Fault detection method, device and equipment of transformer and transformer
By deploying multiple sensors in instrument transformers and utilizing deep learning models and dynamic threshold technology, the problems of high cost and low accuracy in instrument transformer fault detection have been solved, enabling automated, real-time fault diagnosis and maintenance, and ensuring the stability of the power system.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- MEIZHOU POWER SUPPLY BUREAU OF GUANGDONG POWER GRID CORP
- Filing Date
- 2026-01-16
- Publication Date
- 2026-05-29
AI Technical Summary
Existing technologies for detecting instrument transformer faults are costly, susceptible to subjective factors, and produce inaccurate results, making it difficult to meet real-time and comprehensive requirements and posing safety hazards.
By deploying multiple sensors in the current transformer to collect data in real time, using deep learning models to extract feature information, and combining historical data and equipment-related information to determine dynamic fault thresholds, fault scoring and type identification can be achieved.
It enables automatic diagnosis of transformer faults, improves the comprehensiveness and accuracy of detection, meets the needs of real-time detection, saves computing resources, and ensures the stable operation of the power system.
Smart Images
Figure CN122109964A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of power system technology, and in particular to a fault detection method, device, equipment and instrument for an instrument transformer. Background Technology
[0002] In power systems, instrument transformers are core devices that ensure the accuracy of power measurement, relay protection, and metering. Their operating status directly affects the stability of the power system. Therefore, it is crucial to detect potential faults in instrument transformers.
[0003] Generally, instrument transformers can be inspected manually to detect faults in a timely manner and ensure their stable operation. However, this method is not only costly but also susceptible to subjective influences, leading to missed detections and affecting the accuracy of instrument transformer testing.
[0004] Therefore, how to comprehensively, accurately, and efficiently test current transformers has become an urgent technical problem to be solved. Summary of the Invention
[0005] This application provides a fault detection method, apparatus, device, and instrument for instrument transformers, which can comprehensively, accurately, and efficiently detect faults in instrument transformers, thereby helping to ensure the stable operation of the power system.
[0006] In a first aspect, embodiments of this application provide a fault detection method for a current transformer, applied to a processor; the method includes:
[0007] Acquire real-time sensor data from multiple sensors deployed in the current transformer;
[0008] The fault score of the current transformer is determined based on the first feature information corresponding to the multiple sensor data; wherein, the first feature information indicates the fusion feature of the multiple sensor data;
[0009] The historical sensor data of the current transformer is obtained, and the dynamic fault threshold corresponding to the current transformer is determined based on the various sensor data, the historical sensor data and the device-related information of the current transformer.
[0010] If the current transformer is determined to be faulty based on the fault score and the dynamic fault threshold, the fault type of the current transformer is determined based on the second feature information corresponding to the multiple sensor data; wherein, the second feature information indicates the data characteristics corresponding to the multiple sensor data.
[0011] Optionally, based on the first feature information corresponding to the data from the multiple sensors, a fault score for the current transformer is determined, including:
[0012] Based on the first model, local feature extraction is performed on the data from the various sensors to obtain local feature information;
[0013] According to the second model, temporal feature extraction is performed on the local feature information to obtain the first feature information; wherein, the first feature information is a multi-dimensional vector;
[0014] The fault score is obtained by weighted summation of each dimension in the multi-dimensional vector.
[0015] Optionally, the equipment-related information includes, but is not limited to: the operating environment information of the current transformer, the service life information of the current transformer, the operating time information of the current transformer, the model information of the current transformer, and the manufacturer information of the current transformer.
[0016] Optionally, based on the various sensor data, the historical sensor data, and the device-related information of the current transformer, the dynamic fault threshold corresponding to the current transformer is determined, including:
[0017] Based on the historical sensor data and the device-related information corresponding to the historical sensor data, a third model is trained to obtain a trained third model;
[0018] The dynamic fault threshold is determined based on the trained third model, the various sensor data, and the device-related information corresponding to the various sensor data.
[0019] Optionally, the sensor data includes at least: current data, voltage data, temperature data, partial discharge signal data, and vibration signal data.
[0020] Optionally, the number of current transformers is multiple, and different types of sensors are used in different current transformers to acquire the same type of sensor data; acquiring multiple types of sensor data collected in real time by multiple sensors deployed in the current transformers, including:
[0021] Acquire initial data collected in real time by the various sensors deployed in the current transformer;
[0022] Determine the corresponding data processing method based on the type of sensor that acquires the same sensor data;
[0023] The initial data is processed according to the corresponding data processing method to obtain the sensor data.
[0024] Optionally, based on the second feature information corresponding to the various sensor data, the fault type of the current transformer is determined, including:
[0025] Based on the data type of the sensor data, feature extraction processing is performed on the sensor data to obtain the second feature information corresponding to each type of sensor data;
[0026] The second feature information is classified and processed according to the fourth model to obtain the fault type of the current transformer.
[0027] Secondly, embodiments of this application provide a fault detection device for a current transformer, applied to a processor; the device includes:
[0028] The acquisition unit is used to acquire various sensor data collected in real time by various sensors deployed in the current transformer;
[0029] The first determining unit is configured to determine the fault score of the current transformer based on the first feature information corresponding to the multiple sensor data; wherein the first feature information indicates the fusion feature of the multiple sensor data;
[0030] The second determining unit is used to acquire historical sensor data of the current transformer, and determine the dynamic fault threshold corresponding to the current transformer based on the various sensor data, the historical sensor data and the device-related information of the current transformer.
[0031] The fault processing unit is configured to determine the fault type of the current transformer based on the second feature information corresponding to the multiple sensor data if the current transformer is determined to be faulty based on the fault score and the dynamic fault threshold; wherein the second feature information indicates the data characteristics corresponding to the multiple sensor data.
[0032] Optionally, the first determining unit is used for:
[0033] Based on the first model, local feature extraction is performed on the data from the various sensors to obtain local feature information;
[0034] According to the second model, temporal feature extraction is performed on the local feature information to obtain the first feature information; wherein, the first feature information is a multi-dimensional vector;
[0035] The fault score is obtained by weighted summation of each dimension in the multi-dimensional vector.
[0036] Optionally, the equipment-related information includes, but is not limited to: the operating environment information of the current transformer, the service life information of the current transformer, the operating time information of the current transformer, the model information of the current transformer, and the manufacturer information of the current transformer.
[0037] Optionally, the second determining unit is used for:
[0038] Based on the historical sensor data and the device-related information corresponding to the historical sensor data, a third model is trained to obtain a trained third model;
[0039] The dynamic fault threshold is determined based on the trained third model, the various sensor data, and the device-related information corresponding to the various sensor data.
[0040] Optionally, the sensor data includes at least: current data, voltage data, temperature data, partial discharge signal data, and vibration signal data.
[0041] Optionally, the number of current transformers is multiple, and different types of sensors are used in different current transformers to acquire the same type of sensor data; in this case, the acquisition unit is used for:
[0042] Acquire initial data collected in real time by the various sensors deployed in the current transformer;
[0043] Determine the corresponding data processing method based on the type of sensor that acquires the same sensor data;
[0044] The initial data is processed according to the corresponding data processing method to obtain the sensor data.
[0045] Optional, a fault handling unit, used for:
[0046] Based on the data type of the sensor data, feature extraction processing is performed on the sensor data to obtain the second feature information corresponding to each type of sensor data;
[0047] The second feature information is classified and processed according to the fourth model to obtain the fault type of the current transformer.
[0048] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;
[0049] The memory stores computer-executed instructions;
[0050] The processor executes computer execution instructions stored in the memory, causing the processor to perform various possible implementations as described in the first aspect above.
[0051] Fourthly, embodiments of this application provide a current transformer, the current transformer comprising: a current transformer body, a plurality of sensors mounted on the current transformer body, a power supply module, and an electronic device as described in the third aspect; wherein the power supply module is used to supply power to at least one of the plurality of sensors, the electronic device, and the current transformer body.
[0052] Fifthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect described above.
[0053] In a sixth aspect, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.
[0054] The fault detection method, apparatus, device, and transformer provided in this application can determine the fault score of the transformer based on the first feature information corresponding to the various sensor data after acquiring multiple sensor data collected in real time by multiple sensors deployed in the transformer. This allows for real-time fault diagnosis of the transformer by combining multiple types of sensor data, meeting the real-time detection requirements of the transformer and improving the comprehensiveness and accuracy of fault diagnosis. Simultaneously, historical sensor data of the transformer can be acquired, and a dynamic fault threshold corresponding to the transformer can be determined based on the various sensor data, historical sensor data, and device-related information of the transformer, thereby determining a threshold that better matches the current state of the transformer. If a fault is determined in the transformer based on the fault score and dynamic fault threshold, the fault type of the transformer is determined based on the second feature information corresponding to the various sensor data. This implementation method, when a fault is determined in the transformer, further determines the fault type based on the data characteristics corresponding to the various sensor data, achieving automatic fault diagnosis of the transformer and improving efficiency. It also enables timely and accurate maintenance of the transformer based on the determined fault type. Furthermore, this implementation method allows for fault type detection only after a fault is identified, thereby saving computing resources and avoiding waste. Attached Figure Description
[0055] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.
[0056] Figure 1 A flowchart illustrating a fault detection method for a current transformer provided in an embodiment of this application;
[0057] Figure 2 A flowchart illustrating another fault detection method for a current transformer provided in an embodiment of this application;
[0058] Figure 3 This is a schematic diagram of the structure of a fault detection system for a current transformer provided in an embodiment of this application;
[0059] Figure 4 A schematic diagram of the structure of a fault detection device for a current transformer provided in an embodiment of this application;
[0060] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application;
[0061] Figure 6 This is a schematic diagram of the structure of a current transformer provided in an embodiment of this application.
[0062] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation
[0063] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.
[0064] In power systems, the operating status of instrument transformers directly affects the stability of the power system. Therefore, it is crucial to detect potential faults in instrument transformers.
[0065] In one implementation, instrument transformers can be inspected manually to detect faults promptly and ensure stable operation. However, this method is not only time- and labor-intensive, but also susceptible to subjective influences, leading to missed detections and affecting the accuracy of instrument transformer testing.
[0066] The inventors also discovered that fault detection of instrument transformers can be performed through periodic preventative tests. However, this approach typically detects a single fault type based on a single data type, failing to comprehensively assess the transformer's operational status and resulting in low detection efficiency and accuracy. Furthermore, since periodic preventative tests are usually conducted on a fixed schedule, they cannot meet the real-time monitoring requirements of instrument transformers. Moreover, instrument transformers may malfunction between tests, potentially leading to safety incidents. Therefore, this approach also presents certain safety hazards.
[0067] The fault detection method for current transformers provided in this application can perform real-time processing of sensor data collected by various types of sensors and detect faults through dynamically changing thresholds. This not only meets the real-time requirements but also improves the comprehensiveness and accuracy of fault detection, thereby solving the above-mentioned technical problems.
[0068] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.
[0069] Figure 1 This is a flowchart illustrating a fault detection method for a current transformer provided in an embodiment of this application, applied to a processor, such as... Figure 1 As shown, the method includes:
[0070] S101. Acquire multiple sensor data collected in real time from multiple sensors deployed in the current transformer.
[0071] In one example, different types of sensors can be deployed at different locations on the current transformer, and fault detection can be performed based on the sensor data collected by the different types of sensors deployed.
[0072] For example, different types of sensors can be deployed at key locations of the current transformer, such as the winding, core, insulation layer, and housing. For example, the deployed sensors may include, but are not limited to: current sensors (e.g., Rogowski coil current sensors), voltage sensors (e.g., capacitive voltage divider voltage sensors), temperature sensors (e.g., fiber Bragg grating-based temperature sensors), partial discharge sensors (e.g., ultrasonic partial discharge sensors), and vibration sensors (e.g., microelectromechanical vibration sensors).
[0073] Based on this, sensor data includes at least: current data, voltage data, temperature data, partial discharge signal data, and vibration signal data.
[0074] In one possible implementation, when acquiring real-time sensor data from multiple sensors deployed in the current transformer, the data can be converted from analog to digital and then filtered to remove noise and interference. After removing noise and interference, the data is then normalized to obtain the aforementioned sensor data. This allows for unified processing of different types of sensor data, facilitating more accurate data processing.
[0075] S102. Determine the fault score of the current transformer based on the first characteristic information corresponding to the data from multiple sensors.
[0076] The first feature information indicates the fusion characteristics of data from multiple sensors.
[0077] In one example, a deep learning model can be used to process data from multiple sensors to obtain fused features, or first feature information, which can then be used to determine the fault score of the transformer. The specific deep learning model used is not limited here; the key is whether it can be implemented.
[0078] In one example, before using a deep learning model to process data from multiple sensors, a training dataset can be constructed based on the various sensor data and their corresponding fault labels. The deep learning model can then be trained using this dataset, and the fusion features output by the deep learning model can be used to determine the fault score of the transformer.
[0079] S103. Obtain historical sensor data of the current transformer, and determine the dynamic fault threshold corresponding to the current transformer based on various sensor data, historical sensor data and equipment-related information of the current transformer.
[0080] In one example, historical sensor data of the current transformer can be obtained according to a first sliding window of a preset size. For example, the preset size of the first sliding window can be 5 hours, 24 hours, or 72 hours, etc. The preset size is not limited here, but is determined according to actual needs.
[0081] In one example, a dynamic fault threshold can be determined based on historical sensor data, real-time collected data from multiple sensors, and device-related information of the current transformer. This allows for a comprehensive consideration of the impact of sensor data and device-related information on the normal operation of the current transformer. Compared to thresholds determined based on fixed thresholds or only sensor data, the dynamic fault threshold determined in this embodiment is more accurate and flexible.
[0082] S104. If a fault is determined to exist in the current transformer based on the fault score and dynamic fault threshold, the fault type of the current transformer is determined based on the second characteristic information corresponding to the data from multiple sensors.
[0083] The second feature information indicates the data characteristics corresponding to data from multiple sensors.
[0084] In one example, multiple feature extraction methods can be used to extract feature data corresponding to multiple sensor data to obtain second feature information, and then the fault type of the current transformer can be determined based on the second feature information.
[0085] The fault detection method for current transformers provided in this application can determine a fault score of the current transformer based on the first feature information corresponding to the various sensor data after acquiring multiple sensor data collected in real time from multiple sensors deployed in the current transformer. This allows for real-time fault diagnosis of the current transformer by combining multiple types of sensor data, not only meeting the real-time detection requirements of the current transformer but also improving the comprehensiveness and accuracy of fault diagnosis. Simultaneously, historical sensor data of the current transformer can be acquired, and a dynamic fault threshold corresponding to the current transformer can be determined based on the multiple sensor data, historical sensor data, and equipment-related information of the current transformer. This allows for the determination of a threshold that better matches the current state of the current transformer. If a fault is determined in the current transformer based on the fault score and the dynamic fault threshold, the fault type of the current transformer is determined based on the second feature information corresponding to the multiple sensor data. This implementation method, when a fault is determined in the current transformer, can then determine the fault type based on the data characteristics corresponding to the multiple sensor data. This not only achieves automatic fault diagnosis of the current transformer, improving efficiency, but also allows for timely and accurate maintenance of the current transformer based on the determined fault type. Furthermore, this implementation method allows for fault type detection only after a fault is identified, thereby saving computing resources and avoiding waste.
[0086] Figure 2 A flowchart illustrating another method for fault detection of a current transformer provided in this application embodiment is shown below. Figure 2 As shown, in this embodiment... Figure 1 Based on the embodiments, a fault detection method for current transformers is described in detail. This method is applied to a processor and includes:
[0087] S201. Acquire real-time sensor data from various sensors deployed in the current transformer.
[0088] In one possible implementation, multiple instrument transformers are used for fault detection. In this case, different types of sensors can be used in different instrument transformers to acquire the same type of sensor data. For example, different types of sensors can be installed on instrument transformers in different environments to acquire the same type of sensor data. For instance, when collecting temperature data, a fiber optic grating temperature sensor can be installed for outdoor instrument transformers, while an infrared temperature sensor can be installed for indoor or factory-area instrument transformers.
[0089] At this point, when acquiring multiple sensor data collected in real time by various sensors deployed in the current transformer, the initial data collected in real time by the various sensors deployed in the current transformer can be acquired first; then, the corresponding data processing method can be determined according to the type of sensor that acquires the same sensor data; finally, the initial data can be processed according to the corresponding data processing method to obtain the sensor data.
[0090] The data processing method refers to the preprocessing of the initial data to standardize the digital signal. For example, the data processing method may include analog-to-digital conversion and normalization.
[0091] For example, for a fiber Bragg grating temperature sensor, the corresponding data processing method can be determined as follows: first, perform photoelectric signal conversion to obtain an electrical signal; then, perform analog-to-digital conversion on the electrical signal to obtain a digital signal; finally, perform wavelength modulation processing on the digital signal and then normalize it to obtain temperature data. At this time, the initial data collected by the fiber Bragg grating temperature sensor can be processed according to this data processing method to obtain the temperature data.
[0092] For example, for an infrared temperature sensor, the corresponding data processing method can be determined as follows: first, perform thermoelectric signal conversion to obtain an electrical signal; then, perform analog-to-digital conversion on the electrical signal to obtain a data signal; finally, perform energy conversion on the digital signal and then normalize it to obtain temperature data. In this case, the initial data collected by the infrared temperature sensor can be processed according to this data processing method to obtain the temperature data.
[0093] This implementation method allows for the installation of different types of sensors on different current transformers to acquire the same type of sensor data, thereby enabling fault detection of current transformers in various environments and improving the applicability and scope of the fault detection method for current transformers.
[0094] Optionally, the deep learning model used to extract the first feature information can be composed of two parts: a first model and a second model. Based on this, the process of determining the fault score of the mutual inductor based on multiple sensor data can be realized by following the steps described in S202 to S204 below.
[0095] S202. Based on the first model, local feature extraction is performed on data from multiple sensors to obtain local feature information.
[0096] Optionally, the first model can be a convolutional neural network (CNN) model. In this case, multiple sensor data can be simultaneously input into the trained convolutional neural network model for local feature extraction processing to obtain local feature information corresponding to multiple sensor data.
[0097] S203. Based on the second model, perform temporal feature extraction on the local feature information to obtain the first feature information.
[0098] Among them, the fusion feature of multiple sensor data indicated by the first feature information is a multi-dimensional vector.
[0099] Optionally, the second model can be a Long Short-Term Memory (LSTM) network model. In this case, local feature information can be input into the trained LSTM network model for temporal feature extraction, and the first feature information can be obtained after the temporal features are extracted.
[0100] Optionally, the number of LSTM layers included in the second model can be a single layer or multiple layers. There is no limit to the number of LSTM layers included in the second model, which shall be determined according to actual needs.
[0101] In the above embodiments, local feature information from multiple sensor data can be extracted based on the first model, thereby obtaining higher-dimensional and richer features from the multiple sensor data. Then, time series analysis is performed on the local feature information based on the second model to capture the time dependencies between the local feature information, thus making the obtained first feature information richer and more accurate, thereby helping to improve the accuracy of fault detection results.
[0102] S204. After weighted summation of each dimension in the multi-dimensional vector, the fault score is obtained.
[0103] Optionally, for each dimension in the multi-dimensional vector, corresponding weight information can be preset, and the weighted sum of each dimension in the multi-dimensional vector can be performed according to the preset weight information to obtain the fault score.
[0104] Optionally, when determining the weight information corresponding to each dimension in the multi-dimensional vector, you can first set the same initial weight for each dimension (the sum of each initial weight is 1), and then optimize the initial fault score obtained by weighted summation based on the initial weights to obtain the weight information corresponding to each dimension in the multi-dimensional vector.
[0105] This implementation method can assign higher weights to the more important dimensions in the multi-dimensional vector, thereby making the obtained fault scores more accurate and further improving the accuracy of fault detection.
[0106] In one example, after determining the fault score of the current transformer according to the process described above, it can be determined whether the fault score indicates a fault in the current transformer based on the fault threshold of the current transformer (i.e., the dynamic fault threshold below). The method for determining the dynamic fault threshold can be found in the process described in S205 below.
[0107] S205. Obtain historical sensor data of the current transformer, and determine the dynamic fault threshold corresponding to the current transformer based on various sensor data, historical sensor data and equipment-related information of the current transformer.
[0108] Optionally, the equipment-related information includes, but is not limited to: the operating environment information of the instrument transformer, the service life information of the instrument transformer, the operating time information of the instrument transformer, the model information of the instrument transformer, and the manufacturer information of the instrument transformer.
[0109] In one example, the operating environment information of the current transformer may include, but is not limited to, ambient temperature information and ambient humidity information.
[0110] At this point, the impact of the information contained in the equipment on the operating status of the instrument transformer can be considered, thereby making the fault detection of the instrument transformer more accurate.
[0111] Based on this, when determining the dynamic fault threshold corresponding to the current transformer based on multiple sensor data, historical sensor data, and the device-related information of the current transformer, a third model can be trained first based on historical sensor data and the device-related information corresponding to the historical sensor data to obtain a trained third model; then, the dynamic fault threshold can be determined based on the trained third model, multiple sensor data, and the device-related information corresponding to the multiple sensor data.
[0112] In one example, the third model can be understood as a model that can handle time series data. For example, the third model can be an LSTM model or an ARIMA (AutoRegressive Integrated Moving Average) model. The type of the third model is not limited here, as long as it can be implemented.
[0113] In one example, a third model can be trained based on historical sensor data, the device-related information corresponding to the historical sensor data, and the corresponding fault labels, resulting in a trained third model.
[0114] Subsequently, when determining the dynamic fault threshold based on the trained third model, an initial threshold can be determined at the start of detection (e.g., within the first hour of detection or within half an hour of detection), based on historical sensor data and the device-related information corresponding to the historical sensor data. Then, based on the initial threshold and the fault score, it can be determined whether the transformer is faulty.
[0115] Then, the dynamic fault threshold can be determined based on the trained third model, real-time collected data from multiple sensors, and the device-related information corresponding to the real-time collected data from multiple sensors.
[0116] In one example, a second sliding window of a preset size can be used to input real-time collected data from multiple sensors, along with corresponding device-related information, into a trained third model for processing, thereby obtaining the dynamic fault thresholds corresponding to the real-time collected sensor data. The preset size of the second sliding window can be 1 hour, 1 day, etc.; the size is not limited here, but is determined based on actual needs.
[0117] This implementation method can determine the dynamic fault threshold based on real-time collected data from various sensors, historical sensor data, and equipment-related information. It can combine multiple factors such as equipment production information, power grid load changes, and the environment in which the equipment is located to determine the dynamic fault threshold, thus taking into account the influence of multiple factors on the operating status of the instrument transformer, making the determined dynamic fault threshold more accurate and improving the accuracy of fault detection.
[0118] S206. If a fault is determined to exist in the current transformer based on the fault score and dynamic fault threshold, the fault type of the current transformer is determined based on the second characteristic information corresponding to the data from multiple sensors.
[0119] The second feature information indicates the data characteristics corresponding to data from multiple sensors.
[0120] Optionally, when determining the fault type of the current transformer based on the second feature information corresponding to multiple sensor data, the sensor data can first be processed by feature extraction based on the data type of the sensor data to obtain the second feature information corresponding to each type of sensor data; then, the second feature information can be classified according to the fourth model to obtain the fault type of the current transformer.
[0121] In one example, the second feature information may include feature information under multiple feature dimensions. For example, the feature dimension may include at least one of the time domain feature dimension, frequency domain feature dimension, time-frequency domain feature dimension, and temperature feature dimension.
[0122] At this point, different feature extraction processes can be performed on sensor data of different data types under different feature dimensions. For example, for current and voltage data, in the time domain feature dimension, features such as mean, peak value, and root mean square can be extracted to obtain second feature information; in the frequency domain feature dimension, harmonic component features can be extracted; and in the time-frequency domain feature dimension, wavelet transform features and spectral features can be extracted. For temperature data, features such as temperature mean, temperature change rate, and temperature gradient can be extracted in the temperature feature dimension to obtain second feature information. For partial discharge signals and vibration signals, in the time domain feature dimension, features such as standard deviation or variance, peak value, and waveform factor can be extracted; in the frequency domain feature dimension, band energy features can be extracted; and in the time-frequency domain feature dimension, envelope features and spectral features can be extracted.
[0123] In one example, after determining the second feature information, the second feature information can be input into the fourth model for classification processing. At this time, the fourth model is a model that can perform classification, such as the SVM (Support Vector Machine) model. Here, the fourth model is not limited, as long as it can be implemented.
[0124] Optionally, the fault types of the current transformer may include, but are not limited to: winding short circuit fault, core saturation fault, insulation aging fault, partial discharge fault, overload fault, and mechanical fault.
[0125] In the above embodiments, when a fault is determined in the current transformer, more refined second feature information from multiple sensor data can be extracted, and fault classification processing can be performed based on the second feature information, making the fault classification more accurate and thus obtaining more accurate fault detection results.
[0126] In one possible implementation, after determining the fault type of the instrument transformer, a fault detection result can be determined based on the determined fault type. This fault detection result may include, but is not limited to, the fault type, fault cause, fault location, fault consequences, and maintenance recommendations. The fault detection result can then be sent to the power system, which will trigger a fault warning sound and / or illuminate a fault indicator light via a buzzer, thus prompting staff to promptly address the fault and ensure the stable operation of the power system.
[0127] In one possible implementation, in addition to issuing fault warnings within the power system, warning messages can be pushed to staff based on the fault detection results, thereby ensuring that staff receive timely reminders and improving maintenance efficiency.
[0128] In one possible implementation, when there are multiple current transformers, the fault detection result may also include the identification information of the current transformers to facilitate accurate location of the faulty current transformer.
[0129] Figure 3 This is a schematic diagram of the structure of a fault detection system for a current transformer provided in an embodiment of this application, as shown below. Figure 3 As shown, the fault detection system of the current transformer includes: a data acquisition and preprocessing module, a data fusion processing module, a dynamic threshold generation module, a fault diagnosis module, and a fault early warning module.
[0130] The data acquisition and preprocessing module is used to acquire data from various sensors installed on the current transformer, and to perform analog-to-digital conversion, filtering, noise reduction, and normalization on the data acquired by each sensor to obtain various sensor data.
[0131] The data fusion processing module is used to perform feature extraction processing on multiple sensor data according to the first model and the second model mentioned above, to obtain first feature information, and to determine the fault score of the current transformer based on the first feature information.
[0132] The dynamic threshold generation module is used to determine the dynamic fault threshold corresponding to the current transformer based on historical sensor data, real-time sensor data, equipment-related information, and the aforementioned third model, thereby facilitating the determination of whether the current transformer is faulty.
[0133] The fault diagnosis module is used to trigger the fault diagnosis process when a fault is found in the current transformer. That is, by extracting the second feature information corresponding to multiple sensor data, and based on the fourth model and the second feature information mentioned above, the fault diagnosis is performed to determine the fault type of the current transformer.
[0134] The fault early warning module is used to determine the fault detection result based on the identified fault type of the instrument transformer, and to issue an alert in the power system based on the fault detection result, and / or to send the fault detection result to the staff for alert, thereby realizing fault early warning and timely handling of faults.
[0135] Figure 4 This is a schematic diagram of the structure of a fault detection device for a current transformer provided in an embodiment of this application, as shown below. Figure 4 As shown, the device 40 includes:
[0136] The acquisition unit 401 is used to acquire multiple sensor data collected in real time by multiple sensors deployed in the current transformer.
[0137] The first determining unit 402 is used to determine the fault score of the current transformer based on the first feature information corresponding to multiple sensor data; wherein the first feature information indicates the fusion feature of multiple sensor data.
[0138] The second determining unit 403 is used to acquire historical sensor data of the current transformer and determine the dynamic fault threshold corresponding to the current transformer based on various sensor data, historical sensor data and equipment-related information of the current transformer.
[0139] The fault processing unit 404 is used to determine the fault type of the current transformer based on the second feature information corresponding to multiple sensor data if the current transformer is determined to have a fault according to the fault score and dynamic fault threshold; wherein the second feature information indicates the data features corresponding to multiple sensor data.
[0140] Optionally, the first determining unit 402 is used for:
[0141] Based on the first model, local feature extraction is performed on data from multiple sensors to obtain local feature information;
[0142] Based on the second model, temporal features are extracted from local feature information to obtain the first feature information; wherein, the first feature information is a multi-dimensional vector;
[0143] The fault score is obtained by weighted summation of each dimension in the multi-dimensional vector.
[0144] Optionally, the equipment-related information includes, but is not limited to: the operating environment information of the instrument transformer, the service life information of the instrument transformer, the operating time information of the instrument transformer, the model information of the instrument transformer, and the manufacturer information of the instrument transformer.
[0145] Optionally, the second determining unit 403 is used for:
[0146] Based on historical sensor data and the corresponding device information, a third model is trained to obtain a well-trained third model.
[0147] Based on the trained third model, data from multiple sensors, and device-related information corresponding to the data from multiple sensors, the dynamic fault threshold is determined.
[0148] Optionally, the sensor data may include at least: current data, voltage data, temperature data, partial discharge signal data, and vibration signal data.
[0149] Optionally, there can be multiple current transformers, with different types of sensors used in different current transformers to acquire the same type of sensor data; in this case, the acquisition unit 401 is used for:
[0150] Acquire initial data in real time from various sensors deployed in the current transformer;
[0151] Determine the corresponding data processing method based on the type of sensor that acquires the same sensor data;
[0152] The initial data is processed according to the corresponding data processing method to obtain sensor data.
[0153] Optionally, the fault handling unit 404 is used for:
[0154] Based on the data type of the sensor data, feature extraction processing is performed on the sensor data to obtain the second feature information corresponding to each type of sensor data;
[0155] The second feature information is classified and processed according to the fourth model to obtain the fault type of the transformer.
[0156] The fault detection device for the current transformer provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.
[0157] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 5 As shown, the electronic device 50 provided in this embodiment includes at least one processor 501 and a memory 502. Optionally, the device 50 further includes a communication component 503. The processor 501, memory 502, and communication component 503 are connected via a bus 504.
[0158] In a specific implementation, at least one processor 501 executes computer execution instructions stored in memory 502, causing at least one processor 501 to perform the above-described method.
[0159] The specific implementation process of processor 501 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.
[0160] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.
[0161] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.
[0162] 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. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.
[0163] Figure 6 This is a schematic diagram of the structure of a current transformer provided in an embodiment of this application, as shown below. Figure 6 As shown, the current transformer 60 provided in this embodiment includes: a current transformer body, multiple sensors installed on the current transformer body, a power supply module, and... Figure 5 The electronic device shown. The power supply module is used to supply power to at least one of the multiple sensors, electronic devices, and current transformer bodies.
[0164] Optionally, the power supply module can be a power supply module that combines photovoltaic panels and lithium batteries.
[0165] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.
[0166] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.
[0167] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.
[0168] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.
[0169] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.
[0170] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0171] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.
[0172] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0173] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.
[0174] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.
Claims
1. A fault detection method for a current transformer, characterized in that, Applied to a processor; the method includes: Acquire real-time sensor data from multiple sensors deployed in the current transformer; The fault score of the current transformer is determined based on the first feature information corresponding to the multiple sensor data; wherein, the first feature information indicates the fusion feature of the multiple sensor data; The historical sensor data of the current transformer is obtained, and the dynamic fault threshold corresponding to the current transformer is determined based on the various sensor data, the historical sensor data and the device-related information of the current transformer. If the current transformer is determined to be faulty based on the fault score and the dynamic fault threshold, the fault type of the current transformer is determined based on the second feature information corresponding to the multiple sensor data; wherein, the second feature information indicates the data characteristics corresponding to the multiple sensor data.
2. The method according to claim 1, characterized in that, Based on the first feature information corresponding to the various sensor data, a fault score for the current transformer is determined, including: Based on the first model, local feature extraction is performed on the data from the various sensors to obtain local feature information; According to the second model, temporal feature extraction is performed on the local feature information to obtain the first feature information; wherein, the first feature information is a multi-dimensional vector; The fault score is obtained by weighted summation of each dimension in the multi-dimensional vector.
3. The method according to claim 1, characterized in that, The equipment-related information includes, but is not limited to: the operating environment information of the instrument transformer, the service life information of the instrument transformer, the operating time information of the instrument transformer, the model information of the instrument transformer, and the manufacturer information of the instrument transformer.
4. The method according to claim 3, characterized in that, Based on the data from the various sensors, the historical sensor data, and the device-related information of the current transformer, the dynamic fault threshold corresponding to the current transformer is determined, including: Based on the historical sensor data and the device-related information corresponding to the historical sensor data, a third model is trained to obtain a trained third model; The dynamic fault threshold is determined based on the trained third model, the various sensor data, and the device-related information corresponding to the various sensor data.
5. The method according to any one of claims 1 to 4, characterized in that, The sensor data includes at least: current data, voltage data, temperature data, partial discharge signal data, and vibration signal data.
6. The method according to claim 5, characterized in that, The number of current transformers is multiple, and different types of sensors are used in different current transformers to acquire the same type of sensor data; multiple types of sensor data are acquired in real time from various sensors deployed in the current transformers, including: Acquire initial data collected in real time by the various sensors deployed in the current transformer; Determine the corresponding data processing method based on the type of sensor that acquires the same sensor data; The initial data is processed according to the corresponding data processing method to obtain the sensor data.
7. The method according to any one of claims 1 to 4, characterized in that, Based on the second feature information corresponding to the various sensor data, the fault type of the current transformer is determined, including: Based on the data type of the sensor data, feature extraction processing is performed on the sensor data to obtain the second feature information corresponding to each type of sensor data; The second feature information is classified and processed according to the fourth model to obtain the fault type of the current transformer.
8. A fault detection device for a current transformer, characterized in that, Applied to a processor; the device includes: The acquisition unit is used to acquire various sensor data collected in real time by various sensors deployed in the current transformer; The first determining unit is configured to determine the fault score of the current transformer based on the first feature information corresponding to the multiple sensor data; wherein the first feature information indicates the fusion feature of the multiple sensor data; The second determining unit is used to acquire historical sensor data of the current transformer, and determine the dynamic fault threshold corresponding to the current transformer based on the various sensor data, the historical sensor data and the equipment-related information of the current transformer. The fault processing unit is configured to determine the fault type of the current transformer based on the second feature information corresponding to the multiple sensor data if the current transformer is determined to be faulty based on the fault score and the dynamic fault threshold; wherein the second feature information indicates the data characteristics corresponding to the multiple sensor data.
9. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1 to 7.
10. A current transformer, characterized in that, The current transformer includes: a current transformer body, a plurality of sensors mounted on the current transformer body, a power supply module, and the electronic device as described in claim 9; wherein the power supply module is used to supply power to at least one of the plurality of sensors, the electronic device, and the current transformer body.