Transformer operation state analysis method, device, equipment, medium and program product

Abstract

The invention relates to a transformer operation state analysis method, device and equipment, a medium and a program product. The method comprises the following steps: firstly, acquiring a multi-source feature vector of a transformer, and then aiming at each fault type of the transformer, according to at least one first feature vector corresponding to the fault type, a second feature vector corresponding to other fault diagnosis paths and analysis data generated by the second feature vector in the other fault diagnosis paths, determining a multi-source feature vector of the transformer; the multi-source feature vector comprises a first feature vector and a second feature vector, and finally, based on the analysis results corresponding to the multiple fault types of the transformer, the operation state of the transformer is analyzed. By adopting the method, information interaction and cross validation of each fault diagnosis path can be realized, and the accuracy of operation state analysis of the transformer is higher.

Description

Technical Field

[0001] This application relates to the field of power equipment condition monitoring technology, and in particular to a method, device, equipment, medium, and program product for analyzing the operating status of a transformer. Background Technology

[0002] As a core component of the power grid, transformers play a crucial role in voltage transformation, power distribution, and electrical isolation. The operational stability of transformers directly affects the reliability and power quality of the entire power system. Therefore, it is essential to conduct condition monitoring and fault diagnosis of transformers, and to promptly assess their health status.

[0003] Currently, methods for monitoring the operating status of transformers are not very effective. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, device, equipment, medium, and program product for analyzing the operating status of transformers to address the aforementioned technical problems.

[0005] Firstly, this application provides a method for analyzing the operating status of a transformer, the method comprising:

[0006] Obtain the multi-source feature vector of the transformer;

[0007] For each fault type of transformer, the analysis result corresponding to the fault type is determined based on at least one first feature vector corresponding to the fault type, the second feature vector corresponding to other fault diagnosis paths, and the analysis data generated by the second feature vector in other fault diagnosis paths; the multi-source feature vector includes the first feature vector and the second feature vector.

[0008] Based on the analysis results corresponding to various fault types of transformers, the operating status of transformers is analyzed.

[0009] In one embodiment, the analysis result corresponding to the fault type is determined based on at least one first feature vector corresponding to the fault type, a second feature vector related to the fault type, and analysis data generated by the second feature vector in other fault diagnosis paths, including:

[0010] The analysis data corresponding to the first feature vector, the second feature vector, and the second feature vector are input into the analysis model corresponding to the fault type to obtain the analysis results corresponding to the fault type.

[0011] In one embodiment, if the fault type is a discharge fault, the method further includes:

[0012] Obtain winding deformation analysis information of the transformer in the mechanical fault diagnosis path;

[0013] If the winding deformation analysis information indicates that the transformer windings are deformed, then the mechanical vibration characteristic vector of the transformer is determined as the second characteristic vector.

[0014] In one embodiment, if the fault type is an overheating fault, the method further includes:

[0015] Obtain dissolved gas analysis information in the discharge fault diagnosis path, and / or, state analysis information in the mechanical fault diagnosis path;

[0016] If the dissolved gas analysis information indicates that the transformer has a thermal fault, then the oil chromatography feature vector is determined as the second feature vector, and / or, if the condition analysis information indicates that the transformer is in an abnormal operating state, then the electrical stress feature vector and the leakage flux feature vector are determined as the second feature vector.

[0017] In one embodiment, the method further includes:

[0018] If fault information of the target fault type is detected, the detection results of other fault diagnosis paths are obtained;

[0019] Based on the detection results of other fault diagnosis paths, determine the fault cause and handling method corresponding to the fault information.

[0020] In one embodiment, obtaining the multi-source feature vector of the transformer includes:

[0021] Obtain multi-source operating information of the transformer;

[0022] For each type of operational information, the feature vector of the operational information is extracted to obtain a multi-source feature vector.

[0023] Secondly, this application also provides a transformer operating status analysis device, comprising:

[0024] The acquisition module is used to acquire the multi-source feature vector of the transformer;

[0025] The determination module is used to determine the analysis result corresponding to each fault type of the transformer based on at least one first feature vector corresponding to the fault type, second feature vectors corresponding to other fault diagnosis paths, and analysis data generated by the second feature vectors in other fault diagnosis paths; the multi-source feature vectors include the first feature vector and the second feature vector.

[0026] The analysis module is used to analyze the operating status of the transformer based on the analysis results corresponding to various fault types of the transformer.

[0027] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described in any of the first aspects above.

[0028] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any of the first aspects above.

[0029] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described in any of the first aspects above.

[0030] The aforementioned transformer operation status analysis method, device, equipment, medium, and program product first acquires multi-source feature vectors of the transformer. Then, for each fault type of the transformer, based on at least one first feature vector corresponding to the fault type, second feature vectors corresponding to other fault diagnosis paths, and analysis data generated by the second feature vectors in other fault diagnosis paths, the analysis result corresponding to the fault type is determined. The multi-source feature vectors include both first and second feature vectors. Finally, based on the analysis results corresponding to the various fault types of the transformer, the operating status of the transformer is analyzed. In this way, by acquiring the transformer's multi-source feature vectors to analyze its operating status, a comprehensive judgment of the overall transformer status is achieved. Furthermore, when analyzing each fault type, by acquiring the second feature vectors corresponding to other fault diagnosis paths and the analysis data, information exchange and cross-validation among the fault diagnosis paths are achieved, resulting in higher accuracy in the transformer operation status analysis. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a flowchart illustrating a transformer operating status analysis method in one embodiment;

[0033] Figure 2 This is a flowchart illustrating the multi-source feature vector extraction steps in one embodiment;

[0034] Figure 3 This is a flowchart illustrating the discharge fault diagnosis path in one embodiment;

[0035] Figure 4 This is a flowchart illustrating an overheating fault diagnosis path in one embodiment;

[0036] Figure 5 This is a flowchart illustrating the fault detection steps for a target fault type in one embodiment.

[0037] Figure 6 This is a flowchart of a transformer operating status analysis method in one embodiment;

[0038] Figure 7 This is a flowchart illustrating the feature extraction steps in one embodiment;

[0039] Figure 8 A flowchart of a method for multiple fault paths in transformer operating state in one embodiment;

[0040] Figure 9 This is a structural block diagram of a transformer operating status analysis device in one embodiment;

[0041] Figure 10 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0043] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0044] In one exemplary embodiment, such as Figure 1 As shown, a method for analyzing the operating status of a transformer is provided. This embodiment illustrates the application of this method to a terminal. It is understood that this method can also be applied to a server, and to a system including both a terminal and a server, and implemented through interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0045] Step 101: Obtain the multi-source feature vector of the transformer.

[0046] Among them, the multi-source feature vector of the transformer is obtained, such as Figure 2 As shown, it may include:

[0047] Step 201: Obtain multi-source operation information of the transformer.

[0048] Among them, multi-source operation information includes data obtained from multiple aspects of the transformer, such as heterogeneous data collected through different sensors and systems, and data collected through unified communication protocols such as IEC 61850, thereby solving the problem of information silos in related technologies where different monitoring systems are independent of each other and the data lacks correlation.

[0049] Optionally, monitoring data can be collected in real-time or near real-time. For example, transformer operating condition data can be acquired, including load current, ambient temperature, cooler operating status, and other thermal conditions, as well as multi-point temperatures of key components such as transformer winding temperature, top oil temperature, bottom oil temperature, and core temperature. Electrical fault characteristic data of the transformer can also be collected, including data used to detect insulation degradation signals, such as partial discharge signals obtained through ultra-high frequency, ultrasonic, and high-frequency current sensors, the content and ratio of dissolved gases (H2, CH4, C2H2, C2H4, C2H6, etc.) in the oil, and the content of free gases. Transformer structural data can also be collected, including transformer structural information (such as winding type and insulation material), leakage flux signals (reflecting the condition of the core and clamps), and historical maintenance data (such as maintenance records and information on replaced components).

[0050] Then, the acquired multi-source operational information is cleaned and denoised. For example, for partial discharge signals, wavelet threshold denoising can be used to effectively remove white noise interference and retain the true discharge pulse. For multi-point temperature and dissolved gas data, a sliding time window method combined with the Laida criterion can be used to identify and remove transient interference and gross errors.

[0051] Step 202: For each type of operational information, extract the feature vector of the operational information to obtain a multi-source feature vector.

[0052] For example, for each type of operational information after the aforementioned data cleaning and noise reduction, feature extraction is performed. Through a series of targeted signal processing and mathematical transformation methods, various types of raw monitoring data are converted into high-dimensional, standardized numerical features with clear physical meaning, forming multi-source feature vectors for subsequent analysis of the transformer's operating status. For instance, feature extraction is performed on partial discharge signals to obtain a partial discharge feature vector; feature extraction is performed on dissolved gas information in oil to obtain an oil chromatography feature vector; feature extraction is performed on temperature information to obtain a thermal feature vector; feature extraction is performed on mechanical vibration signals to obtain a mechanical vibration feature vector; feature extraction is performed on leakage magnetic flux signals to obtain a leakage magnetic flux feature vector; feature extraction is performed on operating condition data to obtain an electrical stress feature vector; and feature extraction is performed on thermal condition signals to obtain an external heat dissipation capacity feature vector. All feature vectors obtained from feature extraction are used as multi-source feature vectors.

[0053] Optionally, examples of some feature extraction processes are provided. For partial discharge signals, a phase-resolved partial discharge spectrum is generated, and its basic triplet—including discharge amplitude, main discharge phase, and pulse count—is extracted. Higher-order statistical features describing the spectrum distribution are then calculated, such as skewness (reflecting asymmetry), steepness (reflecting concentration), and the cross-correlation coefficients of positive and negative half-cycle discharge pulse sequences, thus forming a partial discharge feature vector characterizing the discharge mode. For dissolved gases in oil, based on monitoring the absolute concentrations and gas production rates of gases such as H2, CH4, C2H2, C2H4, and C2H6, key ratio combinations (such as C2H2 / C2H4, CH4 / H2, and C2H4 / C2H6) are calculated. A discrete fault type code is generated using a modified three-ratio method, and its normalized coordinates in the David triangle are calculated, forming a multidimensional oil chromatographic feature that integrates concentration, ratio, and code, resulting in an oil chromatographic feature vector. For temperature signals, the temperature rise of the top oil layer, the temperature of hot spots in the windings, and the temperature differences between key components (such as the temperature difference between the top and bottom of the oil layer and the temperature difference between phases) are calculated. The average and standard deviation of all temperature measurement points are then statistically analyzed to quantify the uniformity of the temperature field, ultimately forming a thermal feature vector. For mechanical vibration signals, after Fourier transform, the amplitude, phase, and phase difference between the three phases centered at 100Hz and 200Hz are accurately extracted. The harmonic amplitude ratio and the total effective value of vibration are also calculated to form a mechanical vibration feature vector. Simultaneously, the leakage magnetic flux density distribution map is generated from the data of the leakage magnetic flux sensor array after spatial interpolation. Its average density, maximum density, spatial gradient, and difference from historical benchmarks are then calculated to form a leakage magnetic flux feature vector. For operating condition data, i.e., the processing of operating condition signals, the time-series statistics of the standardized load rate (maximum value, mean, standard deviation, and rate of change), as well as power quality indicators such as three-phase current imbalance, voltage deviation, and total harmonic distortion rate, are calculated to form an electrical stress feature vector. The acquisition of thermal condition signals directly relies on ambient temperature sensors, cooler control systems, and load monitoring units installed around the transformer. The acquired raw parameters include real-time ambient temperature measurements, cooler group start / stop status signals (or fan / oil pump operating current and speed feedback), and the transformer's current load current. Based on the principle of thermal balance, by inputting real-time load current and ambient temperature, and combining the cooler's operating status (such as the number of operating groups and start / stop mode), the equivalent heat dissipation coefficient is calculated online using a preset heat dissipation characteristic formula or a simplified heat transfer model. This coefficient dynamically reflects the comprehensive efficiency of the cooling system. The equivalent heat dissipation coefficient value is combined with the weighted average and extreme values ​​of the ambient temperature calculated through a sliding time window, as well as the quantified cooler operating rate, to jointly constitute an external heat dissipation capacity feature vector characterizing the transformer's external heat dissipation capacity.

[0054] Step 102: For each fault type of the transformer, determine the analysis result corresponding to the fault type based on at least one first feature vector corresponding to the fault type, the second feature vector corresponding to other fault diagnosis paths, and the analysis data generated by the second feature vector in other fault diagnosis paths.

[0055] The multi-source feature vector includes a first feature vector and a second feature vector. Transformer fault types can include discharge faults, overheating faults, and mechanical faults. Discharge faults are diagnosed using a discharge fault diagnosis path, overheating faults using an overheating fault diagnosis path, and mechanical faults using a mechanical fault diagnosis path. The first feature vector can be a feature vector directly related to the fault type, while the second feature vector can be another feature vector not directly related to the fault type but used to assist in the analysis and location of the fault type. Information exchange and cross-validation are performed on the analysis data generated by the first, second, and third feature vectors in other fault diagnosis paths to determine the analysis result corresponding to the fault type.

[0056] Optionally, the analysis data corresponding to the first feature vector, the second feature vector, and the second feature vector can be input into the analysis model corresponding to the fault type to obtain the analysis result corresponding to the fault type. The analysis model corresponding to each fault type can be obtained by training a model based on deep neural networks, support vector machines, etc. The analysis model corresponding to each fault type is obtained by training the initial model based on historical feature data and corresponding fault type annotation results.

[0057] Step 103: Based on the analysis results corresponding to various fault types of the transformer, analyze the operating status of the transformer.

[0058] Based on the analysis results of various fault types of transformers determined by multi-source feature vectors, such as discharge fault diagnosis results, overheating fault diagnosis results, and mechanical fault diagnosis results, the operating status of the transformer is determined. Based on the current status and historical data, the remaining life indicators of the transformer, such as dynamic endurance time, tripping risk probability, and maximum withstand current / number of trips, are determined by predictive models, providing forward-looking guidance for condition-based maintenance and operation and maintenance decisions.

[0059] In the above embodiments, firstly, multi-source feature vectors of the transformer are obtained. Then, for each fault type of the transformer, based on at least one first feature vector corresponding to the fault type, second feature vectors corresponding to other fault diagnosis paths, and analysis data generated by the second feature vectors in other fault diagnosis paths, the analysis result corresponding to the fault type is determined. The multi-source feature vectors include first and second feature vectors. Finally, based on the analysis results corresponding to the multiple fault types of the transformer, the operating status of the transformer is analyzed. In this way, by obtaining the multi-source feature vectors of the transformer to analyze its operating status, a comprehensive judgment of the overall transformer status is achieved. Simultaneously, when analyzing each fault type, by obtaining the second feature vectors corresponding to other fault diagnosis paths and the analysis data, information interaction and cross-validation among the fault diagnosis paths are achieved, resulting in higher accuracy in the analysis of the transformer's operating status.

[0060] In one embodiment, if the fault type is a discharge fault, the discharge fault is diagnosed through a discharge fault diagnosis path. Optionally, the first feature vector corresponding to the discharge fault includes a partial discharge feature vector and an oil chromatography feature vector. The fault type is determined to be either a discharge fault or a thermal fault based on the oil chromatography feature vector. If it is a discharge fault, the partial discharge feature vector and the oil chromatography feature vector are input into the discharge signal pattern recognition model corresponding to the discharge fault to identify the discharge type and whether a partial discharge fault exists. The discharge type may include corona discharge, surface discharge, internal discharge, arc discharge, breakdown discharge, etc. At the same time, the severity and trend of the corresponding fault are output.

[0061] Optionally, the discharge signal pattern recognition model can be based on a deep convolutional neural network or an attention-based classifier. By training the classifier network with a massive amount of labeled typical discharge PRPD spectra (images or feature vectors), a discharge signal pattern recognition model is obtained, which directly outputs the preliminary probability distribution of the discharge type. Simultaneously, after identifying a specific discharge fault, preliminary electrical localization based on the UHF signal time difference method can be used to locate the fault. By arranging UHF sensors in an array, the electromagnetic waves generated by the discharge pulse propagate to each sensor with a nanosecond-level time difference. By collecting the time when each sensor receives the same pulse, and based on the equivalent propagation speed of light in the oil-paper insulating composite medium, the three-dimensional spatial coordinates or region of the discharge activity can be calculated, thus obtaining the electrical localization result of the discharge fault.

[0062] For example, to achieve accurate analysis of discharge faults, cross-validation with other fault diagnosis paths can be used to accurately locate the discharge fault, such as... Figure 3 As shown, the method also includes:

[0063] Step 301: Obtain the winding deformation analysis information of the transformer in the mechanical fault diagnosis path.

[0064] For example, when performing fault analysis in a mechanical fault diagnosis path, the transformer winding deformation analysis information is stored in a shared evidence pool. This shared evidence pool is used to save intermediate diagnostic results and feature vectors generated during the diagnosis process of each fault diagnosis path, enabling information exchange and cross-validation among the fault diagnosis paths. The discharge fault diagnosis path obtains the transformer winding deformation analysis information through the shared evidence pool.

[0065] Step 302: If the winding deformation analysis information indicates that the transformer windings are deformed, then the mechanical vibration characteristic vector of the transformer is determined as the second characteristic vector.

[0066] If the winding deformation analysis information indicates that the transformer windings are deformed, the mechanical vibration characteristic vector of the transformer is determined as the second characteristic vector. By analyzing the analysis data corresponding to the second characteristic vector and inputting it into the discharge fault location model, the discharge fault can be accurately located, and the analysis results corresponding to the fault type can be obtained. Assume that the initial location is near the "A-phase winding" using UHF signals, and the analysis data corresponding to the second characteristic vector indicates that "the A-phase winding has moderate radial deformation." Since the conclusions of the two independent diagnostic paths highly overlap spatially, the discharge fault location model will significantly improve the confidence of this location result and comprehensively diagnose it as "mechanical deformation leading to field strength concentration, triggering partial discharge." Simultaneously, it outputs more accurate deformation point coordinates to correct or refine the location results of the UHF signals.

[0067] In the above embodiments, by interacting with information from other paths, the deformation information obtained from the mechanical fault diagnosis path is used in the discharge fault diagnosis path to focus on the specific discharge mode caused by electric field distortion due to mechanical deformation. If the ultra-high frequency signal characteristics are highly correlated with the deformation location and type, the causal chain of "mechanical deformation inducing insulation fault" is jointly confirmed, thereby improving the accuracy of discharge fault diagnosis.

[0068] In the embodiments of this application, if the fault type is an overheating fault, such as Figure 4 As shown, the method also includes:

[0069] Step 401: Obtain dissolved gas analysis information in the discharge fault diagnosis path and / or state analysis information in the mechanical fault diagnosis path.

[0070] If the discharge fault diagnosis path determines the fault type to be a thermal fault based on the oil chromatography feature vector, the dissolved gas analysis information is stored in the shared evidence pool. Simultaneously, the mechanical fault diagnosis path, upon determining that the transformer is under abnormal operating conditions, stores the condition analysis information in the shared evidence pool. The overheating fault diagnosis path is used to diagnose overheating faults, acquiring dissolved gas analysis information and condition analysis information from the shared evidence pool.

[0071] Step 402: If the dissolved gas analysis information indicates that the transformer has a thermal fault, then the oil chromatography feature vector is determined as the second feature vector, and / or, if the condition analysis information indicates that the transformer is in an abnormal working state, then the electrical stress feature vector and the leakage flux feature vector are determined as the second feature vector.

[0072] For example, if the transformer has a thermal fault, the above-mentioned oil chromatography feature vector is determined as the second feature vector, and / or the transformer is in an abnormal operating state, the electrical stress feature vector and the leakage flux feature vector are determined as the second feature vector.

[0073] The analysis data corresponding to the second feature vector and the first feature vector corresponding to the overheating fault, such as the external heat dissipation capacity feature vector, are input into the temperature prediction model corresponding to the overheating fault. This model is established based on historical data and thermodynamic principles, and can predict the theoretical temperature of the transformer under the current operating conditions. By comparing the theoretical temperature with the measured temperature of the transformer, multi-parameter fusion algorithms such as weighted fusion and DS evidence theory are used to determine whether the transformer has an overheating fault and the type of overheating fault.

[0074] Overheating faults are categorized into thermal anomalies caused by abnormal operating conditions and thermal anomalies caused by component overheating. If the thermal anomaly is determined to be caused by abnormal operating conditions, the temperature prediction model outputs the cause of the thermal anomaly, such as the presence of harmonics, poor heat dissipation, overload, and demagnetization, and outputs the fault level, including low overheating, moderate overheating, or high overheating, while also outputting the analyzed fault trend, such as dynamic endurance time. If the thermal anomaly is determined to be caused by component overheating, the model analyzes whether the thermal fault originates from windings (e.g., poor conductor contact), structural components (e.g., overheating of the iron core), or other components, and outputs the faulty component, fault level, and fault trend prediction.

[0075] Optionally, the temperature prediction model may consist of multiple sub-modules. The temperature prediction sub-model outputs a theoretical temperature field based on the input feature vector. The thermal state deviation analysis module compares the theoretical temperature with the measured temperature, generating a thermal state deviation feature vector. This feature vector may include the average temperature deviation reflecting the overall temperature prediction error level, the maximum local deviation used to capture the most severe local overheating or temperature anomalies, and the spatial temperature field distortion index indicating the presence of local heat accumulation. The root cause determination and fusion module receives the thermal state deviation feature vector and combines it with the input first and second feature vectors and analysis data to determine the overheating fault type. The abnormal operating condition analysis module determines the cause of the overheating fault based on the overheating fault type and a built-in rule base. The component overheating location and cross-validation module outputs the final output of the temperature prediction model based on the analysis results in a shared evidence pool.

[0076] In the above embodiments, by interacting with information from other paths, the accuracy of overheating fault diagnosis is improved by utilizing the analytical information from the discharge fault diagnosis path and the mechanical fault diagnosis path in the overheating fault diagnosis path.

[0077] In one embodiment, the mechanical fault diagnosis path is used to diagnose mechanical faults. Based on leakage flux feature vectors, mechanical vibration feature vectors, transformer structural information, and maintenance data, it determines whether the transformer is under abnormal operating conditions using an abnormal operating condition judgment model corresponding to the mechanical fault. If the determination is "yes", correction parameters are output to the shared evidence pool, abnormal operating conditions are output, and input to the temperature prediction model.

[0078] The correction parameters are one or a set of parameters used to convert changes in the mechanical state into quantified values ​​that the temperature prediction model can understand and use, thereby correcting the prediction results of the temperature prediction model. The mechanical state of the transformer windings is directly related to their heat dissipation capacity, and the correction parameters serve as a bridge to quantify this relationship. Correction parameters may include: equivalent thermal resistance increment, which represents how much the internal thermal resistance and the thermal resistance between the windings and the oil passages have increased due to impact, loosening, or deformation of the windings. Increased thermal resistance means that heat is more difficult to dissipate. Oil passage blockage factor, a coefficient between 0 and 1, characterizes the degree to which winding deformation reduces the effective flow area of ​​the cooling oil passages. Hot spot location offset vector, which indicates the possible direction and distance of movement of the hot spot in the three-dimensional space of the windings if mechanical defects such as local deformation cause changes in the current density or heat dissipation conditions in the windings. Other parameters may also be included in the correction parameters, which are not limited in this application. When the mechanical fault diagnosis path detects that the transformer is under abnormal operating conditions, it determines the correction parameters corresponding to the abnormal operating conditions based on the built-in rule base and saves the correction parameters to the shared evidence pool for use by other fault diagnosis paths.

[0079] If the result is "No", the winding defect identification and location model is activated. This model integrates mechanical vibration feature vectors, leakage flux feature vectors, and short-circuit electrodynamic calculations to determine whether a short circuit has occurred and identify the defect location. The model uses a convolutional neural network to identify whether the winding has undergone radial or axial deformation and assesses the degree of deformation (e.g., slight, moderate, severe). Simultaneously, by analyzing parameters such as three-phase current imbalance, loop resistance, and short-circuit electrodynamics, it diagnoses whether a short-circuit fault has occurred, such as inter-turn or inter-layer short circuits, and predicts the dynamic endurance time before tripping. Finally, the model outputs the fault type (radial deformation, axial deformation, short circuit), fault severity (e.g., slight, moderate, severe), fault trend (predicted dynamic endurance time before tripping, number of withstand cycles), and fault location.

[0080] For example, an abnormal operating condition determination model is used to determine whether a transformer is currently or has been subjected to abnormal electrodynamic shocks that may cause mechanical damage, such as near-zone short circuits, external short circuits, frequent inrush currents, or severe three-phase imbalance. The abnormal operating condition determination model can be a hybrid judgment system based on a rule engine and a statistical model. This model has a built-in hierarchical rule base that sets thresholds from multiple dimensions and outputs a determination conclusion regarding the abnormal operating condition based on the input feature vector.

[0081] In one embodiment, such as Figure 5 As shown, the method also includes:

[0082] Step 501: If fault information of the target fault type is detected, obtain the detection results of other fault diagnosis paths.

[0083] Among them, the target fault type can be a severe discharge fault, such as arc discharge or breakdown discharge. In this case, the detection results of other fault diagnosis paths are obtained through a shared evidence pool to jointly assess the fault risk of the transformer.

[0084] Step 502: Based on the detection results of other fault diagnosis paths, determine the fault cause and handling method corresponding to the fault information.

[0085] For example, by retrieving transformer design parameters such as short-circuit impedance from the structural information and combining them with the current load, the maximum withstand current or the number of arc discharges that have occurred can be quickly assessed, providing the most critical basis for deciding whether to immediately shut down the system. Optionally, the system can query whether there is severe overheating in the overheating fault diagnosis path that could lead to insulation carbonization, whether there is severe deformation in the mechanical fault diagnosis path that could lead to short circuits, and whether there is a trend of the discharge fault diagnosis path evolving into a breakdown fault.

[0086] In the above embodiments, by quickly obtaining the detection results of other fault diagnosis paths, the root cause of the target fault type can be quickly located, and it can be distinguished whether the fault is caused by overheating, deformation or insulation breakdown, providing key basis for emergency response and root cause analysis.

[0087] In the embodiments of this application, as shown Figure 6 The diagram shown is a flowchart of a transformer operating status analysis method. Some feature extraction processes in this method can be described as follows: Figure 7 As shown. Figure 8 This document presents flowcharts for the diagnostic paths of discharge faults, overheating faults, and mechanical faults. Using the aforementioned operational status analysis methods, the operational status analysis results of the transformer are obtained, generating a structured diagnostic report. Optionally, the report content may include: 1. Fault Type: Clearly listing all identified faults, including abnormal operating conditions such as harmonics, demagnetization, overload, poor heat dissipation, winding faults, structural component faults, partial discharge faults, breakdown faults, arcing faults, winding deformation, short circuits, etc. 2. Fault Severity: Quantifying the level of each fault type, such as overheating classified as "low / medium / high," deformation as "minor / moderate / severe," and discharge as "minor / severe" defects. 3. Fault Trend: Based on the current status and historical data, predicting the remaining life indicators of the equipment, such as dynamic endurance time, tripping risk probability, and maximum withstand current / number of trips, providing forward-looking guidance for condition-based maintenance and operation and maintenance decisions.

[0088] In the above embodiments, a multi-physical quantity linkage diagnostic method is proposed. Through a multi-path parallel diagnosis and cross-validation mechanism for overheating faults, discharge faults, and mechanical faults in transformers, multi-source information fusion and precise fault location are achieved, and the fault severity and dynamic trend prediction are output. This realizes an upgrade of the operation and maintenance strategy from post-maintenance and preventative maintenance to predictive maintenance.

[0089] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0090] Based on the same inventive concept, this application also provides a transformer operating state analysis device for implementing the transformer operating state analysis method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more embodiments of the transformer operating state analysis device provided below can be found in the limitations of the transformer operating state analysis method described above, and will not be repeated here.

[0091] In one exemplary embodiment, such as Figure 9 As shown, a transformer operating status analysis device is provided, comprising: an acquisition module, a first determination module, and an analysis module, wherein:

[0092] The acquisition module is used to acquire the multi-source feature vector of the transformer;

[0093] The first determining module is used to determine the analysis result corresponding to the fault type for each type of transformer fault, based on at least one first feature vector corresponding to the fault type, second feature vectors corresponding to other fault diagnosis paths, and analysis data generated by the second feature vectors in other fault diagnosis paths; the multi-source feature vectors include the first feature vector and the second feature vector.

[0094] The analysis module is used to analyze the operating status of the transformer based on the analysis results corresponding to various fault types of the transformer.

[0095] In one embodiment, the first determining module is specifically used to input the analysis data corresponding to the first feature vector, the second feature vector, and the second feature vector into the analysis model corresponding to the fault type to obtain the analysis result corresponding to the fault type.

[0096] In one embodiment, if the fault type is a discharge fault, the device further includes a second determining module for obtaining winding deformation analysis information of the transformer in the mechanical fault diagnosis path; if the winding deformation analysis information indicates that the winding of the transformer is deformed, the mechanical vibration feature vector of the transformer is determined as the second feature vector.

[0097] In one embodiment, if the fault type is an overheating fault, the device further includes a third determining module for acquiring dissolved gas analysis information in the discharge fault diagnosis path and / or state analysis information in the mechanical fault diagnosis path; if the dissolved gas analysis information indicates that the transformer has a thermal fault, then the oil chromatography feature vector is determined as the second feature vector, and / or if the state analysis information indicates that the transformer is in an abnormal working state, then the electrical stress feature vector and the leakage flux feature vector are determined as the second feature vector.

[0098] In one embodiment, the device further includes a fourth determining module, which is used to obtain the detection results of other fault diagnosis paths if fault information of the target fault type is detected; and to determine the fault cause and handling method corresponding to the fault information based on the detection results of other fault diagnosis paths.

[0099] In one embodiment, the acquisition module is specifically used to acquire multi-source operating information of the transformer; for each type of operating information, the feature vector of the operating information is extracted to obtain a multi-source feature vector.

[0100] Each module in the aforementioned transformer operation status analysis device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0101] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 10 As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When executed by the processor, the computer program implements a transformer operating state analysis method. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0102] Those skilled in the art will understand that Figure 10The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0103] In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to perform the following steps: acquiring multi-source feature vectors of a transformer; for each fault type of the transformer, determining the analysis result corresponding to the fault type based on at least one first feature vector corresponding to the fault type, a second feature vector corresponding to other fault diagnosis paths, and analysis data generated by the second feature vector in other fault diagnosis paths; the multi-source feature vectors include a first feature vector and a second feature vector; and analyzing the operating status of the transformer based on the analysis results corresponding to the multiple fault types of the transformer.

[0104] In one embodiment, when the processor executes the computer program, it further performs the following steps: inputting the analysis data corresponding to the first feature vector, the second feature vector, and the second feature vector into the analysis model corresponding to the fault type, and obtaining the analysis result corresponding to the fault type.

[0105] In one embodiment, if the fault type is a discharge fault, the processor, when executing the computer program, also performs the following steps: obtaining the winding deformation analysis information of the transformer in the mechanical fault diagnosis path; if the winding deformation analysis information indicates that the transformer winding has deformation, then the mechanical vibration feature vector of the transformer is determined as the second feature vector.

[0106] In one embodiment, if the fault type is an overheating fault, the processor, when executing the computer program, further implements the following steps: obtaining dissolved gas analysis information in the discharge fault diagnosis path, and / or, state analysis information in the mechanical fault diagnosis path; if the dissolved gas analysis information indicates that the transformer has a thermal fault, then the oil chromatography feature vector is determined as the second feature vector, and / or, if the state analysis information indicates that the transformer is in an abnormal working state, then the electrical stress feature vector and the leakage flux feature vector are determined as the second feature vector.

[0107] In one embodiment, when the processor executes the computer program, it further performs the following steps: if fault information of the target fault type is detected, it obtains the detection results of other fault diagnosis paths; based on the detection results of other fault diagnosis paths, it determines the fault cause and handling method corresponding to the fault information.

[0108] In one embodiment, when the processor executes the computer program, it also performs the following steps: acquiring multi-source operating information of the transformer; for each type of operating information, extracting the feature vector of the operating information to obtain a multi-source feature vector.

[0109] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, it performs the following steps: acquiring multi-source feature vectors of a transformer; for each fault type of the transformer, determining the analysis result corresponding to the fault type based on at least one first feature vector corresponding to the fault type, a second feature vector corresponding to other fault diagnosis paths, and analysis data generated by the second feature vector in other fault diagnosis paths; the multi-source feature vectors include a first feature vector and a second feature vector; and analyzing the operating status of the transformer based on the analysis results corresponding to the multiple fault types of the transformer.

[0110] In one embodiment, when the computer program is executed by the processor, it performs the following steps: inputting the analysis data corresponding to the first feature vector, the second feature vector, and the second feature vector into the analysis model corresponding to the fault type, and obtaining the analysis result corresponding to the fault type.

[0111] In one embodiment, if the fault type is a discharge fault, the computer program, when executed by the processor, performs the following steps: obtaining the winding deformation analysis information of the transformer in the mechanical fault diagnosis path; if the winding deformation analysis information indicates that the transformer winding is deformed, then the mechanical vibration feature vector of the transformer is determined as the second feature vector.

[0112] In one embodiment, if the fault type is an overheating fault, the computer program, when executed by the processor, performs the following steps: acquiring dissolved gas analysis information in the discharge fault diagnosis path, and / or, state analysis information in the mechanical fault diagnosis path; if the dissolved gas analysis information indicates that the transformer has a thermal fault, then the oil chromatography feature vector is determined as the second feature vector, and / or, if the state analysis information indicates that the transformer is in an abnormal operating state, then the electrical stress feature vector and the leakage flux feature vector are determined as the second feature vector.

[0113] In one embodiment, when the computer program is executed by the processor, it performs the following steps: if fault information of the target fault type is detected, the detection results of other fault diagnosis paths are obtained; based on the detection results of other fault diagnosis paths, the fault cause and handling method corresponding to the fault information are determined.

[0114] In one embodiment, when the computer program is executed by the processor, it performs the following steps: acquiring multi-source operating information of the transformer; for each type of operating information, extracting the feature vector of the operating information to obtain a multi-source feature vector.

[0115] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps: acquiring multi-source feature vectors of a transformer; for each fault type of the transformer, determining an analysis result corresponding to the fault type based on at least one first feature vector corresponding to the fault type, a second feature vector corresponding to other fault diagnosis paths, and analysis data generated by the second feature vector in other fault diagnosis paths; the multi-source feature vectors include a first feature vector and a second feature vector; and analyzing the operating status of the transformer based on the analysis results corresponding to the multiple fault types of the transformer.

[0116] In one embodiment, when the computer program is executed by the processor, it performs the following steps: inputting the analysis data corresponding to the first feature vector, the second feature vector, and the second feature vector into the analysis model corresponding to the fault type, and obtaining the analysis result corresponding to the fault type.

[0117] In one embodiment, if the fault type is a discharge fault, the computer program, when executed by the processor, performs the following steps: obtaining the winding deformation analysis information of the transformer in the mechanical fault diagnosis path; if the winding deformation analysis information indicates that the transformer winding is deformed, then the mechanical vibration feature vector of the transformer is determined as the second feature vector.

[0118] In one embodiment, if the fault type is an overheating fault, the computer program, when executed by the processor, performs the following steps: acquiring dissolved gas analysis information in the discharge fault diagnosis path, and / or, state analysis information in the mechanical fault diagnosis path; if the dissolved gas analysis information indicates that the transformer has a thermal fault, then the oil chromatography feature vector is determined as the second feature vector, and / or, if the state analysis information indicates that the transformer is in an abnormal operating state, then the electrical stress feature vector and the leakage flux feature vector are determined as the second feature vector.

[0119] In one embodiment, when the computer program is executed by the processor, it performs the following steps: if fault information of the target fault type is detected, the detection results of other fault diagnosis paths are obtained; based on the detection results of other fault diagnosis paths, the fault cause and handling method corresponding to the fault information are determined.

[0120] In one embodiment, when the computer program is executed by the processor, it performs the following steps: acquiring multi-source operating information of the transformer; for each type of operating information, extracting the feature vector of the operating information to obtain a multi-source feature vector.

[0121] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0122] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0123] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0124] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for analyzing the operating status of a transformer, characterized in that, The method includes: Obtain the multi-source feature vector of the transformer; For each fault type of the transformer, the analysis result corresponding to the fault type is determined based on at least one first feature vector corresponding to the fault type, a second feature vector corresponding to other fault diagnosis paths, and analysis data generated by the second feature vector in other fault diagnosis paths; the multi-source feature vector includes the first feature vector and the second feature vector; Based on the analysis results corresponding to various fault types of the transformer, the operating status of the transformer is analyzed.

2. The method according to claim 1, characterized in that, The step of determining the analysis result corresponding to the fault type based on at least one first feature vector corresponding to the fault type, a second feature vector related to the fault type, and analysis data generated by the second feature vector in other fault diagnosis paths includes: The first feature vector, the second feature vector, and the analysis data corresponding to the second feature vector are input into the analysis model corresponding to the fault type to obtain the analysis result corresponding to the fault type.

3. The method according to claim 2, characterized in that, If the fault type is a discharge fault, the method further includes: Obtain the winding deformation analysis information of the transformer mentioned in the mechanical fault diagnosis path; If the winding deformation analysis information indicates that the transformer windings are deformed, then the mechanical vibration characteristic vector of the transformer is determined as the second characteristic vector.

4. The method according to claim 2, characterized in that, If the fault type is an overheating fault, the method further includes: Obtain dissolved gas analysis information in the discharge fault diagnosis path, and / or, state analysis information in the mechanical fault diagnosis path; If the dissolved gas analysis information indicates that the transformer has a thermal fault, then the oil chromatography feature vector is determined as the second feature vector, and / or, if the state analysis information indicates that the transformer is in an abnormal operating state, then the electrical stress feature vector and the leakage flux feature vector are determined as the second feature vector.

5. The method according to claim 1, characterized in that, The method further includes: If fault information of the target fault type is detected, the detection results of other fault diagnosis paths are obtained; Based on the detection results of the other fault diagnosis paths, the cause of the fault and the handling method corresponding to the fault information are determined.

6. The method according to claim 1, characterized in that, The acquisition of the multi-source feature vector of the transformer includes: Obtain multi-source operating information of the transformer; For each type of operational information, the feature vector of the operational information is extracted to obtain the multi-source feature vector.

7. A transformer operating status analysis device, characterized in that, The device includes: The acquisition module is used to acquire the multi-source feature vector of the transformer; The determination module is used to determine the analysis result corresponding to each fault type of the transformer based on at least one first feature vector corresponding to the fault type, a second feature vector corresponding to other fault diagnosis paths, and analysis data generated by the second feature vector in other fault diagnosis paths; the multi-source feature vector includes the first feature vector and the second feature vector; The analysis module is used to analyze the operating status of the transformer based on the analysis results corresponding to various fault types of the transformer.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images