Transformer insulation fault monitoring and early warning device and early warning system thereof

Abstract

The invention discloses a transformer insulation fault monitoring and early warning device and an early warning system thereof, and relates to the technical field of transformer insulation faults. The transformer comprises a transformer body, and a partial discharge on-line monitor, a temperature sensing assembly, an insulating oil photoacoustic spectrum on-line monitor, an oil micro-water analyzer, a vibration monitor and an insulation tester are arranged on the peripheral side of the transformer body. Through arrangement of the partial discharge on-line monitor, the temperature sensing assembly, the insulation oil photoacoustic spectrum on-line monitor, the oil micro-water analyzer, the vibration monitor, the insulation tester and the main control end, the insulation fault of the transformer body can be comprehensively monitored in real time, and the alarm lamp is used for reminding irrelevant personnel to be away from the transformer body. And fault information is timely transmitted to a computer or a communication terminal of related personnel through the early warning module, so that the fault information is timely processed, the early warning and reminding effects are greatly achieved, the occurrence of accident hidden dangers is reduced, and the guarantee is improved for power utilization safety.

Description

Technical Field

[0001] The invention belongs to the field of transformer insulation failure, and in particular relates to a transformer insulation failure monitoring and early warning device and an early warning system thereof. Background Art

[0002] A transformer is a device that uses the principle of electromagnetic induction to change AC voltage. Its main components are primary coils, secondary coils, and iron cores. Its main functions include voltage conversion, current conversion, impedance conversion, isolation, and voltage stabilization. Transformers are basic equipment for power transmission and distribution, and are widely used in industries, agriculture, transportation, urban communities, and other fields. There are about 17 million transformers in operation in my country, with a total capacity of about 11 billion kilovolt-amperes. Transformer losses account for about 40% of power transmission and distribution losses, and have great energy-saving potential. Transformer insulation damage failures account for about 70% to 80% of total failures. Therefore, monitoring and early warning of transformer insulation failures can effectively increase the service life of transformers and ensure the safe operation of electricity.

[0003] A Chinese patent with publication number CN208459496U discloses an online monitoring system for traction transformers used in electrified railways, including an insulating oil photoacoustic spectroscopy online monitoring device, an iron core grounding current online monitoring device, a winding temperature online monitoring device, a partial discharge online monitoring device, a bushing insulation online monitoring device, a vibration online monitoring device, an intelligent merging unit and a fault warning and status analysis system software platform. The insulating oil photoacoustic spectroscopy online monitoring device, the iron core grounding current online monitoring device, the winding temperature online monitoring device, the partial discharge online monitoring device, the bushing insulation online monitoring device, and the vibration online monitoring device complete the collection of traction transformer operating status information, and the intelligent merging unit summarizes all operating status information collected by the online monitoring device.

[0004] Although the above application monitors the insulation fault of the transformer in real time, it cannot ensure whether the fault information after monitoring is viewed, received and processed. There is a possibility that relevant personnel will omit and ignore some fault information, which has a certain probability of potential accidents.

[0005] In view of this, the present invention is proposed. Summary of the invention

[0006] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide a transformer insulation fault monitoring and early warning device and an early warning system thereof, which solve the problems raised in the above-mentioned background technology.

[0007] In order to solve the above technical problems, the basic concept of the technical solution adopted by the present invention is:

[0008] A transformer insulation fault monitoring and early warning device comprises: a transformer body, wherein a partial discharge online monitor, a temperature sensing component, an insulating oil photoacoustic spectroscopy online monitor, an oil micro-water analyzer, a vibration monitor and an insulation tester are respectively installed on the peripheral sides of the transformer body, one end of the partial discharge online monitor, the temperature sensing component, the insulating oil photoacoustic spectroscopy online monitor, the oil micro-water analyzer, the vibration monitor and the insulation tester is connected to a main control end, a protective cover is clamped on the top of the transformer body, two alarm lights are installed in the protective cover, and the two alarm lights are electrically connected to the main control end.

[0009] A transformer insulation fault monitoring and early warning system comprises a main control terminal, an insulating oil photoacoustic spectroscopy online monitoring module, and an oil micro-water analysis module, wherein the main control terminal is connected to an analysis module, a storage module, and a processing module, wherein:

[0010] The storage module is used to record all data information obtained during monitoring;

[0011] A partial discharge online monitoring module is used to monitor partial discharge phenomena of the insulation system in the transformer body in real time. The partial discharge online monitoring module continuously collects high-frequency electromagnetic signals or ultrasonic signals generated during the operation of the transformer through high-frequency sensors, and transmits these signals to the main control system;

[0012] The temperature sensor module is used to monitor the temperature of the transformer insulation coil and transmit the monitored temperature data to the main control system;

[0013] Insulating oil photoacoustic spectroscopy online monitoring module and oil micro-water analysis module are used to monitor the dissolved gas components in the transformer body insulating oil;

[0014] A vibration monitoring module is used to monitor the vibration stress of the transformer body in real time during operation. The vibration monitoring module collects vibration data in real time through sensors and converts it into digital signals for transmission to the main control system;

[0015] The insulation test module is used to monitor the insulation performance of the transformer body in real time. The insulation test module can capture the slight changes in the electrical insulation performance of the transformer during operation, and perform real-time analysis on various insulation indicators, and transmit the results to the main control system for comprehensive evaluation.

[0016] Optionally, the analysis module includes an evaluation module, which is used to further analyze the monitoring data and derive a specific cause of the fault.

[0017] Optionally, the evaluation module is connected to a data visualization module, which includes visualization charts, comparison reference tables, data increase and decrease line charts and bar charts, so that users can intuitively understand the fault analysis results.

[0018] Optionally, the processing module is connected to the early warning module, the early warning module is connected to the module to be processed, the early warning module includes an early warning signal light and a signal sending module, and the signal sending module is used to send the early warning information to a computer or a remote user terminal so that the user can obtain the fault warning information in time.

[0019] Optionally, the vibration monitoring module is used to monitor the vibration stress of the transformer body, and the insulation testing module is used to monitor the insulation performance of the transformer body in the following steps:

[0020] Collect the vibration data generated by the transformer during operation in real time and transmit the collected vibration data to the main control end;

[0021] The main control end constructs a normal vibration mode model based on historical data, and then builds a vibration feature baseline of the transformer under normal operating conditions. Then, a support vector machine vibration analysis model is constructed to identify abnormal vibration modes.

[0022] A threshold is set in the main control terminal according to historical data, and then the difference between the current vibration pattern and the baseline is determined according to the preset threshold. If the degree of deviation from the normal pattern exceeds the preset threshold, it is identified as an abnormal vibration pattern;

[0023] When abnormal vibration is detected, the vibration signal will be compared with historical data and classified.

[0024] Optionally, the steps of constructing a support vector machine vibration analysis model are:

[0025] Extract time domain features from historical data and use Fourier transform to convert data from time domain to frequency domain, extract spectrum features, and perform time-frequency domain analysis on non-stationary signals to extract time-frequency features, including mean, amplitude, energy spectrum, power spectrum density, instantaneous frequency, and instantaneous phase;

[0026] Select linear SVM or nonlinear SVM according to the data distribution, and then set the penalty parameter C and kernel function parameter γ of the linear SVM or nonlinear SVM, which are expressed as follows: Among them, w is the weight vector, b is the bias term, and y i is the label, C controls the penalty for classification error, and the kernel function parameter γ controls the shape of the kernel function;

[0027] The data is divided into a training set and a test set. Then, the training set data is used to train the SVM model. After the training is completed, the test set data is used to verify the model and evaluate the performance of the model. Finally, the trained SVM model is deployed to the main control end for real-time monitoring and anomaly detection.

[0028] Optionally, the pending module has a timed monitoring and automatic detection function for tracking the processing status of the warning information. After the warning information is generated, the pending module starts the timer and continuously monitors the progress of the warning processing within a specified time range. If the warning information is not processed or confirmed in time within the specified response time, the pending module will automatically identify the status as timeout and unprocessed;

[0029] At this point, the processing module collects multi-dimensional data related to the warning processing, and then records the time and frequency of each response to build a comprehensive processing status database;

[0030] After data collection is completed, the processing module will extract key features from the raw data, including average response time, maximum and minimum response time, number of response delays, and the most recent processing time. The key features are then integrated into training sets and validation sets, and the training set is used to train the isolation forest detection model. Then, by repeatedly iteratively adjusting the model parameters, the processing module will enable the model to learn the difference between normal response patterns and potential anomalies.

[0031] Each time the processing module generates new warning information, it will obtain the current processing data in real time and input this data into the trained anomaly detection model;

[0032] During real-time monitoring, the processing module will use the trained model to determine whether the current response mode is abnormal. If the processing time or frequency detected by the model deviates significantly from the normal value, the system will automatically mark the state as a potential delay. This judgment will be based on the output of the model and cross-validated with historical data. In addition, the model will generate a classifier during long-term training to analyze new data inputs in real time to ensure that the model can accurately identify potential processing delays and respond to real-time data. The processing module will also set a threshold. If the current response exceeds the set range, an abnormal alarm will be triggered;

[0033] After detecting an anomaly, the processing module will combine the statistical information of historical data and analyze the patterns of previous processing delays to predict response time delays. Once a delay is predicted, the processing module will immediately mark the status as high risk through internal logic, issue an alert to relevant personnel, and generate a detailed exception handling report. The exception handling report covers the time, type, and predicted processing delay of the anomaly. These reports will be provided to management for subsequent analysis to help identify system weaknesses and potential improvements.

[0034] Optionally, the signal sending module is used to send the monitored warning information to the remote user terminal in real time through wired or wireless communication, and the communication protocol exchanges data with the control center or mobile device to ensure the timely delivery of the warning information. The signal sending module also supports multi-terminal concurrent communication and can send information to multiple user terminals synchronously to ensure that in an emergency, maintenance personnel or managers at different locations can receive relevant warnings at the same time. The transmission methods of warning information include but are not limited to SMS, e-mail, and dedicated client notifications.

[0035] After adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art. Of course, any product implementing the present invention does not necessarily need to achieve all the advantages described below at the same time:

[0036] Through the installation of partial discharge online monitor, temperature sensor assembly, insulating oil photoacoustic spectroscopy online monitor, oil micro-water analyzer, vibration monitor and insulation tester, the transformer body insulation fault can be fully monitored in real time, and the alarm light can be used to remind irrelevant personnel to stay away. The fault information can be transmitted to the computer or communication terminal of the relevant personnel in time through the early warning module, so as to be processed in time. At the same time, the use of the pending module and the security alarm system can further ensure the complete processing of the insulation fault information, effectively avoid the situation where some fault information is omitted and ignored, greatly play a role in early warning, reduce the occurrence of potential accidents, and improve the protection of electricity safety.

[0037] The specific implementation modes of the present invention are further described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The drawings described below are only some embodiments. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0039] In the figure:

[0040] Figure 1 It is a schematic diagram of the transformer body structure;

[0041] Figure 2 Schematic diagram of insulation fault monitoring system.

[0042] In the accompanying drawings, the components represented by the reference numerals are listed as follows:

[0043] Transformer body 1, partial discharge online monitor 2, temperature sensor component 3, insulating oil photoacoustic spectroscopy online monitor 4, oil micro-water analyzer 5, vibration monitor 6, insulation tester 7, protective cover 8, alarm light 9.

[0044] It should be noted that these drawings and textual descriptions are not intended to limit the conceptual scope of the present invention in any way, but are intended to illustrate the concept of the present invention for those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0045] The present invention will now be described in further detail with reference to the accompanying drawings.

[0046] See also Figure 1-2 As shown, in this embodiment, a transformer insulation fault monitoring and early warning device is provided, including a transformer body 1, and a partial discharge online monitor 2, a temperature sensor component 3, an insulating oil photoacoustic spectrometer online monitor 4, an oil micro-water analyzer 5, a vibration monitor 6, and an insulation tester 7 are respectively installed on the surrounding sides of the transformer body 1, and one end of the partial discharge online monitor 2, the temperature sensor component 3, the insulating oil photoacoustic spectrometer online monitor 4, the oil micro-water analyzer 5, the vibration monitor 6 and the insulation tester 7 is connected to a main control end, and a protective cover 8 is clamped on the top of the transformer body 1, and two alarm lights 9 are installed in the protective cover 8, and the two alarm lights 9 are electrically connected to the main control end, wherein the protective cover 8 can be connected to the top of the transformer body 1 by a buckle in the prior art.

[0047] The application of one aspect of this embodiment is: when the transformer body 1 is in operation, the partial discharge online monitor 2, the temperature sensor assembly 3, the insulating oil photoacoustic spectroscopy online monitor 4, the oil micro-water analyzer 5, the vibration monitor 6 and the insulation tester 7 are turned on to monitor the insulation fault, and the data information is transmitted to the main control end for data processing. When the insulation fault is detected, the signal is transmitted to the two alarm lights 9 for early warning, and the protective cover 8 can protect the two alarm lights 9, effectively reducing the occurrence of collision damage to the alarm lights 9, and the protective cover 8 can be opened to replace the damaged alarm lights 9 in time. It should be noted that all electrical equipment involved in this application can be powered by batteries or external power supplies.

[0048] A transformer insulation fault monitoring and early warning system comprises: a main control end, an insulating oil photoacoustic spectroscopy online monitoring module, and an oil micro-water analysis module. The main control end is connected to an analysis module, a storage module, and a processing module, wherein:

[0049] The storage module is used to record all data information obtained during monitoring;

[0050] The partial discharge online monitoring module is used to monitor the partial discharge phenomenon of the insulation system in the transformer body 1 in real time. The partial discharge online monitoring module continuously collects high-frequency electromagnetic signals or ultrasonic signals generated during the operation of the transformer through high-frequency sensors, and transmits these signals to the main control system;

[0051] The temperature sensor module is used to monitor the temperature of the insulation coil of the transformer body 1 and transmit the monitored temperature data to the main control system;

[0052] Insulating oil photoacoustic spectroscopy online monitoring module and oil micro-water analysis module are used to monitor the dissolved gas components in the insulating oil of the transformer body 1;

[0053] A vibration monitoring module is used to monitor the vibration stress of the transformer body 1 in real time during operation. The vibration monitoring module collects vibration data in real time through sensors and converts them into digital signals for transmission to the main control system;

[0054] The insulation test module is used to monitor the insulation performance of the transformer body 1 in real time. The insulation test module can capture the slight changes in the electrical insulation performance during the operation of the transformer, and perform real-time analysis on various insulation indicators, and transmit the results to the main control system for comprehensive evaluation.

[0055] It should be noted that by integrating multiple online monitoring modules, such as partial discharge online monitoring module, temperature sensor module, insulating oil photoacoustic spectroscopy online monitoring module, oil micro-water analysis module, vibration monitoring module and insulation test module, all-round real-time monitoring of multiple key parameters of the transformer body is achieved. These modules are connected to the main control end, and through high-precision sensors and signal processing technology, they can ensure timely detection of small changes in partial discharge, dissolved gas in insulating oil, temperature, vibration stress and insulation performance. Through real-time data collection, processing and storage, the system can provide accurate and reliable monitoring data for the analysis module, which further evaluates the health status of the transformer, helps predict potential faults and generates early warning signals in a timely manner. This multi-module collaborative monitoring solution not only improves the safety of equipment operation, but also extends the service life of the transformer, reduces maintenance costs and the risk of sudden failures.

[0056] In this embodiment, the analysis module includes an evaluation module, and the evaluation module is used to further analyze the monitoring data and obtain the specific cause of the fault.

[0057] It should be noted that the evaluation module can not only identify the abnormal state of the transformer through in-depth analysis of the monitoring data, but also further locate the specific cause of the fault. This not only improves the accuracy of fault diagnosis and avoids simple warnings or false alarms, but also provides maintenance personnel with clear maintenance guidance, helping to quickly identify the problem and reduce diagnosis time. At the same time, the evaluation module can use multi-dimensional data for comprehensive analysis, including cross-comparison of multiple parameters such as temperature, partial discharge, vibration, etc., which greatly improves the intelligence level of the monitoring system, helps to timely discover potential problems, prevent the expansion of faults, and thus improve the safety and reliability of equipment operation.

[0058] In this embodiment, the evaluation module is connected to a data visualization module, and the data visualization module includes visualization charts, comparison reference tables, data increase and decrease line charts, and bar charts, so that users can intuitively understand the fault analysis results.

[0059] It should be noted that the data visualization module presents complex fault analysis results to users in an intuitive and easy-to-understand way through various forms of charts and data presentation. This graphical interface can help users quickly grasp key information and understand the operating status and potential faults of the transformer without having a deep technical background. In addition, the data increase and decrease line chart and the comparison reference table can show the changing trend of the equipment operating parameters, allowing users to clearly see the formation process of abnormal conditions, thereby providing strong support for decision-making. The bar chart can compare the data of different monitoring modules, which is convenient for comprehensive analysis of the abnormality of various indicators. This not only improves the user experience of the system, but also speeds up the diagnosis of problems, supports timely decision-making and maintenance actions, and reduces the risks brought by equipment failures.

[0060] In this embodiment, the processing module is connected to the early warning module, the early warning module is connected to the module to be processed, the early warning module includes an early warning signal light and a signal sending module, and the signal sending module is used to send the early warning information to a computer or a remote user terminal so that the user can obtain the fault warning information in time.

[0061] The early warning module ensures that fault warnings can be quickly and accurately communicated to users through multi-level early warning mechanisms and signal transmission methods. First, the early warning signal light provides an instant visual reminder on site, helping on-site operators to immediately realize equipment abnormalities. Secondly, the signal sending module can transmit early warning information to a computer or remote user terminal in real time through the network. No matter where the user is, he can receive fault alarms in time and respond quickly. This multi-channel early warning method greatly improves the speed and efficiency of information transmission, avoids the lag or omission of early warning information, ensures that equipment can be handled in a timely manner when a fault occurs, reduces the risk of equipment damage or downtime, and improves the responsiveness and security of the entire monitoring system.

[0062] In this embodiment, the vibration monitoring module is used to monitor the vibration stress of the transformer body 1, and the insulation testing module is used to monitor the insulation performance of the transformer body 1 in the following steps:

[0063] Collect the vibration data generated by the transformer during operation in real time and transmit the collected vibration data to the main control end;

[0064] The main control end constructs a normal vibration mode model based on historical data, and then builds a vibration feature baseline of the transformer under normal operating conditions. Then, a support vector machine vibration analysis model is constructed to identify abnormal vibration modes.

[0065] A threshold is set in the main control terminal according to historical data, and then the difference between the current vibration pattern and the baseline is determined according to the preset threshold. If the degree of deviation from the normal pattern exceeds the preset threshold, it is identified as an abnormal vibration pattern;

[0066] When abnormal vibration is detected, the vibration signal will be compared with historical data and classified.

[0067] Through multiple steps, a systematic fault detection process is formed to ensure the safe operation of the transformer. First, the ability to collect vibration data in real time can continuously track the operating status of the transformer, and process it through the main control end, using the built-in vibration analysis model for in-depth analysis to ensure data accuracy. Secondly, by establishing a normal mode based on the vibration feature baseline of historical data, combined with the support vector machine (SVM) model, the system can automatically identify abnormal vibration modes, improving the intelligence level of fault detection. By setting the threshold, the difference between the current vibration and the normal operating state can be accurately judged to ensure that the alarm is triggered only when there is a significant deviation, reducing the false alarm rate. Finally, through the comparison and classification of historical data, the system can more comprehensively judge the severity and nature of the problem and support decision makers to take appropriate maintenance measures. This progressive processing method ensures that the system not only has high-precision fault identification capabilities, but also can perform intelligent learning and self-optimization based on historical data to improve the reliability and safety of the equipment.

[0068] In this embodiment, the steps of constructing the support vector machine vibration analysis model are:

[0069] Extract time domain features from historical data and use Fourier transform to convert data from time domain to frequency domain, extract spectrum features, and perform time-frequency domain analysis on non-stationary signals to extract time-frequency features, including mean, amplitude, energy spectrum, power spectrum density, instantaneous frequency, and instantaneous phase;

[0070] Select linear SVM or nonlinear SVM according to the data distribution, and then set the penalty parameter C and kernel function parameter γ of the linear SVM or nonlinear SVM, which are expressed as follows: Among them, w is the weight vector, b is the bias term, and y i is the label, C controls the penalty for classification error, and the kernel function parameter γ controls the shape of the kernel function;

[0071] The data is divided into a training set and a test set. Then, the training set data is used to train the SVM model. After the training is completed, the test set data is used to verify the model and evaluate the performance of the model. Finally, the trained SVM model is deployed to the main control terminal for real-time monitoring and anomaly detection. By extracting time domain and frequency domain features, Fourier transform and time-frequency domain analysis (such as wavelet transform) are used to process non-stationary signals to ensure that the system can capture multi-dimensional change information of the data, greatly improving the accuracy and completeness of the data. Secondly, a suitable SVM model (linear or nonlinear) is selected according to the data distribution, and key parameters (such as penalty parameters and kernel function parameters) are adjusted to optimize the performance of the model, indicating that the method has a high degree of flexibility and adaptability. The present invention uses the trained SVM model to be deployed to the main control terminal to achieve real-time monitoring and anomaly detection, which means that the system can automatically determine whether the equipment is operating normally based on the new vibration data and discover potential problems in a timely manner.

[0072] In this embodiment, the processing module has a timed monitoring and automatic detection function, which is used to track the processing status of the warning information. After the warning information is generated, the processing module starts the timer and continuously monitors the progress of the warning processing within a specified time range. If the warning information is not processed or confirmed in time within the specified response time, the processing module will automatically identify the status as timeout and unprocessed;

[0073] At this point, the processing module collects multi-dimensional data related to the warning processing, and then records the time and frequency of each response to build a comprehensive processing status database;

[0074] After data collection is completed, the processing module will extract key features from the raw data, including average response time, maximum and minimum response time, number of response delays, and the most recent processing time. The key features are then integrated into training sets and validation sets, and the training set is used to train the isolation forest detection model. Then, by repeatedly iteratively adjusting the model parameters, the processing module will enable the model to learn the difference between normal response patterns and potential anomalies.

[0075] Each time the processing module generates new warning information, it will obtain the current processing data in real time and input this data into the trained anomaly detection model;

[0076] During real-time monitoring, the processing module will use the trained model to determine whether the current response mode is abnormal. If the processing time or frequency detected by the model deviates significantly from the normal value, the system will automatically mark the state as a potential delay. This judgment will be based on the output of the model and cross-validated with historical data. In addition, the model will generate a classifier during long-term training to analyze new data inputs in real time to ensure that the model can accurately identify potential processing delays and respond to real-time data. The processing module will also set a threshold. If the current response exceeds the set range, an abnormal alarm will be triggered;

[0077] After detecting an anomaly, the processing module will combine the statistical information of historical data and analyze the pattern of previous processing delays to predict the response time delay. Once a delay is predicted, the processing module will immediately mark the status as high risk through internal logic, and issue an alarm to relevant personnel and generate a detailed anomaly handling report. The anomaly handling report covers the time, type and predicted processing delay of the anomaly. These reports will be provided to management for subsequent analysis to help identify system weaknesses and potential improvement points. Through a variety of early warning methods (such as sound, light signals, remote notifications, etc.), ensure that relevant personnel can be informed in time and respond quickly to avoid further deterioration of the fault. This mechanism not only improves the efficiency and reliability of early warning information processing, but also ensures the safe operation of the equipment and reduces the risks brought by sudden failures.

[0078] In this embodiment, the signal sending module is used to send the monitored warning information to the remote user terminal in real time through wired or wireless communication, and the communication protocol interacts with the control center or mobile device to ensure the timely transmission of the warning information. The signal sending module also supports multi-terminal concurrent communication, and can send information to multiple user terminals synchronously to ensure that in an emergency, maintenance personnel or managers at different locations can receive relevant warnings at the same time. The transmission method of the warning information includes but is not limited to text messages, emails, and client notifications. The signal sending module transmits the warning information in real time through wired or wireless communication to ensure that the remote user can receive the fault alarm in time. This diversified communication method adapts to different network environments and improves the flexibility and adaptability of the system. The multi-terminal concurrent communication function of the signal sending module means that the warning information can be sent to multiple user terminals at the same time, and the maintenance personnel or managers can obtain the warning information synchronously regardless of where they are distributed. This function is particularly suitable for emergency situations, ensuring the wide dissemination and timely response of information, and avoiding handling errors caused by delays in a single terminal or incomplete information. The transmission methods of this module are diverse, including text messages, emails, client notifications, etc., making information transmission more flexible and diverse, and able to adapt to different user needs and scenarios. Overall, this modular design significantly improves the transmission efficiency, response speed and system security of early warning information, and ensures safe monitoring and fault management during equipment operation.

[0079] The present invention is not limited to the above-mentioned embodiments. Anyone should be aware that any structural changes made under the enlightenment of the present invention, and any technical solutions that are the same or similar to the present invention, fall within the protection scope of the present invention. The technology, shape, and structural parts not described in detail in the present invention are all well-known technologies.

Claims

1. A transformer insulation fault monitoring and early warning device, characterized in that: include: A transformer body (1) is provided. A partial discharge online monitor (2), a temperature sensor assembly (3), an insulating oil photoacoustic spectroscopy online monitor (4), an oil micro-water analyzer (5), a vibration monitor (6), and an insulation tester (7) are respectively installed on the peripheral side of the transformer body (1). One end of the partial discharge online monitor (2), the temperature sensor assembly (3), the insulating oil photoacoustic spectroscopy online monitor (4), the oil micro-water analyzer (5), the vibration monitor (6), and the insulation tester (7) is connected to a main control end. A protective cover (8) is clamped on the top of the transformer body (1). Two alarm lights (9) are installed in the protective cover (8). The two alarm lights (9) are electrically connected to the main control end.

2. A transformer insulation fault monitoring and early warning system, characterized in that: It includes a main control terminal, an insulating oil photoacoustic spectroscopy online monitoring module, and an oil-micro-water analysis module. The main control terminal is connected to an analysis module, a storage module, and a processing module, wherein: The storage module is used to record all data information obtained during monitoring; A partial discharge online monitoring module is used to monitor partial discharge phenomena of the insulation system in the transformer body (1) in real time, wherein the partial discharge online monitoring module continuously collects high-frequency electromagnetic signals generated during the operation of the transformer through a high-frequency sensor and transmits these signals to a main control system; A temperature sensing module, used to monitor the temperature of the insulation coil of the transformer body (1) and transmit the monitored temperature data to a main control system; Insulating oil photoacoustic spectroscopy online monitoring module and oil micro-water analysis module are used to monitor the dissolved gas components in the insulating oil of the transformer body (1); A vibration monitoring module, used for real-time monitoring of the vibration stress of the transformer body (1) during operation, wherein the vibration monitoring module collects vibration data in real time through a sensor and converts the data into digital signals for transmission to a main control system; The insulation test module is used to monitor the insulation performance of the transformer body (1) in real time. The insulation test module can capture changes in the electrical insulation performance of the transformer during operation and perform real-time analysis on various insulation indicators.

3. A transformer insulation fault monitoring and early warning system according to claim 2, characterized in that: The analysis module includes an evaluation module, and the evaluation module is used to further analyze the monitoring data and obtain the specific cause of the fault.

4. A transformer insulation fault monitoring and early warning system according to claim 2, characterized in that: The evaluation module is connected to a data visualization module, which includes a visualization chart, a comparison reference table, a data increase and decrease line chart and a bar chart, so that the user can intuitively understand the fault analysis result.

5. A transformer insulation fault monitoring and early warning system according to claim 2, characterized in that: The processing module is connected to the early warning module, the early warning module is connected to the module to be processed, the early warning module includes an early warning signal light and a signal sending module, and the signal sending module is used to send early warning information to a computer or a remote user terminal so that the user can obtain fault early warning information in time.

6. A transformer insulation fault monitoring and early warning system according to claim 2, characterized in that: The vibration monitoring module is used to monitor the vibration stress of the transformer body (1), and the insulation testing module is used to monitor the insulation performance of the transformer body (1) in the following steps: Collect the vibration data generated by the transformer during operation in real time and transmit the collected vibration data to the main control end; The main control end constructs a normal vibration mode model based on historical data, and then builds a vibration feature baseline of the transformer under normal operating conditions. Then, a support vector machine vibration analysis model is constructed to identify abnormal vibration modes. A threshold is set in the main control terminal according to historical data, and then the difference between the current vibration pattern and the baseline is determined according to the preset threshold. If the degree of deviation from the normal pattern exceeds the preset threshold, it is identified as an abnormal vibration pattern; When abnormal vibration is detected, the vibration signal will be compared with historical data and classified.

7. A transformer insulation fault monitoring and early warning system according to claim 6, characterized in that: The steps to construct a support vector machine vibration analysis model are: Extract time domain features from historical data and use Fourier transform to convert data from time domain to frequency domain, extract spectrum features, and perform time-frequency domain analysis on non-stationary signals to extract time-frequency features, including mean, amplitude, energy spectrum, power spectrum density, instantaneous frequency, and instantaneous phase; Select linear SVM or nonlinear SVM according to the data distribution, and then set the penalty parameter C and kernel function parameter γ of the linear SVM or nonlinear SVM, which are expressed as follows: Among them, w is the weight vector, b is the bias term, and y i is the label, C controls the penalty for classification error, and the kernel function parameter γ controls the shape of the kernel function; The data is divided into a training set and a test set. Then, the training set data is used to train the SVM model. After the training is completed, the test set data is used to verify the model and evaluate the performance of the model. Finally, the trained SVM model is deployed to the main control end for real-time monitoring and anomaly detection.

8. A transformer insulation fault monitoring and early warning system according to claim 5, characterized in that: The pending module has the functions of timed monitoring and automatic detection, which is used to track the processing status of the warning information. After the warning information is generated, the pending module starts the timer and continuously monitors the progress of the warning processing within the specified time range. If the warning information is not processed or confirmed in time within the specified response time, the pending module will automatically identify the status as timeout and unprocessed; At this point, the processing module collects multi-dimensional data related to the warning processing, and then records the time and frequency of each response to build a comprehensive processing status database; After data collection is completed, the processing module will extract key features from the raw data, including average response time, maximum and minimum response time, number of response delays, and the most recent processing time. The key features are then integrated into training sets and validation sets, and the training set is used to train the isolation forest detection model. Then, by repeatedly iteratively adjusting the model parameters, the processing module will enable the model to learn the difference between normal response patterns and potential anomalies. Each time the processing module generates new warning information, it will obtain the current processing data in real time and input this data into the trained anomaly detection model; During real-time monitoring, the processing module will use the trained model to determine whether the current response mode is abnormal. If the processing time or frequency detected by the model deviates significantly from the normal value, the system will automatically mark the state as a potential delay. This judgment will be based on the output of the model and cross-validated with historical data. In addition, the model will generate a classifier during long-term training to analyze new data inputs in real time to ensure that the model can accurately identify potential processing delays and respond to real-time data. The processing module will also set a threshold. If the current response exceeds the set range, an abnormal alarm will be triggered; After detecting an anomaly, the processing module will combine the statistical information of historical data and analyze the patterns of previous processing delays to predict response time delays. Once a delay is predicted, the processing module will immediately mark the status as high risk through internal logic, issue an alert to relevant personnel, and generate a detailed exception handling report. The exception handling report covers the time, type, and predicted processing delay of the anomaly. These reports will be provided to management for subsequent analysis to help identify system weaknesses and potential improvements.

9. A transformer insulation fault monitoring and early warning system according to claim 5, characterized in that: The signal sending module is used to send the monitored warning information to the remote user end in real time through wired or wireless communication, and the communication protocol interacts with the control center or mobile device to ensure the timely transmission of the warning information. The signal sending module also supports multi-terminal concurrent communication and can send information to multiple user ends synchronously to ensure that in an emergency, maintenance personnel or managers at different locations can receive relevant warnings at the same time.

Images