is comprehensively traversed and recognized from each root node of the partial discharge recognition decision tree. If the recognition sample feature can be traversed and recognized to the branch end node or leaf node, the recognition action is stopped to output the partial discharge mode corresponding to the branch end node or leaf node, and the current partial discharge mode of the target insulation measurement test object is obtained.

[0030] Furthermore, in a preferred embodiment of the present invention, the feature extraction of the insulation defects of the target insulation measurement test object based on the insulation defect layout diagram is performed to obtain the edge direction matrix of each insulation defect, and the initial insulation defect heat map is drawn in combination with the distribution of the Manhattan distance between the edge direction matrix of each insulation defect and the edge direction normalized matrix, so as to analyze the insulation health state of the target insulation measurement test object and obtain the insulation health monitoring result, which specifically includes the following steps:

[0031] Using an industrial camera and based on the insulation defect layout diagram, the defect position of the target insulation measurement test object is photographed at a fixed point to obtain image data of each insulation defect in the insulation defect layout diagram;

[0032] An edge direction matrix algorithm is introduced. In the edge direction matrix algorithm, the image data of each insulation defect is converted into a grayscale image and Gaussian blur processing is performed. The Sobel operator is input to detect the grayscale image data after Gaussian blur processing to obtain the gradient and edge direction angle of the grayscale image data.

[0033] Preset a quantization direction interval of each insulation defect based on the gradient of the grayscale image data, quantize the edge direction angle to the quantization direction interval of each insulation defect, and then count the number of edge pixels in each edge direction in each quantization direction interval after quantization to generate an edge direction matrix for each insulation defect;

[0034] The discharge frequency and discharge intensity of each insulation defect in the current partial discharge mode are extracted by analyzing the multi-dimensional partial discharge map, and the edge direction normalized matrix of each insulation defect in the current partial discharge mode is obtained in the big data network based on the discharge frequency and discharge intensity;

[0035] Calculate the Manhattan distance between the edge direction matrix of each insulation defect and the corresponding edge direction normalized matrix, and obtain the Manhattan distance of the edge direction matrix that is relatively larger than the edge direction normalized matrix, which is defined as the positive Manhattan distance, and the Manhattan distance of the edge direction matrix that is relatively smaller than the edge direction normalized matrix, which is defined as the negative Manhattan distance;

[0036] A heat map framework is constructed based on the insulation defect layout map, and all positive Manhattan distances and all negative Manhattan distances are marked on the corresponding insulation defects in the heat map framework to obtain an initial insulation defect heat map;

[0037] Performing a health assessment on the discharge frequency and discharge intensity of each insulation defect through an insulation health status assessment model preset by an AC insulation measurement test device to obtain an insulation health status assessment value of each insulation defect, and generating a corresponding thermal region area in the initial insulation defect thermodynamic map based on the insulation health status assessment value;

[0038] If the ratio between the sum of the thermal area areas of all positive Manhattan distances and the sum of the thermal area areas of all negative Manhattan distances in the initial insulation defect thermogram is greater than the preset ratio, it means that the insulation health is good, otherwise the insulation health is poor, and the insulation health monitoring result is generated.

[0039] Further, in a preferred embodiment of the present invention, a breakdown test is performed on the target insulation measurement test object according to the insulation monitoring results, a tree-like discharge model diagram corresponding to each uniform time stamp of the target insulation measurement test object is constructed, adjacent tree-like discharge model diagrams are aligned in the model diagram alignment field and the alignment coincidence rate is calculated, and whether the breakdown test generates a dielectric loss factor is estimated according to the alignment coincidence rate to generate an insulation quality monitoring result, specifically including the following steps:

[0040] Analyzing the insulation health monitoring result, if the insulation health monitoring result shows that the insulation health state of the target insulation measurement test object is poor, disconnecting the target insulation measurement test object from the AC insulation measurement test equipment and marking it as unqualified;

[0041] If the insulation health monitoring result shows that the insulation health status of the target insulation measurement test object is good, then the AC voltage of the AC insulation measurement test equipment is planned according to the ratio shown by the insulation health monitoring result to reach the highest safety voltage critical value of the breakdown test, and the AC voltage adjustment range of the breakdown test is obtained;

[0042] Preset a number of uniform time stamps, continue to perform breakdown tests on target insulation measurement test objects with good insulation health status, and continuously adjust and increase the AC voltage of the AC insulation measurement test equipment according to the AC voltage adjustment interval;

[0043] During the process of continuously adjusting and increasing the AC voltage, the discharge signal strength and the discharge signal distribution of the target insulation measurement test object at each uniform time stamp are monitored by a capacitive coupling sensor, and a tree-like discharge model diagram corresponding to each uniform time stamp is constructed based on the discharge signal strength and the discharge signal distribution.

[0044] Constructing a model graph registration field, registering the tree-shaped discharge model graphs corresponding to adjacent uniform time stamps one by one in the model graph registration field, and calculating the registration coincidence rate between the tree-shaped discharge model graphs of adjacent uniform time stamps after the registration is completed, to obtain multiple registration coincidence rates;

[0045] The cubic spline interpolation method is introduced to construct N interpolation polynomial equation groups, and M interpolation functions are preset based on multiple registration coincidence rates. The M interpolation functions are imported into the N interpolation polynomial equation groups for solving, and the registration coincidence rate variation curve is generated;

[0046] A Bayesian network is constructed, and the registration coincidence rate change curve is imported into the Bayesian network for estimation to obtain several dielectric loss factor estimation values. If at least one dielectric loss factor estimation value greater than a preset dielectric loss factor estimation value is generated in the Bayesian network, it indicates that the target insulation measurement test object has poor compressive performance, and an insulation quality monitoring result is generated.

[0047] A second aspect of the present invention provides a data monitoring system for AC insulation measurement test equipment, the data monitoring system for AC insulation measurement test equipment comprising a memory and a processor, the memory storing a data monitoring method program for AC insulation measurement test equipment, and when the data monitoring method program for AC insulation measurement test equipment is executed by the processor, the following steps are implemented:

[0048] An insulation test is performed on the target insulation measurement test object through an AC insulation measurement test device to obtain a sequential real-time temperature chain of each insulation monitoring area block, and a temperature storage model of each insulation monitoring area block is constructed. Based on the temperature storage model and the sequential real-time temperature chain, the probability of occurrence of a three-dimensional thermal energy volume diagram under different partial discharge modes is analyzed to obtain a temperature anomaly monitoring result;

[0049] Perform a partial discharge test on the target insulation measurement test object according to the temperature monitoring result to obtain the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal in each insulation monitoring area block, determine the current partial discharge mode by analyzing the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal, and generate an insulation defect layout map based on the monitoring and positioning of the current partial discharge mode;

[0050] Based on the insulation defect layout diagram, the insulation defects of the target insulation measurement test object are feature extracted to obtain the edge direction matrix of each insulation defect. The initial insulation defect heat map is drawn based on the distribution of the Manhattan distance between the edge direction matrix of each insulation defect and the edge direction normalized matrix to analyze the insulation health status of the target insulation measurement test object and obtain the insulation health monitoring results.

[0051] According to the insulation monitoring results, a breakdown test is carried out on the target insulation measurement test object, and a tree-like discharge model diagram corresponding to the target insulation measurement test object at each uniform time stamp is constructed. The adjacent tree-like discharge model diagrams are aligned in the model diagram alignment field and the alignment overlap rate is calculated. According to the alignment overlap rate, it is estimated whether the breakdown test produces a dielectric loss factor to generate the insulation quality monitoring result.

[0052] The present invention solves the technical defects existing in the background technology, and the beneficial technical effects of the present invention are:

[0053] The timing real-time temperature chain of each insulation monitoring area block of the target insulation measurement test object is obtained, and the probability of occurrence of the three-dimensional graph of the thermal energy volume under different local discharge modes is analyzed based on the timing real-time temperature chain to obtain the temperature monitoring result; the local discharge test is performed on the target insulation measurement test object according to the temperature monitoring result, and the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area block are obtained to determine the current local discharge mode, and the insulation defect layout map is generated based on the monitoring positioning of the current local discharge mode; the distribution of the Manhattan distance between the edge direction matrix and the edge direction normalized matrix of the insulation defect is calculated based on the insulation defect layout map to draw the initial insulation defect thermal map to analyze the insulation health state of the target insulation measurement test object and obtain the insulation health monitoring result; the target insulation measurement test object is subjected to a breakdown test according to the insulation monitoring result, and a tree-like discharge model diagram corresponding to each uniform time stamp of the target insulation measurement test object is constructed, and the breakdown test is estimated by analyzing the tree-like discharge model diagram to generate the insulation breakdown monitoring result. The present invention can accurately monitor the test data of the AC insulation measurement test equipment to ensure that more accurate insulation measurement data is obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0055] Figure 1A first method flow chart of a data monitoring method for AC insulation measurement test equipment is shown;

[0056] Figure 2 A second method flow chart of a data monitoring method for AC insulation measurement test equipment is shown;

[0057] Figure 3 A third method flow chart of a data monitoring method for AC insulation measurement test equipment is shown;

[0058] Figure 4 A system framework diagram of a data monitoring system for AC insulation measurement and testing equipment is shown. DETAILED DESCRIPTION

[0059] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0060] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0061] The first aspect of the present invention provides a data monitoring method for AC insulation measurement test equipment, such as Figure 1 As shown, the following steps are included:

[0062] S102: performing an insulation test on a target insulation measurement test object through an AC insulation measurement test device to obtain a sequential real-time temperature chain of each insulation monitoring area block, and constructing a temperature storage model for each insulation monitoring area block; analyzing the probability of occurrence of a three-dimensional thermal energy volume diagram under different partial discharge modes based on the temperature storage model and the sequential real-time temperature chain, and obtaining a temperature anomaly monitoring result;

[0063] S104: performing a partial discharge test on the target insulation measurement test object according to the temperature monitoring result to obtain the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area block, determining the current partial discharge mode by analyzing the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal, and generating an insulation defect layout map based on the monitoring and positioning of the current partial discharge mode;

[0064] S106: extracting features of insulation defects of the target insulation measurement test object based on the insulation defect layout diagram, obtaining an edge direction matrix of each insulation defect, and drawing an initial insulation defect heat map based on the distribution of the Manhattan distance between the edge direction matrix of each insulation defect and the edge direction normalized matrix, so as to analyze the insulation health state of the target insulation measurement test object and obtain an insulation health monitoring result;

[0065] S108: Perform a breakdown test on the target insulation measurement test object according to the insulation monitoring results, construct a tree-like discharge model diagram corresponding to each uniform time stamp of the target insulation measurement test object, align adjacent tree-like discharge model diagrams in the model diagram alignment field and calculate the alignment overlap rate, and estimate whether the breakdown test generates a dielectric loss factor based on the alignment overlap rate to generate insulation quality monitoring results.

[0066] Furthermore, in a preferred embodiment of the present invention, the insulation test is performed on the target insulation measurement test object by the AC insulation measurement test equipment to obtain the time series real-time temperature chain of each insulation monitoring area block, and a temperature storage model of each insulation monitoring area block is constructed. Based on the temperature storage model and the time series real-time temperature chain, the probability of occurrence of the three-dimensional thermal energy volume under different partial discharge modes is analyzed to obtain the temperature anomaly monitoring result, which specifically includes the following steps:

[0067] Obtain model parameters of an AC insulation measurement test device and type information of a target insulation measurement test object, and obtain a safe measurement voltage threshold of the AC insulation measurement test device for the target insulation measurement test object in a big data network based on the model parameters and the type information;

[0068] Obtaining a circuit layout diagram inside a target insulation measurement test object, presetting and calibrating a plurality of insulation monitoring area blocks of the target insulation measurement test object according to the circuit layout diagram, and extracting the circuit layout density and insulation component setting points in each insulation monitoring area block;

[0069] A quadratic polynomial regression algorithm is introduced to calculate the quadratic polynomial coefficients between the circuit arrangement density and the insulation component setting points in each insulation monitoring area block to obtain a plurality of quadratic polynomial coefficients, and a heat concentration index of each insulation monitoring area block is determined according to the plurality of quadratic polynomial coefficients;

[0070] Based on the preset heat accumulation index, a temperature storage model for each insulation monitoring area block is constructed. At this time, the safety measurement voltage threshold is input into the AC insulation measurement test equipment for insulation measurement, and the sequential real-time temperature chain of each insulation monitoring area block is monitored by an infrared temperature detector;

[0071] Import the storage corresponding to each time series real-time temperature chain into the temperature storage model of each insulation monitoring area block to obtain the heat energy storage height of each temperature storage model. If the heat energy storage height is greater than the preset heat storage height, mark the temperature storage model corresponding to the heat storage height greater than the preset heat storage height, and define it as one or more high temperature storage models;

[0072] A naive Bayesian network of different partial discharge modes is constructed, and the thermal energy volume stored in one or more high-temperature storage models is stripped to form a three-dimensional thermal energy volume map. The probability of the three-dimensional thermal energy volume map appearing under different partial discharge modes is calculated in the naive Bayesian network. Based on the occurrence probability, whether the temperature performance of the target insulation measurement test object causes partial discharge is analyzed to obtain the temperature anomaly monitoring result.

[0073] It should be noted that AC insulation measurement test equipment usually measures the insulation of electrical equipment or electrical components by gradually increasing the size of the AC current. Since there are differences in the circuit distribution inside different electrical equipment or electrical components and the installation and setting points of the insulation component materials, if the insulation performance of the circuit-dense area is poor, it may cause overheating on the surface and inside of the electrical equipment or electrical components. The cause of the overheating may be partial discharge caused by abnormal insulation defects. Therefore, the present invention can divide the target insulation measurement test object into several insulation monitoring area blocks according to the circuit layout and the setting points of the insulation components, and then construct a temperature storage model by calculating the heat concentration index of each insulation monitoring area block. The model is used to store the time series real-time temperature chain generated by each insulation monitoring area block during the test process, wherein the time series real-time temperature chain is a temperature chain formed by continuous real-time temperatures generated during the insulation measurement test period to reflect the continuity of the real-time temperature; by analyzing the heat storage height of the time series real-time temperature chain in the temperature storage model, if the heat storage height exceeds the preset heat storage height, it means that the corresponding insulation monitoring area block has overheating, which is likely to be caused by partial discharge, so the heat volume in the temperature storage model of the insulation monitoring area block is finally formed into a three-dimensional heat volume map, and the probability of occurrence of this overheating phenomenon caused by this partial discharge is estimated in the naive Bayesian network and a comprehensive analysis is performed. The method can quickly and accurately analyze the overheating phenomenon inside the insulation test object measured by the AC insulation measurement test equipment, and further accurately judge the correlation with partial discharge, thereby improving the data monitoring efficiency and quality of the temperature performance of the insulation test object by the AC insulation measurement test equipment, and providing a reliable basis for the insulation fault analysis of the insulation test object.

[0074] Furthermore, in a preferred embodiment of the present invention, a partial discharge test is performed on the target insulation measurement test object according to the temperature monitoring result to obtain the discharge signal spectrum amplitude distribution characteristics and the discharge signal time-frequency density distribution characteristics of each insulation monitoring area block, and the current partial discharge mode is determined by analyzing the discharge signal spectrum amplitude distribution characteristics and the discharge signal time-frequency density distribution characteristics, and an insulation defect layout diagram is generated based on the monitoring positioning of the current partial discharge mode, such as Figure 2 As shown, the specific steps include:

[0075] S202: Analyze the temperature monitoring result by using AC insulation measurement test equipment. If the temperature monitoring result shows that there is at least one result in which the probability of occurrence of the three-dimensional diagram of the thermal energy volume in different partial discharge modes is greater than a preset probability of occurrence, perform a partial discharge test on the target insulation measurement test object, and preset an AC voltage safety measurement threshold of the partial discharge test according to the probability of occurrence;

[0076] S204: inputting an AC voltage safety measurement threshold into the AC insulation measurement test equipment to perform insulation measurement. During the measurement process, an ultrasonic discharge point detector is used to detect the original discharge signal of each insulation monitoring area block, and the original discharge signal of each insulation monitoring area block is denoised to be converted into a sinusoidal signal sequence.

[0077] S206: Introduce a discrete Fourier transform algorithm, preset a rectangular window function in the discrete Fourier transform algorithm based on the sinusoidal signal sequence of each insulation monitoring area block, output the discrete Fourier transform of the positively selected signal sequence according to the rectangular window function, obtain the sinusoidal frequency domain signal of each insulation monitoring area block, and calculate the amplitude spectrum and phase of the sinusoidal frequency domain signal to extract the spectrum amplitude distribution characteristics of the discharge signal of each insulation monitoring area block;

[0078] S208: Introducing the Hilbert transform method to calculate the complex amplitude of each sinusoidal signal sequence, extracting the real part and the imaginary part of each complex amplitude, analyzing the envelope contour of the sinusoidal signal based on the real part and the imaginary part, and determining the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area according to the envelope contour;

[0079] S210: constructing a partial discharge identification decision tree and a multi-dimensional partial discharge map based on the spectrum amplitude distribution characteristics of the discharge signal of each insulation monitoring area block and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area, and performing identification and analysis on the multi-dimensional partial discharge map based on the partial discharge identification decision tree to obtain the current partial discharge mode of the target insulation measurement test object;

[0080] S212: Obtain the standard spectral density grayscale value of the current partial discharge mode of the target insulation measurement test object. If the actual spectral density grayscale value displayed by the multi-dimensional partial discharge spectrum exceeds the standard spectral density grayscale value, monitor and locate the defect position of the target insulation measurement test object based on all phase resolutions, generate an insulation defect layout map and upload it to the data acquisition terminal of the AC insulation measurement test equipment.

[0081] It should be noted that if the temperature monitoring results show that the probability of occurrence is greater than the preset probability of occurrence, it means that the overheating phenomenon is caused by defects such as air gaps and cracks in the target insulation measurement test object itself, which causes local discharge during the insulation measurement test. Therefore, it is necessary to conduct a local discharge test on the target insulation measurement test object to further determine the specific location of the insulation defect, so as to facilitate the subsequent analysis of the insulation health status of the target insulation measurement test object. The method extracts the spectrum amplitude distribution characteristics and time-frequency density distribution characteristics of the original discharge signal of the partial discharge test in each insulation monitoring area block by introducing discrete Fourier transform algorithm and discrete Fourier transform algorithm respectively. According to these two distribution characteristics, a partial discharge recognition decision tree and a multi-dimensional partial discharge map can be further constructed. The multi-dimensional partial discharge map can quickly and accurately identify and locate the current partial discharge mode in the partial discharge recognition decision tree, which is more efficient, accurate and reliable than the traditional partial discharge mode recognition and positioning. The multi-dimensional partial discharge map is used to analyze whether the actual spectrum density grayscale value of the current partial discharge mode exceeds the standard spectrum density grayscale value. If it exceeds, it means that there is an insulation defect at the position corresponding to the actual spectrum density grayscale value, so as to draw an insulation defect layout map, which replaces the traditional manual step-by-step insulation defect judgment and positioning steps, saves manpower output costs, greatly improves the rate of insulation defect data monitoring, and has high economic benefits.

[0082] Furthermore, in a preferred embodiment of the present invention, a partial discharge identification decision tree and a multidimensional partial discharge map are constructed through the spectrum amplitude distribution characteristics of the discharge signal of each insulation monitoring area block and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area, and the multidimensional partial discharge map is identified and analyzed based on the partial discharge identification decision tree to obtain the current partial discharge mode of the target insulation measurement test object, such as Figure 3 As shown, the specific steps include:

[0083] S302: Obtain the information gain of the discharge signal spectrum amplitude distribution characteristics and the discharge signal time-frequency density distribution characteristics of each insulation monitoring area relative to the insulation monitoring area, and determine whether the information gain is greater than a preset information gain. If so, extract the discharge signal spectrum amplitude distribution characteristics and the discharge signal time-frequency density distribution characteristics corresponding to the information gain greater than the preset information gain;

[0084] S304: defining the spectral amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal corresponding to the preset information gain as segmentation characteristics, and recursively dividing each insulation monitoring area into N sub-identification areas by the segmentation characteristics until a preset recursive frequency is reached, thereby obtaining a partial discharge identification decision tree;

[0085] S306: construct a multidimensional spectrum framework, import the spectrum amplitude distribution characteristics of the discharge signal of each insulation monitoring area block and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area into the multidimensional spectrum framework for phase resolution description, and obtain a multidimensional partial discharge spectrum;

[0086] S308: Based on the multi-dimensional partial discharge spectrum, all phase resolutions and corresponding actual spectrum density grayscale values ​​of each insulation monitoring area block in a partial discharge cycle are obtained, which are defined as identification samples, and feature values ​​are taken according to the phase resolutions and corresponding spectrum density grayscale values ​​to obtain a feature value domain of the identification samples;

[0087] S310: Based on the recognition sample feature value domain, the entire local discharge recognition decision tree is comprehensively traversed and recognized from each root node of the local discharge recognition decision tree. If the recognition sample feature can be traversed and recognized to the branch end node or leaf node, the recognition action is stopped to output the local discharge mode corresponding to the branch end node or leaf node to obtain the current local discharge mode of the target insulation measurement test object.

[0088] It should be noted that the traditional determination of local discharge patterns is usually to directly construct and analyze local discharge maps. Although this method has a fast calculation and analysis rate, the display results of the local discharge maps may often have errors, which also leads to certain deviations in the final analysis of local discharge signals, thus causing errors in the local discharge pattern recognition results. In addition, traditional local discharge maps usually use two-dimensional maps such as phase distribution maps or amplitude-time maps. The discharge signal trend analysis and data display of two-dimensional maps often have certain limitations, making it difficult to intuitively reflect the discharge situation. Therefore, this method first calculates the information gain based on the spectrum amplitude distribution characteristics and the time-frequency density distribution characteristics. The information gain can reflect and screen out the best distribution characteristics as the segmentation features for constructing the decision tree, and construct the local discharge recognition decision tree through the segmentation features; then the key to constructing a multidimensional local discharge map is to calculate the phase resolution of the distribution characteristics, so as to combine the phase, amplitude and number of discharges to display the overall characteristics of the local discharge in three-dimensional form, which can more intuitively reflect the discharge situation; finally, based on all the phase resolutions of each insulation monitoring area block displayed in the multidimensional local discharge map in a local discharge cycle and the corresponding actual spectrum density grayscale values ​​as identification samples, the corresponding classified local discharge mode is traversed in the local discharge recognition decision tree to improve the accuracy of local discharge mode recognition, with fast recognition rate and high sensitivity, and can provide rich data monitoring information to support in-depth discharge mechanism research and fault diagnosis.

[0089] Furthermore, in a preferred embodiment of the present invention, the feature extraction of the insulation defects of the target insulation measurement test object based on the insulation defect layout diagram is performed to obtain the edge direction matrix of each insulation defect, and the initial insulation defect heat map is drawn in combination with the distribution of the Manhattan distance between the edge direction matrix of each insulation defect and the edge direction normalized matrix, so as to analyze the insulation health state of the target insulation measurement test object and obtain the insulation health monitoring result, which specifically includes the following steps:

[0090] Using an industrial camera and based on the insulation defect layout diagram, the defect position of the target insulation measurement test object is photographed at a fixed point to obtain image data of each insulation defect in the insulation defect layout diagram;

[0091] An edge direction matrix algorithm is introduced. In the edge direction matrix algorithm, the image data of each insulation defect is converted into a grayscale image and Gaussian blur processing is performed. The Sobel operator is input to detect the grayscale image data after Gaussian blur processing to obtain the gradient and edge direction angle of the grayscale image data.

[0092] Preset a quantization direction interval of each insulation defect based on the gradient of the grayscale image data, quantize the edge direction angle to the quantization direction interval of each insulation defect, and then count the number of edge pixels in each edge direction in each quantization direction interval after quantization to generate an edge direction matrix for each insulation defect;

[0093] The discharge frequency and discharge intensity of each insulation defect in the current partial discharge mode are extracted by analyzing the multi-dimensional partial discharge map, and the edge direction normalized matrix of each insulation defect in the current partial discharge mode is obtained in the big data network based on the discharge frequency and discharge intensity;

[0094] Calculate the Manhattan distance between the edge direction matrix of each insulation defect and the corresponding edge direction normalized matrix, and obtain the Manhattan distance of the edge direction matrix that is relatively larger than the edge direction normalized matrix, which is defined as the positive Manhattan distance, and the Manhattan distance of the edge direction matrix that is relatively smaller than the edge direction normalized matrix, which is defined as the negative Manhattan distance;

[0095] A heat map framework is constructed based on the insulation defect layout map, and all positive Manhattan distances and all negative Manhattan distances are marked on the corresponding insulation defects in the heat map framework to obtain an initial insulation defect heat map;

[0096] Performing a health assessment on the discharge frequency and discharge intensity of each insulation defect through an insulation health status assessment model preset by an AC insulation measurement test device to obtain an insulation health status assessment value of each insulation defect, and generating a corresponding thermal region area in the initial insulation defect thermodynamic map based on the insulation health status assessment value;

[0097] If the ratio between the sum of the thermal area areas of all positive Manhattan distances and the sum of the thermal area areas of all negative Manhattan distances in the initial insulation defect thermogram is greater than the preset ratio, it means that the insulation health is good, otherwise the insulation health is poor, and the insulation health monitoring result is generated.

[0098] It should be noted that after determining the insulation defect layout diagram of the target insulation measurement test object, the insulation health status can be further analyzed through the insulation defect layout diagram. However, since some insulation defects may be larger and some smaller, and insulation defects of different sizes will produce different insulation performance, if the insulation performance of the current target insulation measurement test object is to be accurately determined, the characteristics of the insulation defects need to be analyzed. Therefore, this method extracts the image features of each insulation defect by introducing the edge direction matrix algorithm, and generates the edge direction matrix of each insulation defect. Since the edge direction matrix of each insulation defect may be larger or smaller than the corresponding edge direction standardized matrix, the edge direction standardized matrix is ​​the ideal insulation defect image feature that meets the insulation health defect range of each insulation defect in the current local discharge mode, that is, the standard edge direction matrix; if the edge direction matrix is ​​larger than the edge direction standardized matrix, it means that the existence of the insulation defect does not meet the insulation health defect requirements, reduces the insulation performance of the target insulation measurement test object, and causes the phenomenon of being unable to insulate alternating current; otherwise, it meets the insulation health defect requirements and will not damage the insulation performance of the target insulation measurement test object. Therefore, this method constructs an initial insulation defect heat map by the Manhattan distance between the edge direction matrix and the edge direction standardized matrix to analyze whether the existence of the insulation defect meets the insulation health defect requirements, thereby improving the efficiency of insulation health status data monitoring, ensuring the accuracy of the AC insulation measurement test equipment in monitoring the insulation defect health and the quality of insulation performance judgment, and has high reliability.

[0099] Further, in a preferred embodiment of the present invention, a breakdown test is performed on the target insulation measurement test object according to the insulation monitoring results, a tree-like discharge model diagram corresponding to each uniform time stamp of the target insulation measurement test object is constructed, adjacent tree-like discharge model diagrams are aligned in the model diagram alignment field and the alignment coincidence rate is calculated, and whether the breakdown test generates a dielectric loss factor is estimated according to the alignment coincidence rate to generate an insulation quality monitoring result, specifically including the following steps:

[0100] Analyzing the insulation health monitoring result, if the insulation health monitoring result shows that the insulation health state of the target insulation measurement test object is poor, disconnecting the target insulation measurement test object from the AC insulation measurement test equipment and marking it as unqualified;

[0101] If the insulation health monitoring result shows that the insulation health status of the target insulation measurement test object is good, then the AC voltage of the AC insulation measurement test equipment is planned according to the ratio shown by the insulation health monitoring result to reach the highest safety voltage critical value of the breakdown test, and the AC voltage adjustment range of the breakdown test is obtained;

[0102] Preset a number of uniform time stamps, continue to perform breakdown tests on target insulation measurement test objects with good insulation health status, and continuously adjust and increase the AC voltage of the AC insulation measurement test equipment according to the AC voltage adjustment interval;

[0103] During the process of continuously adjusting and increasing the AC voltage, the discharge signal strength and the discharge signal distribution of the target insulation measurement test object at each uniform time stamp are monitored by a capacitive coupling sensor, and a tree-like discharge model diagram corresponding to each uniform time stamp is constructed based on the discharge signal strength and the discharge signal distribution.

[0104] Constructing a model graph registration field, registering the tree-shaped discharge model graphs corresponding to adjacent uniform time stamps one by one in the model graph registration field, and calculating the registration coincidence rate between the tree-shaped discharge model graphs of adjacent uniform time stamps after the registration is completed, to obtain multiple registration coincidence rates;

[0105] The cubic spline interpolation method is introduced to construct N interpolation polynomial equation groups, and M interpolation functions are preset based on multiple registration coincidence rates. The M interpolation functions are imported into the N interpolation polynomial equation groups for solving, and the registration coincidence rate variation curve is generated;

[0106] A Bayesian network is constructed, and the registration coincidence rate change curve is imported into the Bayesian network for estimation to obtain several dielectric loss factor estimation values. If at least one dielectric loss factor estimation value greater than a preset dielectric loss factor estimation value is generated in the Bayesian network, it indicates that the target insulation measurement test object has poor compressive performance, and an insulation quality monitoring result is generated.

[0107] It should be noted that the insulation health status of the target insulation measurement test object can be analyzed through the insulation health monitoring results. If the insulation health monitoring results show that the insulation health status of the target insulation measurement test object is poor, it may simultaneously have a certain impact on the compressive performance of the target insulation measurement test object, resulting in the occurrence of breakdown phenomenon, and the breakdown phenomenon is often one of the indicators for evaluating the insulation quality of the target insulation measurement test object. If the breakdown phenomenon occurs, it can be proved that the insulation quality of the target insulation measurement test object is poor; therefore, the method can perform a breakdown test on the target insulation measurement test object according to the insulation health monitoring results, and The discharge signal intensity and the discharge signal distribution at the uniform time stamp are fitted to construct a tree-like discharge model diagram. The tree-like discharge model diagram is a precursor of electrical breakdown. By analyzing the distribution change of the tree-like discharge model diagram over time, the insulation quality performance of the target insulation measurement test object in the breakdown test can be determined, that is, by estimating the dielectric loss factor of the target insulation measurement test object that is broken down under the tree-like discharge model diagram, if at least one dielectric loss factor estimate greater than the preset dielectric loss factor estimate is generated in the Bayesian network, it means that the dielectric loss generated when the target insulation measurement test object is broken down is large, so the compressive performance and insulation quality of the target insulation measurement test object are poor. This method can monitor the compressive performance and insulation quality of the target insulation measurement test object through the breakdown test of the AC insulation measurement test equipment, further improving the efficient and accurate monitoring of the insulation quality of the target insulation measurement test object by the AC insulation measurement test equipment, and ensuring that the user can quickly obtain and fully know the insulation quality data of the insulation test object.

[0108] In addition, the data monitoring method of the AC insulation measurement test equipment further includes the following steps:

[0109] Obtaining measurement abnormality evaluation results of AC insulation measurement test equipment in a preset time period, and obtaining several environmental factors that cause abnormalities in the AC insulation measurement test equipment based on a big data network;

[0110] Introducing a factor analysis algorithm to calculate the factor correlation between the measurement abnormality evaluation result and each environmental factor, if the factor correlation is greater than a preset factor correlation, then the environmental factor corresponding to the factor correlation greater than the preset factor correlation is defined as an abnormal correlation environmental factor;

[0111] Obtain historical insulation measurement test data generated by each sub-measurement mechanism in the AC insulation measurement test equipment under different abnormal associated environmental factors, build a prediction model based on a deep learning network, import the historical insulation measurement test data generated by each sub-measurement mechanism under different abnormal associated environmental factors into the prediction model for training, and obtain a trained prediction model;

[0112] Acquire the actual abnormal associated environmental factors of the AC insulation measurement test equipment, and import the actual abnormal associated environmental factors into the trained prediction model to output insulation measurement test prediction data of each sub-measurement mechanism in the AC insulation measurement test equipment under the actual abnormal associated environmental factors;

[0113] Acquire actual insulation measurement test data of each sub-measurement mechanism of the AC insulation measurement test equipment within a preset time period, introduce a hash algorithm to calculate a hash function between the actual insulation measurement test data and the insulation measurement test prediction data;

[0114] If the hash function is greater than a preset hash function, the sub-measurement mechanism corresponding to the hash function greater than the preset hash function is marked as a faulty mechanism and uploaded to the maintenance log of the AC insulation measurement test equipment.

[0115] It should be noted that the internal structure of AC insulation measurement and test equipment is usually composed of multiple measuring mechanisms, and as the operating time increases, the internal measuring mechanisms will experience wear, aging, or short circuit failures. Traditionally, the failure phenomena of the measuring mechanisms of AC insulation measurement and test equipment need to be determined by manual inspection and analysis, which is slow, time-consuming, and wastes a lot of manpower and material resources, which is not conducive to maintenance personnel's maintenance of AC insulation measurement and test equipment. The method can construct a prediction model through the insulation measurement test data of each sub-measurement mechanism in the AC insulation measurement test equipment under abnormal associated environmental factors, and estimate the insulation measurement test prediction data of each sub-measurement mechanism in the AC insulation measurement test equipment under actual abnormal associated environmental factors with the help of the prediction model. Finally, by calculating the hash function between the insulation measurement test prediction data and the actual insulation measurement test data and determining whether the hash function is greater than a preset hash function, if it is greater, it means that the actual insulation measurement test data of the sub-measurement mechanism has a large deviation from the insulation measurement test prediction data due to fault phenomena such as wear, aging or short circuit, and the sub-measurement mechanism is marked as a faulty mechanism; the speed and accuracy of fault location are improved, the cumbersome steps of traditional manual fault location are replaced, time and labor are saved, and the intelligent automation level of fault location is greatly improved.

[0116] A second aspect of the present invention provides a data monitoring system for an AC insulation measurement test device, the data monitoring system for the AC insulation measurement test device comprising a memory 41 and a processor 42, the memory 41 storing a data monitoring method program for the AC insulation measurement test device, and when the data monitoring method program for the AC insulation measurement test device is executed by the processor 42, Figure 4 As shown, implement the following steps:

[0117] An insulation test is performed on the target insulation measurement test object through an AC insulation measurement test device to obtain a sequential real-time temperature chain of each insulation monitoring area block, and a temperature storage model of each insulation monitoring area block is constructed. Based on the temperature storage model and the sequential real-time temperature chain, the probability of occurrence of a three-dimensional thermal energy volume diagram under different partial discharge modes is analyzed to obtain a temperature anomaly monitoring result;

[0118] Perform a partial discharge test on the target insulation measurement test object according to the temperature monitoring result to obtain the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal in each insulation monitoring area block, determine the current partial discharge mode by analyzing the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal, and generate an insulation defect layout map based on the monitoring and positioning of the current partial discharge mode;

[0119] Based on the insulation defect layout diagram, the insulation defects of the target insulation measurement test object are feature extracted to obtain the edge direction matrix of each insulation defect. The initial insulation defect heat map is drawn based on the distribution of the Manhattan distance between the edge direction matrix of each insulation defect and the edge direction normalized matrix to analyze the insulation health status of the target insulation measurement test object and obtain the insulation health monitoring results.

[0120] According to the insulation monitoring results, a breakdown test is carried out on the target insulation measurement test object, and a tree-like discharge model diagram corresponding to the target insulation measurement test object at each uniform time stamp is constructed. The adjacent tree-like discharge model diagrams are aligned in the model diagram alignment field and the alignment overlap rate is calculated. According to the alignment overlap rate, it is estimated whether the breakdown test produces a dielectric loss factor to generate the insulation breakdown monitoring result.

[0121] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A data monitoring method for AC insulation measurement test equipment, characterized in that: The following steps are involved: An insulation test is performed on the target insulation measurement test object through an AC insulation measurement test device to obtain a sequential real-time temperature chain of each insulation monitoring area block, and a temperature storage model of each insulation monitoring area block is constructed. Based on the temperature storage model and the sequential real-time temperature chain, the probability of occurrence of a three-dimensional thermal energy volume diagram under different partial discharge modes is analyzed to obtain a temperature anomaly monitoring result; Perform a partial discharge test on the target insulation measurement test object according to the temperature monitoring result to obtain the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal in each insulation monitoring area block, determine the current partial discharge mode by analyzing the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal, and generate an insulation defect layout map based on the monitoring and positioning of the current partial discharge mode; Based on the insulation defect layout diagram, the insulation defects of the target insulation measurement test object are feature extracted to obtain the edge direction matrix of each insulation defect. The initial insulation defect heat map is drawn based on the distribution of the Manhattan distance between the edge direction matrix of each insulation defect and the edge direction normalized matrix to analyze the insulation health status of the target insulation measurement test object and obtain the insulation health monitoring results. According to the insulation monitoring results, a breakdown test is performed on the target insulation measurement test object, and a tree-like discharge model diagram corresponding to each uniform time stamp of the target insulation measurement test object is constructed. Adjacent tree-like discharge model diagrams are registered in the model diagram registration field and the registration coincidence rate is calculated. According to the registration coincidence rate, it is estimated whether the breakdown test generates a dielectric loss factor, so as to generate insulation quality monitoring results. Among them, based on the insulation defect layout diagram, the insulation defects of the target insulation measurement test object are feature extracted to obtain the edge direction matrix of each insulation defect, and the initial insulation defect heat map is drawn in combination with the distribution of the Manhattan distance between the edge direction matrix of each insulation defect and the edge direction normalized matrix, so as to analyze the insulation health status of the target insulation measurement test object and obtain the insulation health monitoring result, which specifically includes the following steps: Using an industrial camera and based on the insulation defect layout diagram, the defect position of the target insulation measurement test object is photographed at a fixed point to obtain image data of each insulation defect in the insulation defect layout diagram; An edge direction matrix algorithm is introduced. In the edge direction matrix algorithm, the image data of each insulation defect is converted into a grayscale image and Gaussian blur processing is performed. The Sobel operator is input to detect the grayscale image data after Gaussian blur processing to obtain the gradient and edge direction angle of the grayscale image data. Preset a quantization direction interval of each insulation defect based on the gradient of the grayscale image data, quantize the edge direction angle to the quantization direction interval of each insulation defect, and then count the number of edge pixels in each edge direction in each quantization direction interval after quantization to generate an edge direction matrix for each insulation defect; The discharge frequency and discharge intensity of each insulation defect in the current partial discharge mode are extracted by analyzing the multi-dimensional partial discharge map, and the edge direction normalized matrix of each insulation defect in the current partial discharge mode is obtained in the big data network based on the discharge frequency and discharge intensity; Calculate the Manhattan distance between the edge direction matrix of each insulation defect and the corresponding edge direction normalized matrix, and obtain the Manhattan distance of the edge direction matrix that is relatively larger than the edge direction normalized matrix, which is defined as the positive Manhattan distance, and the Manhattan distance of the edge direction matrix that is relatively smaller than the edge direction normalized matrix, which is defined as the negative Manhattan distance; A heat map framework is constructed based on the insulation defect layout map, and all positive Manhattan distances and all negative Manhattan distances are marked on the corresponding insulation defects in the heat map framework to obtain an initial insulation defect heat map; Performing a health assessment on the discharge frequency and discharge intensity of each insulation defect through an insulation health status assessment model preset by an AC insulation measurement test device to obtain an insulation health status assessment value of each insulation defect, and generating a corresponding thermal region area in the initial insulation defect thermodynamic map based on the insulation health status assessment value; If the ratio between the sum of the thermal area areas of all positive Manhattan distances and the sum of the thermal area areas of all negative Manhattan distances in the initial insulation defect thermogram is greater than the preset ratio, it means that the insulation health is good, otherwise the insulation health is poor, and the insulation health monitoring result is generated.

2. The data monitoring method for AC insulation measurement test equipment according to claim 1 is characterized in that: The method of performing an insulation test on a target insulation measurement test object by using an AC insulation measurement test device to obtain a sequential real-time temperature chain of each insulation monitoring area block, and constructing a temperature storage model for each insulation monitoring area block, and analyzing the probability of occurrence of a three-dimensional thermal energy volume graph under different partial discharge modes based on the temperature storage model combined with the sequential real-time temperature chain to obtain a temperature anomaly monitoring result, specifically includes the following steps: Acquire model parameters of an AC insulation measurement test device and type information of a target insulation measurement test object, and acquire a safe measurement voltage threshold of the AC insulation measurement test device for the target insulation measurement test object in a big data network based on the model parameters and the type information; Obtaining a circuit layout diagram inside a target insulation measurement test object, presetting and calibrating a plurality of insulation monitoring area blocks of the target insulation measurement test object according to the circuit layout diagram, and extracting the circuit layout density and insulation component setting points in each insulation monitoring area block; A quadratic polynomial regression algorithm is introduced to calculate the quadratic polynomial coefficients between the circuit arrangement density and the insulation component setting points in each insulation monitoring area block to obtain a plurality of quadratic polynomial coefficients, and a heat concentration index of each insulation monitoring area block is determined according to the plurality of quadratic polynomial coefficients; Based on the preset heat accumulation index, a temperature storage model for each insulation monitoring area block is constructed. At this time, the safety measurement voltage threshold is input into the AC insulation measurement test equipment for insulation measurement, and the sequential real-time temperature chain of each insulation monitoring area block is monitored by an infrared temperature detector; Import the storage corresponding to each time series real-time temperature chain into the temperature storage model of each insulation monitoring area block to obtain the heat energy storage height of each temperature storage model. If the heat energy storage height is greater than the preset heat storage height, mark the temperature storage model corresponding to the heat storage height greater than the preset heat storage height, and define it as one or more high temperature storage models; A naive Bayesian network of different partial discharge modes is constructed, and the thermal energy volume stored in one or more high-temperature storage models is stripped to form a three-dimensional thermal energy volume map. The probability of the three-dimensional thermal energy volume map appearing under different partial discharge modes is calculated in the naive Bayesian network. Based on the occurrence probability, whether the temperature performance of the target insulation measurement test object causes partial discharge is analyzed to obtain the temperature anomaly monitoring result.

3. The data monitoring method for AC insulation measurement test equipment according to claim 1, characterized in that: The method of performing a partial discharge test on the target insulation measurement test object according to the temperature monitoring result to obtain the discharge signal spectrum amplitude distribution characteristics and the discharge signal time-frequency density distribution characteristics of each insulation monitoring area block, determining the current partial discharge mode by analyzing the discharge signal spectrum amplitude distribution characteristics and the discharge signal time-frequency density distribution characteristics, and generating an insulation defect layout map based on the monitoring and positioning of the current partial discharge mode, specifically includes the following steps: Analyze the temperature monitoring result by using AC insulation measurement test equipment, and if the temperature monitoring result shows that there is at least one result situation in which the probability of occurrence of the three-dimensional diagram of the thermal energy volume under different partial discharge modes is greater than a preset probability of occurrence, perform a partial discharge test on the target insulation measurement test object, and preset an AC voltage safety measurement threshold of the partial discharge test according to the probability of occurrence; The AC voltage safety measurement threshold is input into the AC insulation measurement test equipment to perform insulation measurement. During the measurement process, the original discharge signal of each insulation monitoring area block is detected by an ultrasonic discharge point detector, and the original discharge signal of each insulation monitoring area block is denoised to be converted into a sinusoidal signal sequence; The discrete Fourier transform algorithm is introduced. A rectangular window function is preset in the discrete Fourier transform algorithm based on the sinusoidal signal sequence of each insulation monitoring area block. The positively selected signal sequence is discretely transformed and output according to the rectangular window function to obtain the sinusoidal frequency domain signal of each insulation monitoring area block. The amplitude spectrum and phase of the sinusoidal frequency domain signal are calculated to extract the spectrum amplitude distribution characteristics of the discharge signal of each insulation monitoring area block. The Hilbert transform method is introduced to calculate the complex amplitude of each sinusoidal signal sequence, and the real part and imaginary part of each complex amplitude are extracted. The envelope contour of the sinusoidal signal is analyzed based on the real part and the imaginary part, and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area are determined according to the envelope contour; A partial discharge identification decision tree and a multi-dimensional partial discharge map are constructed based on the spectral amplitude distribution characteristics of the discharge signal of each insulation monitoring area block and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area. Based on the partial discharge identification decision tree, the multi-dimensional partial discharge map is identified and analyzed to obtain the current partial discharge mode of the target insulation measurement test object. The standard spectral density grayscale value of the current partial discharge mode of the target insulation measurement test object is obtained. If the actual spectral density grayscale value displayed by the multi-dimensional partial discharge spectrum exceeds the standard spectral density grayscale value, the defect position of the target insulation measurement test object is monitored and located based on all phase resolutions, and an insulation defect layout map is generated and uploaded to the data acquisition terminal of the AC insulation measurement test equipment.

4. The data monitoring method for AC insulation measurement test equipment according to claim 3 is characterized in that: The method constructs a partial discharge identification decision tree and a multidimensional partial discharge map through the spectrum amplitude distribution characteristics of the discharge signal of each insulation monitoring area block and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area, and identifies and analyzes the multidimensional partial discharge map based on the partial discharge identification decision tree to obtain the current partial discharge mode of the target insulation measurement test object, specifically including the following steps: Obtain the information gain of the discharge signal spectrum amplitude distribution characteristics and the discharge signal time-frequency density distribution characteristics of each insulation monitoring area relative to the insulation monitoring area, and determine whether the information gain is greater than the preset information gain. If so, extract the discharge signal spectrum amplitude distribution characteristics and the discharge signal time-frequency density distribution characteristics corresponding to the information gain greater than the preset information gain; The spectral amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal corresponding to the preset information gain are defined as segmentation characteristics, and each insulation monitoring area is recursively divided into N sub-identification areas by the segmentation characteristics until a preset recursive frequency is reached to obtain a partial discharge identification decision tree; Constructing a multidimensional spectrum framework, importing the spectrum amplitude distribution characteristics of the discharge signal of each insulation monitoring area block and the time-frequency density distribution characteristics of the discharge signal of each insulation monitoring area into the multidimensional spectrum framework for phase resolution description, and obtaining a multidimensional partial discharge spectrum; Based on the multi-dimensional partial discharge spectrum, all phase resolutions and corresponding actual spectrum density grayscale values ​​of each insulation monitoring area block in a partial discharge cycle are obtained, which are defined as identification samples. Feature values ​​are taken according to the phase resolutions and corresponding spectrum density grayscale values ​​to obtain the feature value domain of the identification samples. Based on the recognition sample feature value range, the entire partial discharge recognition decision tree is comprehensively traversed and recognized from each root node of the partial discharge recognition decision tree. If the recognition sample feature can be traversed and recognized to the branch end node or leaf node, the recognition action is stopped to output the partial discharge mode corresponding to the branch end node or leaf node, and the current partial discharge mode of the target insulation measurement test object is obtained.

5. The data monitoring method for AC insulation measurement and testing equipment according to claim 1, characterized in that: The method of performing a breakdown test on a target insulation measurement test object according to the insulation monitoring result, constructing a tree-like discharge model diagram corresponding to each uniform time stamp of the target insulation measurement test object, registering adjacent tree-like discharge model diagrams in a model diagram registration field and calculating the registration coincidence rate, and estimating whether the breakdown test generates a dielectric loss factor according to the registration coincidence rate to generate an insulation quality monitoring result, specifically includes the following steps: Analyzing the insulation health monitoring result, if the insulation health monitoring result shows that the insulation health state of the target insulation measurement test object is poor, disconnecting the target insulation measurement test object from the AC insulation measurement test equipment and marking it as unqualified; If the insulation health monitoring result shows that the insulation health status of the target insulation measurement test object is good, then the AC voltage of the AC insulation measurement test equipment is planned according to the ratio shown by the insulation health monitoring result to reach the highest safety voltage critical value of the breakdown test, and the AC voltage adjustment range of the breakdown test is obtained; Preset a number of uniform time stamps, continue to perform breakdown tests on target insulation measurement test objects with good insulation health status, and continuously adjust and increase the AC voltage of the AC insulation measurement test equipment according to the AC voltage adjustment interval; During the process of continuously adjusting and increasing the AC voltage, the discharge signal strength and the discharge signal distribution of the target insulation measurement test object at each uniform time stamp are monitored by a capacitive coupling sensor, and a tree-like discharge model diagram corresponding to each uniform time stamp is constructed based on the discharge signal strength and the discharge signal distribution. Constructing a model graph registration field, registering the tree-shaped discharge model graphs corresponding to adjacent uniform time stamps one by one in the model graph registration field, and calculating the registration coincidence rate between the tree-shaped discharge model graphs of adjacent uniform time stamps after the registration is completed, to obtain multiple registration coincidence rates; The cubic spline interpolation method is introduced to construct N interpolation polynomial equation groups, and M interpolation functions are preset based on multiple registration coincidence rates. The M interpolation functions are imported into the N interpolation polynomial equation groups for solving, and the registration coincidence rate variation curve is generated; A Bayesian network is constructed, and the registration coincidence rate change curve is imported into the Bayesian network for estimation to obtain several dielectric loss factor estimation values. If at least one dielectric loss factor estimation value greater than a preset dielectric loss factor estimation value is generated in the Bayesian network, it indicates that the target insulation measurement test object has poor compressive performance, and an insulation quality monitoring result is generated.

6. A data monitoring system for AC insulation measurement and testing equipment, characterized in that: The data monitoring system of an AC insulation measurement test device comprises a memory and a processor, wherein the memory stores a data monitoring method program of an AC insulation measurement test device, and when the data monitoring method program of an AC insulation measurement test device is executed by the processor, the following steps are implemented: An insulation test is performed on the target insulation measurement test object through an AC insulation measurement test device to obtain a sequential real-time temperature chain of each insulation monitoring area block, and a temperature storage model of each insulation monitoring area block is constructed. Based on the temperature storage model and the sequential real-time temperature chain, the probability of occurrence of a three-dimensional thermal energy volume diagram under different partial discharge modes is analyzed to obtain a temperature anomaly monitoring result; Perform a partial discharge test on the target insulation measurement test object according to the temperature monitoring result to obtain the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal in each insulation monitoring area block, determine the current partial discharge mode by analyzing the spectrum amplitude distribution characteristics of the discharge signal and the time-frequency density distribution characteristics of the discharge signal, and generate an insulation defect layout map based on the monitoring and positioning of the current partial discharge mode; Based on the insulation defect layout diagram, the insulation defects of the target insulation measurement test object are feature extracted to obtain the edge direction matrix of each insulation defect. The initial insulation defect heat map is drawn based on the distribution of the Manhattan distance between the edge direction matrix of each insulation defect and the edge direction normalized matrix to analyze the insulation health status of the target insulation measurement test object and obtain the insulation health monitoring results. According to the insulation monitoring results, a breakdown test is performed on the target insulation measurement test object, and a tree-like discharge model diagram corresponding to each uniform time stamp of the target insulation measurement test object is constructed. Adjacent tree-like discharge model diagrams are registered in the model diagram registration field and the registration coincidence rate is calculated. According to the registration coincidence rate, it is estimated whether the breakdown test generates a dielectric loss factor, so as to generate insulation quality monitoring results. Among them, based on the insulation defect layout diagram, the insulation defects of the target insulation measurement test object are feature extracted to obtain the edge direction matrix of each insulation defect, and the initial insulation defect heat map is drawn in combination with the distribution of the Manhattan distance between the edge direction matrix of each insulation defect and the edge direction normalized matrix, so as to analyze the insulation health status of the target insulation measurement test object and obtain the insulation health monitoring result, which specifically includes the following steps: Using an industrial camera and based on the insulation defect layout diagram, the defect position of the target insulation measurement test object is photographed at a fixed point to obtain image data of each insulation defect in the insulation defect layout diagram; An edge direction matrix algorithm is introduced. In the edge direction matrix algorithm, the image data of each insulation defect is converted into a grayscale image and Gaussian blur processing is performed. The Sobel operator is input to detect the grayscale image data after Gaussian blur processing to obtain the gradient and edge direction angle of the grayscale image data. Preset a quantization direction interval of each insulation defect based on the gradient of the grayscale image data, quantize the edge direction angle to the quantization direction interval of each insulation defect, and then count the number of edge pixels in each edge direction in each quantization direction interval after quantization to generate an edge direction matrix for each insulation defect; The discharge frequency and discharge intensity of each insulation defect in the current partial discharge mode are extracted by analyzing the multi-dimensional partial discharge map, and the edge direction normalized matrix of each insulation defect in the current partial discharge mode is obtained in the big data network based on the discharge frequency and discharge intensity; Calculate the Manhattan distance between the edge direction matrix of each insulation defect and the corresponding edge direction normalized matrix, and obtain the Manhattan distance of the edge direction matrix that is relatively larger than the edge direction normalized matrix, which is defined as the positive Manhattan distance, and the Manhattan distance of the edge direction matrix that is relatively smaller than the edge direction normalized matrix, which is defined as the negative Manhattan distance; A heat map framework is constructed based on the insulation defect layout map, and all positive Manhattan distances and all negative Manhattan distances are marked on the corresponding insulation defects in the heat map framework to obtain an initial insulation defect heat map; Performing a health assessment on the discharge frequency and discharge intensity of each insulation defect through an insulation health status assessment model preset by an AC insulation measurement test device to obtain an insulation health status assessment value of each insulation defect, and generating a corresponding thermal region area in the initial insulation defect thermodynamic map based on the insulation health status assessment value; If the ratio between the sum of the thermal area areas of all positive Manhattan distances and the sum of the thermal area areas of all negative Manhattan distances in the initial insulation defect thermogram is greater than the preset ratio, it means that the insulation health is good, otherwise the insulation health is poor, and the insulation health monitoring result is generated.

Images