GIS insulation defect evaluation method and device based on phase resolution photon counting

Abstract

The invention relates to the technical field of electrical equipment insulation state evaluation, in particular to a GIS insulation defect evaluation method and device based on phase resolution photon counting, and the method comprises the steps: collecting photon pulses released by partial discharge of different defects, and building a phase resolution mode map according to a time phase relation, the method comprises the following steps: acquiring a phase analysis mode map of a gas insulated switchgear (GIS), then extracting characteristic parameters of multiple dimensions from the phase analysis mode map, and screening characteristics meeting a preset optimization condition from the characteristic parameters of the multiple dimensions based on a pre-constructed multi-channel diagnosis algorithm framework based on a parallel coordination and information fusion mechanism to obtain an insulation defect evaluation result of the gas insulated switchgear (GIS). Therefore, the problems that in optical detection in the prior art, due to the fact that diagnosis depends on a single time domain feature and is not combined with multi-dimensional features, insulation defects are difficult to accurately recognize, adaptability to defect types is lacked, generalization ability is insufficient, and detection requirements under actual working conditions are difficult to meet are solved.

Description

Technical Field

[0001] This application relates to the field of electrical equipment insulation condition assessment technology, and in particular to a method and apparatus for assessing GIS insulation defects based on phase-resolved photon counting. Background Technology

[0002] During the manufacturing, transportation, and assembly of GIS (Gas Insulated Switchgear), internal insulation defects (such as metal particles, burrs on the insulator surface, scratches, etc.) may occur. These defects can induce partial discharge under a strong electric field, causing insulation degradation and ultimately leading to equipment failure. Therefore, methods for detecting insulation defects in GIS have attracted much research attention.

[0003] Among the related technologies, there are pulsed current method, ultra-high frequency method and ultrasonic method, but they are susceptible to electromagnetic interference and are not sensitive to early micro-defects; there are also optical detection methods, which have the characteristics of high anti-interference and high sensitivity. For example, a high voltage is first applied to a typical discharge defect model of GIS to destroy the defect model. During the destruction process, the released electrical signals and electromagnetic waves are detected to obtain electrical data. At the same time, the released light signals are detected to obtain optical data to evaluate GIS insulation defects.

[0004] However, in optical inspection, the relevant technologies rely on single time-domain features and do not combine multi-dimensional features for diagnosis, which makes it difficult to accurately identify insulation defects, lacks adaptability to different defect types, and has insufficient generalization ability, making it difficult to meet the inspection needs under actual working conditions, and urgently needs to be solved. Summary of the Invention

[0005] This application provides a GIS insulation defect assessment method and apparatus based on phase-resolved photon counting to solve the problems of related technologies in optical detection, which rely on single time-domain features and do not combine multi-dimensional features for diagnosis, resulting in difficulty in accurately identifying insulation defects, lack of adaptability to defect types, insufficient generalization ability, and difficulty in meeting the detection needs under actual working conditions.

[0006] The first aspect of this application provides a method for assessing GIS insulation defects based on phase-resolved photon counting, comprising the following steps: collecting photon pulses released by partial discharges of different defects to establish a phase resolution pattern map according to the time-phase relationship; extracting feature parameters of multiple dimensions from the phase resolution pattern map; and selecting at least one feature that meets preset preference conditions from the feature parameters of the multiple dimensions based on a pre-constructed multi-channel diagnostic algorithm framework based on parallel collaboration and information fusion mechanism to obtain the GIS insulation defect assessment result.

[0007] Based on the above technical means, this application embodiment collects photon pulse signals and correlates them with AC voltage phase to establish a phase analysis pattern map, accurately captures weak discharge signals of early micro-defects, and completely preserves the phase distribution characteristics of different defect discharges. Then, by extracting multi-dimensional feature parameters, it makes up for the one-sidedness of single-dimensional features in defect characterization. Furthermore, through a parallel and collaborative multi-channel diagnostic algorithm framework, on the one hand, it focuses on optimizing features through feature screening to reduce redundant information interference; on the other hand, it combines image recognition to mine deep features that are difficult to characterize manually, significantly improving the accuracy, reliability, and cross-condition generalization ability of GIS insulation defect identification, providing efficient and reliable technical support for subsequent operation and maintenance decisions.

[0008] Optionally, in one embodiment of this application, the step of collecting photon pulses released by partial discharges of different defects to establish a phase analysis mode map according to the time-phase relationship includes: collecting photon pulse signals released by different insulation defects during partial discharge using a photon counting detection system; and constructing the phase analysis mode map according to the synchronization relationship between the photon pulse signals and the applied AC voltage.

[0009] Based on the above technical means, the embodiments of this application, on the one hand, accurately capture the photon pulse signals released by early micro-defects through a high-sensitivity photon counting detection system, and on the other hand, retain the phase distribution characteristics of different defect discharges by constructing a phase analysis mode map, so that the discharge patterns of various defects present differentiated characteristics in the map, providing structured and highly recognizable raw data support for subsequent multi-dimensional feature extraction.

[0010] Optionally, in one embodiment of this application, the step of extracting multiple dimensions of feature parameters from the phase resolution pattern map includes: extracting feature parameters covering multiple dimensions in the time domain, phase domain, frequency domain, and image morphology from the phase resolution pattern map.

[0011] Based on the above technical means, the embodiments of this application can comprehensively cover the differentiated discharge characteristics of different defects by extracting feature parameters in multiple dimensions such as time domain, phase domain, frequency domain and image morphology, significantly reduce the probability of feature confusion between different defects, and provide sufficient feature support for subsequent multi-channel diagnostic algorithms. This effectively improves the accuracy of insulation defect identification and cross-condition generalization ability, avoids the limitation of single-dimensional features on the adaptability of specific defects or conditions, and meets the high-precision defect assessment requirements in complex GIS operation scenarios.

[0012] Optionally, in one embodiment of this application, before screening the at least one feature that meets the preset preferred conditions, the method further includes: constructing a feature selection-classifier channel and an image recognition channel based on a dual-channel diagnostic framework with parallel collaboration and information fusion mechanism, so as to use the feature selection-classifier channel to screen at least one target feature to obtain a preferred feature subset, send the preferred feature subset into the classifier, and use the image recognition channel to directly process the phase resolution pattern map using a neural network, and extract deep features through an end-to-end learning method.

[0013] Based on the above technical means, in the feature optimization-classifier channel, the present application embodiment eliminates redundant information and focuses on core features through feature screening, thereby reducing the computational complexity of the classifier and improving diagnostic efficiency. In addition, in the image recognition channel, by using the end-to-end learning of the convolutional neural network, deep features are automatically mined to make up for the feature omissions or one-sided characterization problems that may exist when manually extracting multi-dimensional features, effectively covering the differentiated representation of different defects, and significantly improving the anti-interference ability and robustness of defect recognition.

[0014] A second aspect of this application provides a GIS insulation defect assessment device based on phase-resolved photon counting, comprising: an acquisition module for acquiring photon pulses released by partial discharges of different defects to establish a phase resolution pattern map according to the time-phase relationship; an extraction module for extracting feature parameters of multiple dimensions from the phase resolution pattern map; and an assessment module for selecting at least one feature that meets preset preference conditions from the feature parameters of the multiple dimensions based on a pre-constructed multi-channel diagnostic algorithm framework based on parallel collaboration and information fusion mechanism to obtain the GIS insulation defect assessment result.

[0015] Based on the above technical means, this application embodiment collects photon pulse signals and correlates them with AC voltage phase to establish a phase analysis pattern map, accurately captures weak discharge signals of early micro-defects, and completely preserves the phase distribution characteristics of different defect discharges. Then, by extracting multi-dimensional feature parameters, it makes up for the one-sidedness of single-dimensional features in defect characterization. Furthermore, through a parallel and collaborative multi-channel diagnostic algorithm framework, on the one hand, it focuses on optimizing features through feature screening to reduce redundant information interference; on the other hand, it combines image recognition to mine deep features that are difficult to characterize manually, significantly improving the accuracy, reliability, and cross-condition generalization ability of GIS insulation defect identification, providing efficient and reliable technical support for subsequent operation and maintenance decisions.

[0016] Optionally, in one embodiment of this application, the acquisition module includes: an acquisition unit, used to acquire photon pulse signals released by different insulation defects during partial discharge using a photon counting detection system; and a construction unit, used to construct the phase analysis mode map based on the synchronization relationship between the photon pulse signals and the applied AC voltage phase.

[0017] Based on the above technical means, the embodiments of this application, on the one hand, accurately capture the photon pulse signals released by early micro-defects through a high-sensitivity photon counting detection system, and on the other hand, retain the phase distribution characteristics of different defect discharges by constructing a phase analysis mode map, so that the discharge patterns of various defects present differentiated characteristics in the map, providing structured and highly recognizable raw data support for subsequent multi-dimensional feature extraction.

[0018] Optionally, in one embodiment of this application, the extraction module includes: an extraction unit, used to extract feature parameters covering multiple dimensions of time domain, phase domain, frequency domain and image morphology from the phase analysis pattern map.

[0019] Based on the above technical means, the embodiments of this application can comprehensively cover the differentiated discharge characteristics of different defects by extracting feature parameters in multiple dimensions such as time domain, phase domain, frequency domain and image morphology, significantly reduce the probability of feature confusion between different defects, and provide sufficient feature support for subsequent multi-channel diagnostic algorithms. This effectively improves the accuracy of insulation defect identification and cross-condition generalization ability, avoids the limitation of single-dimensional features on the adaptability of specific defects or conditions, and meets the high-precision defect assessment requirements in complex GIS operation scenarios.

[0020] Optionally, in one embodiment of this application, it further includes: a construction module for constructing a feature optimization-classifier channel and an image recognition channel of a dual-channel diagnostic framework based on parallel collaboration and information fusion mechanism, so as to use the feature optimization-classifier channel to filter at least one target feature to obtain a preferred feature subset, send the preferred feature subset into the classifier, and use the image recognition channel to directly process the phase resolution pattern map using a neural network, and extract deep features through an end-to-end learning method.

[0021] Based on the above technical means, in the feature optimization-classifier channel, the present application embodiment eliminates redundant information and focuses on core features through feature screening, thereby reducing the computational complexity of the classifier and improving diagnostic efficiency. In addition, in the image recognition channel, by using the end-to-end learning of the convolutional neural network, deep features are automatically mined to make up for the feature omissions or one-sided characterization problems that may exist when manually extracting multi-dimensional features, effectively covering the differentiated representation of different defects, and significantly improving the anti-interference ability and robustness of defect recognition.

[0022] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the GIS insulation defect assessment method based on phase-resolved photon counting as described in the above embodiments.

[0023] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for assessing GIS insulation defects based on phase-resolved photon counting.

[0024] A fifth aspect of this application provides a computer program product, including a computer program that, when executed, implements the above-described method for assessing GIS insulation defects based on phase-resolved photon counting.

[0025] This application embodiment acquires photon pulse signals and correlates them with AC voltage phase to establish a phase analysis pattern map, accurately capturing weak discharge signals of early micro-defects and fully preserving the phase distribution characteristics of different defect discharges. Then, by extracting multi-dimensional feature parameters, it compensates for the limitations of single-dimensional features in defect characterization. Furthermore, through a parallel and collaborative multi-channel diagnostic algorithm framework, it focuses on optimizing features through feature screening to reduce redundant information interference, and combines image recognition to mine deep features that are difficult to characterize manually. This significantly improves the accuracy, reliability, and cross-condition generalization ability of GIS insulation defect identification, providing efficient and reliable technical support for subsequent operation and maintenance decisions. Therefore, it solves the problem that related technologies in optical inspection rely on single time-domain features without combining multi-dimensional features for diagnosis, resulting in difficulty in accurately identifying insulation defects, lack of adaptability to defect types, insufficient generalization ability, and inability to meet the detection needs under actual working conditions.

[0026] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0027] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart of a GIS insulation defect assessment method based on phase-resolved photon counting, according to an embodiment of this application. Figure 2 This is a schematic diagram illustrating the acquisition of photon pulse signals and the construction of a phase-analyzed mode map according to an embodiment of this application; Figure 3(a) is a schematic diagram of the spectrum data of surface metal defects under multiple voltage conditions provided according to an embodiment of this application; Figure 3(b) is a schematic diagram of the spectral data of scratch defects under multiple voltage conditions provided according to an embodiment of this application; Figure 3(c) is a schematic diagram of the spectral data of metal tip defects under multiple voltage conditions provided according to an embodiment of this application; Figure 4 This is a flowchart illustrating the principle of a GIS insulation defect assessment method based on phase-resolved photon counting according to an embodiment of this application. Figure 5 This is a block diagram of a GIS insulation defect assessment device based on phase-resolved photon counting provided in an embodiment of this application; Figure 6 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application.

[0028] Figure label: 50-GIS insulation defect assessment device; 100-acquisition module, 200-extraction module, 300-assessment module; 601-memory, 602-processor, 603-communication interface. Detailed Implementation

[0029] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0030] The following description, with reference to the accompanying drawings, illustrates a GIS insulation defect assessment method and apparatus based on phase-resolved photon counting, according to embodiments of this application. Addressing the issues raised in the background section regarding the reliance on single-time-domain features in optical inspection, which fails to incorporate multi-dimensional features for diagnosis, leading to difficulties in accurately identifying insulation defects, a lack of adaptability to different defect types, insufficient generalization ability, and an inability to meet the detection needs of actual operating conditions, this application provides a GIS insulation defect assessment method based on phase-resolved photon counting. This method acquires photon pulse signals and correlates them with AC voltage phases to establish a phase resolution pattern map, accurately capturing weak discharge signals of early micro-defects and fully preserving the phase distribution characteristics of different defect discharges. Furthermore, by extracting multi-dimensional feature parameters, it overcomes the limitations of single-dimensional features in defect characterization. Through a parallel and collaborative multi-channel diagnostic algorithm framework, it focuses on optimizing features through feature screening to reduce redundant information interference, and combines image recognition to uncover deep features that are difficult to characterize manually. This significantly improves the accuracy, reliability, and cross-condition generalization ability of GIS insulation defect identification, providing efficient and reliable technical support for subsequent operation and maintenance decisions. This solves the problem that related technologies in optical inspection rely on single time-domain features, fail to combine multi-dimensional features for diagnosis, resulting in difficulty in accurately identifying insulation defects, lack of adaptability to defect types, insufficient generalization ability, and difficulty in meeting the inspection needs under actual working conditions.

[0031] Specifically, Figure 1 This is a flowchart of a GIS insulation defect assessment method based on phase-resolved photon counting, according to an embodiment of this application.

[0032] like Figure 1 As shown, the GIS insulation defect assessment method based on phase-resolved photon counting includes the following steps: In step S101, photon pulses released by partial discharges from different defects are collected to establish a phase analysis mode map according to the time-phase relationship.

[0033] The following details how embodiments of this application collect photon pulses released by partial discharges from different defects to establish a phase analysis mode map according to the time-phase relationship.

[0034] Specifically, in one embodiment of this application, photon pulses released by partial discharges of different defects are collected to establish a phase resolution pattern map according to the time-phase relationship, including: collecting photon pulse signals released by different insulation defects during partial discharge using a photon counting detection system; and constructing a phase resolution pattern map based on the synchronization relationship between the photon pulse signals and the applied AC voltage phase.

[0035] In the embodiments of this application, the photon counting detection system can be understood as a device that can accurately respond to the single photons released during partial discharge of different insulation defects, convert the photon signals into identifiable electrical pulse signals, count them after filtering out environmental noise interference, and simultaneously support synchronous acquisition with AC voltage phase, so as to provide the original signal input for the construction of phase analysis mode map.

[0036] Furthermore, insulation defects can be understood as various flaws or abnormal states that deviate from design standards and affect insulation performance in the internal insulation structure or components of GIS during manufacturing, transportation, assembly, or long-term operation. These may include, but are not limited to, metal tip protrusions, residual free metal particles, burrs and scratches on the surface of insulating parts. This application does not impose specific limitations.

[0037] In addition, the photon pulse signal can be understood as a discrete optical signal formed by photon radiation generated during the energy transition in the discharge process when a partial discharge occurs in the insulation defect inside the GIS. This signal has the characteristics of being extremely weak, short time domain, and pulsed. Its parameters such as photon release intensity, pulse interval, and duration are directly related to the defect type and discharge intensity, and are used to reflect the physical characteristics of insulation defect discharge.

[0038] Furthermore, the phase analysis mode map can be understood as a structured data map constructed based on the strong correlation between photon pulse signals and AC voltage phase. It uses the AC voltage phase (0°~360°) applied to GIS as the horizontal axis and the photon count within the corresponding phase interval as the vertical axis to intuitively present the phase distribution pattern of photon counts for different insulation defects.

[0039] In actual implementation, the embodiments of this application can use a photon counting detection system to collect extremely weak photon pulse signals released by different GIS insulation defects during partial discharge, and simultaneously collect the phase of the applied AC voltage to accurately associate each photon event with the corresponding phase value. Then, the phase range of 0°~360° is divided into several phase windows, and the photon counts falling into each phase window in multiple discharge cycles are counted to construct a phase analysis mode map.

[0040] In this application, on the one hand, a high-sensitivity photon counting detection system is used to accurately capture photon pulse signals released in the early stages of micro-defects. On the other hand, by constructing a phase analysis mode map, the phase distribution characteristics of different defect discharges are preserved, so that the discharge patterns of various defects are presented in a differentiated manner in the map, providing structured and highly recognizable raw data support for subsequent multi-dimensional feature extraction.

[0041] like Figure 2As shown, based on the description of other embodiments, by way of example, the embodiments of this application first simulate the actual operating conditions of GIS through a test circuit module. Specifically, the power supply is a transformer, which outputs an AC voltage of 10 kV to 50 kV and applies it to the GIS defect sample (containing insulation defects such as metal tip protrusions and free metal particle residues). The defect sample undergoes partial discharge under the action of a strong electric field, and at the same time releases extremely weak photon pulse signals.

[0042] Furthermore, in this embodiment of the application, photon pulse signals are captured by a signal acquisition module (i.e., a photon counting and detection system). Specifically, a photon probe is aimed at the discharge area of ​​the defect sample to capture the photon pulse signals released by the partial discharge in real time. The data acquisition card simultaneously acquires the AC voltage phase output by the test circuit module to accurately associate each photon event with the corresponding phase value.

[0043] Furthermore, after receiving the correlation data between the photon pulse signal and the AC voltage phase transmitted by the data acquisition card, the computer statistically analyzes the data collected over 20 power frequency cycles to reduce random errors. Then, the phase range of 0° to 360° is divided into several phase windows, and the photon count in each phase window is counted. The spectrum on the left is generated first (the horizontal axis represents the AC voltage phase, the vertical axis represents the photon count, and the color depth represents different power frequency cycles). Then, the features of the spectrum are extracted to obtain the phase analysis pattern spectrum on the right (the horizontal axis represents the AC voltage phase, and the vertical axis represents the photon count).

[0044] As shown in Figure 3, based on the description of other embodiments, by way of example, the embodiments of this application can obtain spectrum data under multiple voltage conditions based on the phase analysis mode spectrum. Figure 3(a) corresponds to surface metal defects. As the test voltage increases, the number of photon pulses (i.e., photon count) shrinks towards 90° and 270°, and the value gradually increases (the statistical range covers 0~850). Figure 3(b) corresponds to scratch defects. As the test voltage increases, the number of photon pulses shrinks towards 90° and 270°, and the value gradually increases, but the value is generally at a low level (the statistical range covers 0~5). Figure 3(c) corresponds to metal tip defects. As the test voltage increases, the number of photon pulses shrinks towards 90° and 270°, and the value gradually increases. The increase in value is much higher than that of the above two types of defects (the statistical range covers 0~950).

[0045] Understandably, the spectral data under multiple voltage conditions not only reflect the differences in discharge mechanisms of different insulation defects, but also the evolution of discharge characteristics of different insulation defects with test voltage, which can provide a data foundation for the identification of insulation defects under multiple voltage conditions.

[0046] In step S102, feature parameters of multiple dimensions are extracted from the phase analysis mode map.

[0047] The following details how embodiments of this application extract multi-dimensional feature parameters from phase analysis pattern maps.

[0048] Specifically, in one embodiment of this application, multiple-dimensional feature parameters are extracted from the phase resolution pattern map, including: extracting feature parameters covering multiple dimensions in the time domain, phase domain, frequency domain and image morphology from the phase resolution pattern map.

[0049] In the embodiments of this application, the time-domain characteristic parameters can be understood as quantitative indicators extracted based on the time evolution law of photon pulse signals, used to quantify the overall intensity and distribution shape of the discharge from a macroscopic perspective. These parameters may include, but are not limited to, the total number of photon pulses, the standard deviation of photon pulses, skewness and kurtosis. This application does not impose any specific limitations.

[0050] Furthermore, the characteristic parameters of the phase domain can be understood as quantitative indicators extracted based on the correlation between the photon pulse signal and the phase of the AC voltage. These parameters are used to accurately describe the occurrence pattern and symmetry of the discharge activity within the AC cycle. They may include, but are not limited to, the positive and negative half-cycle correlation coefficients, the peak phase range, and the discharge initiation and extinction phases. This application does not impose any specific limitations.

[0051] In addition, the characteristic parameters in the frequency domain can be understood as quantitative indicators extracted after converting the time-domain photon pulse signal into a frequency-domain signal through signal processing methods such as wavelet transform. These parameters are used to reveal the energy distribution characteristics of the discharge pulse in the frequency domain and may include, but are not limited to, the energy entropy of different frequency bands and the energy proportion of the main frequency band. This application does not impose any specific limitations.

[0052] Furthermore, the feature parameters of image morphology can be understood as quantitative indicators extracted by the phase resolution pattern map based on the phase resolution pattern map pre-converted to grayscale image and the pyramid gradient histogram algorithm. These indicators may include, but are not limited to, texture features and contour shape features. This application does not impose any specific limitations.

[0053] In actual implementation, since the phase analysis pattern map is a structured data map constructed based on the strong correlation between the photon pulse signal and the phase of the AC voltage, and different insulation defect discharge mechanisms are different, it can form unique and distinguishable features in multiple dimensions such as time domain, phase domain, frequency domain and image morphology. Therefore, the embodiments of this application can extract feature parameters covering multiple dimensions such as time domain, phase domain, frequency domain and image morphology from the phase analysis pattern map.

[0054] This application embodiment extracts feature parameters from multiple dimensions, including time domain, phase domain, frequency domain, and image morphology, to comprehensively cover the differentiated discharge characteristics of different defects, significantly reduce the probability of feature confusion between different defects, and provide sufficient feature support for subsequent multi-channel diagnostic algorithms. This effectively improves the accuracy of insulation defect identification and its cross-condition generalization ability, avoids the limitations of single-dimensional features in adapting to specific defects or conditions, and meets the high-precision defect assessment requirements in complex GIS operation scenarios.

[0055] In step S103, based on a pre-built multi-channel diagnostic algorithm framework based on parallel collaboration and information fusion mechanism, features that meet certain optimization conditions are selected from feature parameters of multiple dimensions to obtain GIS insulation defect assessment results.

[0056] In the embodiments of this application, the multi-channel diagnostic algorithm framework can be understood as an algorithm framework with multi-path parallel processing, feature complementary verification, and result fusion decision as its design logic. The framework includes a parallel feature optimization-classifier channel and an image recognition channel. The feature optimization-classifier channel focuses on the screening and classification of multi-dimensional quantitative features, while the image recognition channel focuses on deep feature mining of phase analysis pattern maps. These channels capture defect features from the dimensions of quantitative statistics and spatial morphology, respectively. At the same time, the framework has a built-in information fusion mechanism that can integrate and analyze the intermediate results output by the two channels to eliminate the diagnostic bias of a single channel.

[0057] In addition, the GIS insulation defect assessment results can be understood as the specific type identification results of GIS insulation defects (such as metal tip defects, surface metal defects, and insulation surface scratch defects) after being processed by a multi-channel diagnostic algorithm framework.

[0058] The following details how the embodiments of this application, based on a pre-built multi-channel diagnostic algorithm framework based on parallel collaboration and information fusion mechanisms, select features that meet certain preferred conditions from feature parameters of multiple dimensions to obtain GIS insulation defect assessment results.

[0059] Specifically, in one embodiment of this application, before screening features that meet certain preferred conditions, the method further includes: constructing a feature selection-classifier channel and an image recognition channel based on a dual-channel diagnostic framework with parallel collaboration and information fusion mechanism, so as to use the feature selection-classifier channel to screen target features, obtain a preferred feature subset, send the preferred feature subset into the classifier, and use the image recognition channel to directly process the phase resolution pattern map using a neural network, and extract deep features through an end-to-end learning method.

[0060] In the embodiments of this application, certain preferred conditions can be understood as quantitative standards and judgment rules for selecting preferred features from multi-dimensional feature parameters. These may include, but are not limited to, the correlation threshold between features and insulation defects (e.g., mutual information value greater than a certain threshold indicates that the feature is closely related to the defect type), the upper limit of redundancy between features (e.g., feature similarity less than a certain threshold to avoid interference from duplicate information), the variance contribution rate of features (e.g., features with the top N variance contribution rates to ensure that the features have strong characterization capabilities), and the stability index of features (e.g., the coefficient of variation under different voltage conditions is less than a certain value to ensure the adaptability of features across operating conditions). This application does not impose specific limitations.

[0061] Furthermore, deep features can be understood as abstract features that cannot be directly obtained through manual design, automatically extracted after end-to-end learning of the phase resolution pattern map by a neural network (such as a convolutional neural network) in the image recognition channel. These features are used to compensate for the limitations of manual feature design.

[0062] In actual implementation, the embodiments of this application can achieve comprehensive mining and complementary verification of defect features through dual-channel parallel processing. Specifically, in the feature optimization-classifier channel, multiple dimensions of feature parameters (i.e., target features) in the extracted time domain, phase domain, frequency domain, and image morphology can be selected based on certain optimization conditions such as the correlation between features and insulation defects and the redundancy between features. A subset of optimized features is generated, and then a classifier (such as random forest or SVM) is used to learn the statistical correlation between the optimized features and insulation defects to achieve defect diagnosis based on quantitative features. In the image recognition channel, a neural network can be used to directly input the phase resolution pattern map. Through hierarchical processing of convolutional layers and pooling layers and end-to-end learning, deep features (such as texture of clustered regions and phase distribution topology) can be automatically mined.

[0063] In the feature optimization-classifier channel of this application, redundant information is eliminated and core features are focused through feature screening, which reduces the computational complexity of the classifier and improves diagnostic efficiency. In addition, in the image recognition channel, deep features are automatically mined by using end-to-end learning of convolutional neural networks, which makes up for the feature omissions or one-sided characterization problems that may exist when manually extracting multi-dimensional features, effectively covering the differentiated representation of different defects, and significantly improving the anti-interference ability and robustness of defect recognition.

[0064] Furthermore, the embodiments of this application can utilize the DS evidence theory to perform fusion decision-making. Specifically, based on the intermediate results of the two channels, the confidence probability corresponding to the insulation defect is calculated separately and assigned a corresponding weight. Then, the confidence probabilities corresponding to the two channels are weighted and fused to select the insulation defect with the highest confidence as the specific type identification result. This effectively combines the logical judgment advantage of numerical features with the pattern recognition ability of image recognition, significantly improving the reliability and accuracy of the identification results.

[0065] The principle of the GIS insulation defect assessment method based on phase-resolved photon counting proposed in this application is illustrated below with a specific embodiment.

[0066] Figure 4 This is a flowchart illustrating the principle of a GIS insulation defect assessment method based on phase-resolved photon counting according to an embodiment of this application.

[0067] Step S401: Construct a phase resolution mode map.

[0068] In this embodiment, a photon counting detection system can be used to collect extremely weak photon pulse signals released by different GIS insulation defects during partial discharge, and simultaneously collect the phase of the applied AC voltage to accurately associate each photon event with the corresponding phase value. Then, the phase range of 0°~360° is divided into several phase windows, and the photon counts falling into each phase window in multiple discharge cycles are counted to construct a phase analysis mode map.

[0069] Step S402: Extract multi-dimensional feature parameters.

[0070] In this embodiment, feature parameters covering multiple dimensions including time domain, phase domain, frequency domain, and image morphology can be extracted from the phase analysis pattern map.

[0071] Step S403: Extract deep features through the image recognition channel.

[0072] In this embodiment, a neural network can be used to directly input the phase resolution pattern map. Through hierarchical processing of convolutional layers and pooling layers and end-to-end learning, deep features (such as clustered region texture and phase distribution topology) can be automatically mined.

[0073] Step S404: Filter the optimal features through the feature optimization-classifier channel.

[0074] In this embodiment, the extracted feature parameters (i.e., target features) in multiple dimensions of time domain, phase domain, frequency domain and image morphology can be selected based on certain optimization conditions such as the correlation between features and insulation defects and the redundancy between features. The optimized feature subset is generated, and then a classifier (such as random forest or SVM) is used to learn the statistical correlation between the optimized features and insulation defects to achieve defect diagnosis based on quantitative features.

[0075] Step S405: Make a decision based on the DS evidence theory.

[0076] In this embodiment, the confidence probability of the insulation defect is first calculated based on the intermediate results of the two channels, and corresponding weights are assigned. Then, the confidence probabilities of the two channels are weighted and fused to select the insulation defect with the highest confidence as the specific type identification result.

[0077] The GIS insulation defect assessment method based on phase-resolved photon counting proposed in this application collects photon pulse signals and correlates them with AC voltage phases to establish a phase resolution pattern map. This accurately captures weak discharge signals of early micro-defects and fully preserves the phase distribution characteristics of different defect discharges. Furthermore, by extracting multi-dimensional feature parameters, it overcomes the limitations of single-dimensional features in defect characterization. Through a parallel and collaborative multi-channel diagnostic algorithm framework, it focuses on optimizing features through feature screening to reduce redundant information interference. Simultaneously, it combines image recognition to mine deep features that are difficult to characterize manually. This significantly improves the accuracy, reliability, and cross-condition generalization ability of GIS insulation defect identification, providing efficient and reliable technical support for subsequent operation and maintenance decisions. Therefore, it solves the problem that related technologies in optical inspection rely on single time-domain features without combining multi-dimensional features for diagnosis, resulting in difficulty in accurately identifying insulation defects, lack of adaptability to defect types, insufficient generalization ability, and inability to meet the detection needs of actual working conditions.

[0078] Next, referring to the accompanying drawings, a GIS insulation defect assessment device based on phase-resolved photon counting, according to an embodiment of this application, is described.

[0079] Figure 5 This is a block diagram of a GIS insulation defect assessment device based on phase-resolved photon counting provided in an embodiment of this application.

[0080] like Figure 5 As shown, the GIS insulation defect assessment device 50 based on phase-resolved photon counting includes: a data acquisition module 100, an extraction module 200, and an assessment module 300.

[0081] Among them, the acquisition module 100 is used to acquire photon pulses released by partial discharges of different defects, so as to establish a phase analysis mode map according to the time-phase relationship; Extraction module 200 is used to extract feature parameters of multiple dimensions from the phase analysis mode map; The evaluation module 300 is used to select at least one feature that meets the preset optimization conditions from multiple dimensions of feature parameters based on a pre-built multi-channel diagnostic algorithm framework based on parallel collaboration and information fusion mechanism, so as to obtain the GIS insulation defect evaluation result.

[0082] Optionally, in one embodiment of this application, the acquisition module 100 includes: an acquisition unit and a construction unit.

[0083] The acquisition unit is used to acquire photon pulse signals released during partial discharge of different insulation defects using a photon counting detection system. The building unit is used to construct a phase resolution mode map based on the synchronization relationship between the photon pulse signal and the applied AC voltage phase.

[0084] Optionally, in one embodiment of this application, the extraction module 200 includes: an extraction unit, used to extract feature parameters covering multiple dimensions of time domain, phase domain, frequency domain and image morphology from the phase analysis pattern map.

[0085] Optionally, in one embodiment of this application, it further includes: a construction module for constructing a feature optimization-classifier channel and an image recognition channel based on a parallel collaboration and information fusion mechanism for a dual-channel diagnostic framework, so as to use the feature optimization-classifier channel to filter at least one target feature to obtain a preferred feature subset, send the preferred feature subset into the classifier, and use the image recognition channel to directly process the phase resolution pattern map using a neural network, and extract deep features through an end-to-end learning method.

[0086] It should be noted that the foregoing explanation of the GIS insulation defect assessment method based on phase-resolved photon counting also applies to the GIS insulation defect assessment device based on phase-resolved photon counting in this embodiment, and will not be repeated here.

[0087] The GIS insulation defect assessment device based on phase-resolved photon counting proposed in this application collects photon pulse signals and correlates them with AC voltage phase to establish a phase resolution pattern map. This accurately captures weak discharge signals of early micro-defects and fully preserves the phase distribution characteristics of different defect discharges. Furthermore, by extracting multi-dimensional feature parameters, it overcomes the limitations of single-dimensional features in defect characterization. Through a parallel and collaborative multi-channel diagnostic algorithm framework, it focuses on optimizing features through feature screening to reduce redundant information interference. Simultaneously, it combines image recognition to uncover deep features that are difficult to characterize manually. This significantly improves the accuracy, reliability, and cross-condition generalization ability of GIS insulation defect identification, providing efficient and reliable technical support for subsequent operation and maintenance decisions. Therefore, it solves the problem that related technologies in optical inspection rely on single time-domain features without combining multi-dimensional features for diagnosis, resulting in difficulty in accurately identifying insulation defects, lack of adaptability to defect types, insufficient generalization ability, and inability to meet the detection needs of actual working conditions.

[0088] Figure 6 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application. The electronic device may include: The memory 601, the processor 602, and the computer program stored on the memory 601 and capable of running on the processor 602.

[0089] When the processor 602 executes the program, it implements the GIS insulation defect assessment method based on phase-resolved photon counting provided in the above embodiments.

[0090] Furthermore, electronic devices also include: Communication interface 603 is used for communication between memory 601 and processor 602.

[0091] The memory 601 is used to store computer programs that can run on the processor 602.

[0092] The memory 601 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0093] If the memory 601, processor 602, and communication interface 603 are implemented independently, then the communication interface 603, memory 601, and processor 602 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0094] Optionally, in a specific implementation, if the memory 601, processor 602, and communication interface 603 are integrated on a single chip, then the memory 601, processor 602, and communication interface 603 can communicate with each other through an internal interface.

[0095] The processor 602 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0096] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described GIS insulation defect assessment method based on phase-resolved photon counting.

[0097] This application also provides a computer program product, including a computer program that, when executed, implements the above-described GIS insulation defect assessment method based on phase-resolved photon counting.

[0098] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0099] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0100] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0101] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0102] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, it can be implemented using any one or more of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0103] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it includes one or a combination of the steps of the method embodiments.

[0104] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0105] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for assessing GIS insulation defects based on phase-resolved photon counting, characterized in that, Includes the following steps: Photon pulses emitted by partial discharges from different defects are collected to establish a phase-analyzed mode map based on the time-phase relationship; Multiple feature parameters are extracted from the phase analysis pattern map; Based on a pre-built multi-channel diagnostic algorithm framework with parallel collaboration and information fusion mechanism, at least one feature that meets the preset optimization conditions is selected from the feature parameters of the multiple dimensions to obtain the GIS insulation defect assessment result of gas-insulated switchgear.

2. The method according to claim 1, characterized in that, The process of collecting photon pulses emitted from partial discharges of different defects to establish a phase-analyzed mode map according to the time-phase relationship includes: A photon counting detection system was used to collect photon pulse signals released during partial discharge of different insulation defects; The phase resolution mode map is constructed based on the synchronization relationship between the photon pulse signal and the applied AC voltage phase.

3. The method according to claim 1 or 2, characterized in that, The extraction of multi-dimensional feature parameters from the phase resolution pattern map includes: From the phase analysis pattern map, feature parameters covering multiple dimensions including time domain, phase domain, frequency domain, and image morphology are extracted.

4. The method according to claim 1, characterized in that, Before selecting at least one feature that satisfies the preset preferred conditions, the process further includes: A dual-channel diagnostic framework based on parallel collaboration and information fusion mechanism is constructed, consisting of a feature optimization-classifier channel and an image recognition channel. The feature optimization-classifier channel is used to filter at least one target feature to obtain a preferred feature subset. The preferred feature subset is then fed into the classifier, and the image recognition channel is used to directly process the phase resolution pattern map using a neural network. Deep features are extracted through an end-to-end learning approach.

5. A GIS insulation defect assessment device based on phase-resolved photon counting, characterized in that, include: The acquisition module is used to acquire photon pulses released by partial discharges from different defects in order to establish a phase analysis mode map according to the time-phase relationship; The extraction module is used to extract feature parameters of multiple dimensions from the phase analysis mode map; The evaluation module is used to select at least one feature that meets the preset preference conditions from the feature parameters of the multiple dimensions based on a pre-built multi-channel diagnostic algorithm framework based on parallel collaboration and information fusion mechanism, so as to obtain the GIS insulation defect evaluation result of gas-insulated switchgear.

6. The apparatus according to claim 5, characterized in that, The acquisition module includes: The acquisition unit is used to acquire photon pulse signals released during partial discharge of different insulation defects using a photon counting detection system. The construction unit is used to construct the phase resolution mode map based on the synchronization relationship between the photon pulse signal and the applied AC voltage phase.

7. The apparatus according to claim 5 or 6, characterized in that, The extraction module includes: The extraction unit is used to extract feature parameters covering multiple dimensions, including time domain, phase domain, frequency domain, and image morphology, from the phase analysis pattern map.

8. An electronic device, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, the processor executing the program to implement the GIS insulation defect assessment method based on phase-resolved photon counting as described in any one of claims 1-4.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the GIS insulation defect assessment method based on phase-resolved photon counting as described in any one of claims 1-4.

10. A computer program product, comprising a computer program, characterized in that, The computer program is executed to implement the GIS insulation defect assessment method based on phase-resolved photon counting as described in any one of claims 1-4.

Images