Transformer operation state monitoring method and device, electronic equipment and storage medium

Abstract

The invention provides a transformer operation state monitoring method and device, electronic equipment and a storage medium, and the method comprises the steps: collecting audio data during the operation of a target transformer in real time, and carrying out the preprocessing of the audio data; analyzing the preprocessed audio data through a trained voice recognition model to obtain an operation state analysis result corresponding to the target transformer; based on the operation state analysis result, judging whether the target transformer has an abnormal operation state, and generating a corresponding judgment result; and after determining that the target transformer has the abnormal operation state based on the judgment result, generating corresponding alarm information and pushing the alarm information. According to the transformer operation state monitoring scheme provided by the invention, comprehensive real-time monitoring of the operation state of the transformer can be realized, potential faults can be found and early warned in time, the risk of fault occurrence is reduced, the operation efficiency and safety of a power system are improved, and the intelligent level of the power system is further improved.

Description

Technical Field

[0001] The present application relates to the field of transformer detection technology, and in particular to a transformer operating status monitoring method, device, electronic equipment and storage medium. Background Art

[0002] In the power system, transformers are key power equipment, and their stable operation is crucial to ensuring the reliability and safety of power supply. However, transformers may fail during operation due to a variety of reasons, such as winding short circuit, insulation aging, overload, etc. If these failures are not discovered and handled in time, they may cause equipment damage, power supply interruption, and even safety accidents.

[0003] At present, transformer operation status monitoring mainly relies on regular manual inspections and some traditional monitoring methods, such as temperature monitoring, oil level monitoring, etc. Although these methods can reflect the operation status of the transformer to a certain extent, they have problems such as low monitoring frequency, single data, and inability to monitor in real time. Especially in the early stage of a fault, these traditional methods often fail to detect abnormalities in time, resulting in missing the best maintenance opportunity.

[0004] With the continuous development and improvement of the intelligent level of power systems, the demand for real-time and accurate monitoring of transformer operating status is growing. Existing monitoring technologies can no longer meet these needs, and a new technical solution is urgently needed to achieve comprehensive and real-time monitoring of transformer operating status, so as to timely detect and warn of potential faults, reduce the risk of faults, and improve the operating efficiency and safety of power systems.

[0005] The preceding description is intended to provide general background information and does not necessarily constitute prior art. Summary of the invention

[0006] In response to the above technical problems, the present application provides a transformer operating status monitoring method, device, electronic device and storage medium to solve the problems of low monitoring frequency, single data and inability to monitor in real time in the prior art, thereby improving the operating efficiency and safety of the power system.

[0007] In order to solve the above technical problems, the present application provides a transformer operation status monitoring method, comprising the following steps:

[0008] Collecting audio data of the target transformer in real time during operation, and preprocessing the audio data;

[0009] The preprocessed audio data is analyzed by a trained speech recognition model to obtain an operating status analysis result corresponding to the target transformer;

[0010] Based on the analysis result of the operating state, determine whether the target transformer has an abnormal operating state, and generate a corresponding judgment result;

[0011] After determining that the target transformer has an abnormal operating state based on the judgment result, generate a corresponding alarm message and push it.

[0012] Further, in some embodiments of the present application, the real-time acquisition of the audio data during the operation of the target transformer includes:

[0013] Install at least one acoustic vibration signal sensor on the target transformer; the acoustic vibration signal sensor includes a microphone array and a signal amplifier;

[0014] Based on the acoustic vibration signal sensor, real-time collect the acoustic vibration signal during the operation of the target transformer within a preset acquisition period to obtain corresponding audio data.

[0015] Further, in some embodiments of the present application, the preprocessing of the audio data includes:

[0016] Perform waveform denoising and voice endpoint detection on the audio data to obtain denoised audio data;

[0017] After uniformly processing the sampling rate of the denoised audio data, perform framing to obtain a corresponding two-dimensional framed matrix;

[0018] Perform windowing on the two-dimensional framed matrix;

[0019] Perform fast Fourier transform on the windowed two-dimensional framed matrix to obtain a corresponding frequency domain signal;

[0020] Calculate the audio features corresponding to the frequency domain signal.

[0021] Further, in some embodiments of the present application, the analysis of the preprocessed audio data by the trained speech recognition model to obtain the analysis result of the operating state corresponding to the target transformer includes:

[0022] Obtain the audio features corresponding to the preprocessed audio data;

[0023] Input the audio features into the trained speech recognition model, and output a comparison result;

[0024] Based on the comparison result, determine the analysis result of the operating state corresponding to the target transformer.

[0025] Further, in some embodiments of the present application, the determination of whether the target transformer has an abnormal operating state based on the analysis result of the operating state and the generation of a corresponding judgment result include:

[0026] Based on the state analysis result, determine whether the target transformer has an abnormal operating state to obtain an abnormal state determination result;

[0027] Based on the abnormal state determination result, perform a fault analysis on the target transformer to obtain the corresponding fault type and fault level.

[0028] Further, in some embodiments of the present application, after determining that the target transformer has an abnormal operating state based on the determination result, generating a corresponding alarm message and pushing it includes:

[0029] After determining that the target transformer has an abnormal operating state based on the determination result, determine the abnormal state determination result, fault type, and fault level corresponding to the target transformer;

[0030] Based on the abnormal state determination result, the fault type, and the fault level, generate a corresponding alarm message;

[0031] Push the alarm message to the terminal device of the monitoring personnel.

[0032] Further, in some embodiments of the present application, the training method of the voice recognition model includes:

[0033] Obtain the historical operation audio data of the transformer and perform preprocessing as sample audio data;

[0034] Obtain the sample labels corresponding to the sample audio data; the sample labels include positive sample labels for characterizing the normal operating state and negative sample labels for characterizing different types of abnormal operating states;

[0035] Based on the initial voice recognition model, identify the corresponding sample audio features according to the sample audio data;

[0036] Based on the initial voice recognition model, identify the sample audio features to obtain the corresponding operating state prediction result;

[0037] Based on the difference loss between the operating state prediction result and the sample label, adjust the model parameters of the initial voice recognition model;

[0038] Continue to train the initial voice recognition model based on the adjusted model parameters until the training conditions are met to obtain a trained initial voice recognition model.

[0039] Correspondingly, the present application also provides a transformer operating state monitoring device, including:

[0040] A data module for collecting audio data of a target transformer in real time during operation and preprocessing the audio data;

[0041] An analysis module for analyzing the preprocessed audio data through a trained speech recognition model to obtain an operation status analysis result corresponding to the target transformer;

[0042] A judgment module for judging whether the target transformer has an abnormal operation status based on the operation status analysis result and generating a corresponding judgment result;

[0043] An alarm module for generating a corresponding alarm message and pushing it after determining that the target transformer has an abnormal operation status based on the judgment result.

[0044] This application also provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor. Wherein, when the processor executes the program, the steps of the above-mentioned transformer operation status monitoring method are implemented.

[0045] This application also provides a storage medium storing a computer program that can be loaded and executed by a processor to implement the above-mentioned transformer operation status monitoring method.

[0046] Implementing the embodiments of this application has the following beneficial effects:

[0047] As described above, a transformer operation status monitoring method, device, electronic device, and storage medium provided by this application. The transformer operation status monitoring method includes: collecting audio data of a target transformer in real time during operation and preprocessing the audio data; analyzing the preprocessed audio data through a trained speech recognition model to obtain an operation status analysis result corresponding to the target transformer; judging whether the target transformer has an abnormal operation status based on the operation status analysis result and generating a corresponding judgment result; generating a corresponding alarm message and pushing it after determining that the target transformer has an abnormal operation status based on the judgment result. The transformer operation status monitoring solution provided by this application can accurately identify the operation status and potential faults of the transformer by using acoustic vibration signal recognition technology, reduce false alarms and missed alarms, realize comprehensive real-time monitoring of the operation status of the monitored transformer, and immediately issue a warning when an abnormality is detected, reducing the risk of faults, thereby ensuring the stable operation of the transformer, reducing power supply interruptions caused by transformer faults, improving the operation efficiency and safety of the entire power system, and further improving the intelligent level of the power system. Description of the Drawings

[0048] The accompanying drawings here are incorporated into and form a part of this specification, showing embodiments consistent with this application, and are used together with the specification to explain the principles of this application. To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, for those of ordinary skill in the art, other accompanying drawings can also be obtained based on these drawings without creative efforts.

[0049] Figure 1 It is a schematic diagram of an application scenario of the transformer operation state monitoring method provided by an embodiment of this application;

[0050] Figure 2 It is a schematic flowchart of the transformer operation state monitoring method provided by an embodiment of this application;

[0051] Figure 3 It is a schematic structural diagram of the transformer operation state monitoring device provided by an embodiment of this application;

[0052] Figure 4 It is a schematic structural diagram of the electronic device provided by an embodiment of this application.

[0053] The realization of the purpose of this application, functional features and advantages will be further described in conjunction with the embodiments with reference to the accompanying drawings. Through the above accompanying drawings, the clear embodiments of this application have been shown, and there will be more detailed descriptions later. These accompanying drawings and textual descriptions are not intended to limit the scope of the concept of this application in any way, but to illustrate the concept of this application to those skilled in the art by referring to specific embodiments. Detailed implementation manners

[0054] Here, the exemplary embodiments will be described in detail, and the examples are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different accompanying drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with this application. On the contrary, they are merely examples of devices and methods consistent with some aspects of this application as detailed in the appended claims.

[0055] It should be noted that in this text, the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the phrase "comprising a..." does not exclude the presence of additional identical elements in the process, method, article or device comprising that element. In addition, components, features, and elements with the same name in different embodiments of this application may have the same meaning or different meanings, and their specific meanings need to be determined based on their explanations in the specific embodiments or further in combination with the context of the specific embodiments.

[0056] It should be understood that the specific embodiments described herein are merely for explaining this application and are not used to limit this application.

[0057] In subsequent descriptions, the use of suffixes such as "module", "component" or "unit" to denote elements is only for the convenience of explaining this application and has no specific meaning in itself. Therefore, "module", "component" or "unit" can be used interchangeably.

[0058] This application provides a method, device, electronic device and storage medium for monitoring the operating state of a transformer.

[0059] Among them, the transformer operating state monitoring device can be specifically integrated in an electronic device, and the electronic device can be a smart phone, a tablet computer, a laptop computer or a desktop computer, but is not limited thereto. The electronic device can be directly or indirectly connected to the server through wired or wireless communication means. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. This application does not make any restrictions here.

[0060] Please refer to Figure 1 , Figure 1 which is an application environment diagram of the transformer operating state monitoring method in an embodiment. Refer to Figure 1, The transformer operation status monitoring method can be applied to a transformer operation status monitoring system. Among them, the transformer operation status monitoring system may include a terminal 110 and a server 120. The terminal 110 and the server 120 are connected through a network. The terminal 110 may specifically be a desktop terminal or a mobile terminal, and the mobile terminal may specifically be at least one of a mobile phone, a tablet computer, a laptop computer, etc. The server 120 may be implemented by an independent server or a server cluster composed of multiple servers. The terminal 110 is used to collect the audio data of the target transformer during operation in real time, and preprocess the audio data; analyze the preprocessed audio data through a trained speech recognition model to obtain the operation status analysis result corresponding to the target transformer; based on the operation status analysis result, judge whether the target transformer has an abnormal operation status, and generate a corresponding judgment result; after determining that the target transformer has an abnormal operation status based on the judgment result, generate a corresponding alarm message and push it.

[0061] Once a substation transformer fails, it will cause heavy losses. At the lightest, it will cause equipment failures and bring serious economic losses; at the heaviest, it will trigger a fire, seriously endangering normal safety and production. As the main substation equipment, the transformer may be affected by vibration during operation. When a short-circuit fault occurs in the side winding, each winding bears a considerable short-circuit force. For a transformer in long-term operation, due to the influence of electromagnetic force, thermal stress, electrical corrosion, operation vibration, mechanical damage, moisture, chemical corrosion, etc., various faults and hidden dangers will occur in the transformer. In order to ensure the safe operation of the transformer, parts and components that do not meet the regulations and requirements should be replaced and repaired in time, and the hidden danger parts found through detection and inspection should be regularly overhauled. Therefore, in order to work properly and extend the service life, it is very necessary to carry out daily maintenance and overhaul of the transformer. Carry out necessary maintenance and overhaul of the transformer to restore its performance, ensure the operation technical state of the transformer, and meet the needs of safety and reliability.

[0062] The following will be described in detail respectively. It should be noted that the description order of the following embodiments does not limit the priority order of the embodiments.

[0063] This application provides a transformer operation status monitoring method, including: collecting the audio data of the target transformer during operation in real time, and preprocessing the audio data; analyzing the preprocessed audio data through a trained speech recognition model to obtain the operation status analysis result corresponding to the target transformer; based on the operation status analysis result, judge whether the target transformer has an abnormal operation status, and generate a corresponding judgment result; after determining that the target transformer has an abnormal operation status based on the judgment result, generate a corresponding alarm message and push it.

[0064] Please refer to Figure 2 , Figure 2It is a schematic flowchart of the transformer operation status monitoring method provided by an embodiment of the present application. The transformer operation status monitoring method provided by this embodiment may specifically include the following steps:

[0065] S1. Real-time collect the audio data during the operation of the target transformer, and preprocess the audio data;

[0066] Specifically, for step S1, at least one acoustic vibration signal sensor is installed on the transformer. This acoustic vibration sensor includes a microphone array and a signal amplifier. The acoustic vibration sensor real-time collects the acoustic vibration signals generated during the operation of the transformer within a preset collection period, and converts the acoustic vibration signals into audio data. In addition, besides the acoustic vibration signal sensor, other types of sensors (such as temperature and pressure sensors) can also be integrated to provide more comprehensive transformer operation status information.

[0067] In addition, the accuracy and reliability of signal collection can be improved, and the efficiency of signal processing can be enhanced through the microphone array for voiceprint data collection and signal processing algorithms.

[0068] S2. Analyze the preprocessed audio data through a trained speech recognition model to obtain the operation status analysis result corresponding to the target transformer;

[0069] Specifically, for step S2, the collected audio data needs to be preprocessed, including waveform denoising and voice endpoint detection, to obtain the denoised audio data. Then, after the denoised audio data is subjected to sampling rate unification processing, it is framed to form a two-dimensional framed matrix. Then, windowing processing is performed on these matrices, and the time-domain signal is converted into a frequency-domain signal through the fast Fourier transform (FFT). Finally, audio features are calculated.

[0070] S3. Based on the operation status analysis result, determine whether the target transformer has an abnormal operation status, and generate a corresponding judgment result;

[0071] Specifically, for step S3, the preprocessed audio data is input into a trained speech recognition model. This model can recognize audio features (such as MFCC features) and output comparison results. Based on these comparison results, the operation status analysis result of the transformer is determined. In addition, potential faults can be predicted by collecting and analyzing historical data and using machine learning algorithms, so as to achieve predictive maintenance and reduce unexpected downtime. Through further research and development of more advanced audio processing and pattern recognition algorithms, such as deep learning techniques, the accuracy and efficiency of fault detection can be improved.

[0072] S4. After it is determined based on the judgment result that the target transformer has an abnormal operation status, generate a corresponding warning message and push it;

[0073] Specifically, for step S4, when it is determined based on the judgment result that the transformer is in an abnormal operating state, the system will automatically generate corresponding alarm information and push it to the monitoring personnel's terminal device for timely maintenance actions.

[0074] In a specific embodiment, a recording device is installed on the transformer. During the operation of the transformer, recordings are made and uploaded to the server through a communication network. When the transformer produces abnormal sounds and uploads them to the server, the server uses Python for voice feature analysis and feature comparison. If it is found that the current sound belongs to the alarm range after analysis and comparison, an alarm is automatically executed. Among them, the management system uses the Python Flask framework, can perform real-time monitoring on a large screen by combining Echarts, and manages data and voice files through a MySQL database.

[0075] It can be seen that the transformer operating state monitoring method provided in this embodiment can prevent safety accidents caused by transformer failures and ensure the stable operation of the power system through real-time monitoring and timely alarming; predictive maintenance reduces the need for unexpected downtime and emergency repairs, thereby reducing maintenance costs; the automated monitoring and diagnosis process reduces the need for manual inspections and improves the efficiency of maintenance work; by detecting and repairing faults in a timely manner, the service life of the transformer can be extended and the equipment replacement frequency can be reduced. In addition, more accurate maintenance and operation decisions can be helped for management personnel by providing detailed fault analysis and status reports.

[0076] Furthermore, in some embodiments, "real-time collecting audio data during the operation of the target transformer" in step S1 may specifically include:

[0077] Install at least one acoustic-vibration signal sensor on the target transformer; the acoustic-vibration signal sensor includes a microphone array and a signal amplifier;

[0078] Based on the acoustic-vibration signal sensor, in a preset acquisition period, real-time collect the acoustic-vibration signal during the operation of the target transformer to obtain corresponding audio data.

[0079] Specifically, for the acquisition of audio data in step S1, first install at least one acoustic-vibration signal sensor on the transformer, which consists of a microphone array and a signal amplifier. The microphone array is responsible for capturing the sound waves generated during the operation of the transformer, while the signal amplifier is used to enhance these signals for more accurate subsequent processing. Based on the acoustic-vibration signal sensor, in a preset acquisition period, real-time collect the acoustic-vibration signal generated during the operation of the transformer. These signals are directly related to the operating state of the transformer, and any abnormal vibration or noise may be a precursor to a fault. The acoustic-vibration signals collected by the acoustic-vibration signal sensor are converted into audio data, which will be used for subsequent preprocessing and analysis steps.

[0080] In specific embodiments, more sensitive and durable acoustic and vibration signal sensors are studied and developed to improve the quality and reliability of signal acquisition. For example, sensors with waterproof, dustproof, and anti-interference capabilities are used to adapt to various environmental conditions. A wireless transmission module, such as Wi-Fi or 4G / 5G, can also be integrated to enable real-time wireless transmission of audio data to the central processing system, reducing wiring complexity and enhancing the flexibility and scalability of the system. In addition to audio data, other types of sensor data (such as temperature, humidity, pressure, etc.) can be integrated to provide a more comprehensive analysis of the transformer's operating status through multi-source data fusion technology. An adaptive sampling rate technology is provided to dynamically adjust the sampling rate according to the actual operating status of the transformer to optimize data quality and processing efficiency.

[0081] In this embodiment, through high-quality acoustic and vibration signal sensors and advanced signal processing technologies, the operating sound of the transformer can be captured and analyzed more accurately, improving the accuracy of fault detection; the application of wireless transmission technology and adaptive sampling rate makes the system more stable and reliable, and it can work properly even in harsh environments or high-interference conditions; through real-time monitoring and predictive maintenance, the need for unexpected downtime and emergency repairs can be reduced, thereby reducing maintenance costs and increasing economic benefits; wireless transmission and multi-source data fusion technologies make the system more flexible and can be easily extended to more transformers or integrated with other monitoring systems.

[0082] Further, in some embodiments, the "preprocessing of audio data" in step S1 may specifically include:

[0083] Performing waveform denoising and voice endpoint detection on the audio data to obtain denoised audio data;

[0084] Performing unified sampling rate processing on the denoised audio data and then framing it to obtain a corresponding two-dimensional framed matrix;

[0085] Performing windowing processing on the two-dimensional framed matrix;

[0086] Performing fast Fourier transform on the windowed two-dimensional framed matrix to obtain a corresponding frequency-domain signal;

[0087] Calculating the audio features corresponding to the frequency-domain signal.

[0088] Specifically, for the data preprocessing in step S1, first, perform waveform denoising on the collected audio data to eliminate background noise and irrelevant signals and improve the accuracy of subsequent analysis. This usually involves applying filters and signal processing techniques. Perform voice activity detection on the denoised audio data to determine the effective audio segments, that is, the parts containing the operating sound of the transformer, for further analysis. Perform unified sampling rate processing on the denoised audio data to ensure that all data is analyzed at the same sampling rate for easy comparison and processing. Perform frame segmentation on the audio data with unified sampling rate, that is, divide the continuous audio signal into a series of short-time frames to form a two-dimensional frame matrix. Apply windowing to the frame matrix to reduce the discontinuity between frames, reduce spectral leakage, and improve the accuracy of frequency-domain analysis. Perform fast Fourier transform on the windowed two-dimensional frame matrix to convert the time-domain signal into a frequency-domain signal for analyzing the frequency components of the audio. Calculate the audio features corresponding to the frequency-domain signal, which will be used for subsequent fault diagnosis and condition monitoring.

[0089] In a specific embodiment, first perform filter denoising and voice activity detection on the audio. Define a filter, set the filter threshold size, and apply mean filter denoising to the audio data. After denoising, read the time series of the audio by unifying the sampling rate, convert it into a two-dimensional frame matrix by frame segmentation, and further apply windowing to the frame matrix to eliminate the discontinuity between frames after frame segmentation and reduce spectral leakage. Then, perform fast Fourier transform (FFT) on the frames to convert the time-domain signal of the frames into frequency-domain information, divide the continuous frequencies into a series of Mel frequencies, then design a filter at each Mel frequency, perform logarithmic operation on the distributed spectra of different frequencies, perform Mel-frequency cepstral coefficient calculation with enhanced low frequencies and weakened high frequencies, and finally perform cepstrum and discrete cosine transform (DCT) to obtain the cepstral coefficient mfcc feature vector.

[0090] In addition, more advanced denoising algorithms, such as adaptive filters and wavelet transforms, can be integrated to improve the denoising effect and retain more useful signal features. Utilize deep learning techniques, such as convolutional neural networks (CNNs) or recurrent neural networks (RNNs), for more accurate voice activity detection; develop adaptive frame segmentation techniques to dynamically adjust the frame size and overlap according to the characteristics of the audio signal to optimize feature extraction. In addition to FFT, other frequency-domain analysis techniques, such as wavelet transforms, can be applied for multi-scale frequency-domain analysis to capture features at different frequency levels. Automatically select the most representative audio features through machine learning techniques to reduce the data dimension and improve the analysis efficiency and accuracy.

[0091] In this embodiment, through effective denoising and endpoint detection, the quality of the audio signal can be significantly improved, providing clearer data for subsequent fault diagnosis; unified sampling rate and precise frame division and windowing processing ensure the accuracy of frequency-domain analysis and provide a reliable basis for feature extraction; through multi-scale frequency-domain analysis and optimized feature selection, the abnormal operating state of the transformer can be more accurately identified, improving the accuracy of fault diagnosis; the adaptive frame division technology and feature optimization can reduce unnecessary data processing and improve the efficiency of the entire monitoring system; through adaptive technology and multi-scale analysis, the system can better adapt to different operating environments and conditions, improving the robustness and reliability of the system.

[0092] Further, in some embodiments, step S2, "analyze the preprocessed audio data through the trained speech recognition model to obtain the analysis result of the operating state corresponding to the target transformer", may specifically include:

[0093] Obtain the audio features corresponding to the preprocessed audio data;

[0094] Input the audio features into the trained speech recognition model and output the comparison result;

[0095] Based on the comparison result, determine the analysis result of the operating state corresponding to the target transformer.

[0096] Specifically, for step S2, first, extract audio features from the preprocessed audio data. The audio features may specifically include frequency components, energy distribution, time series features, etc., which can reflect the operating state of the transformer. Then, input the extracted audio features into a pre-trained speech recognition model. This model can recognize and analyze these features and associate them with the known operating states of the transformer. Then, the speech recognition model will output the comparison result, which is based on the model's analysis and indicates the degree of matching between the input audio features and the normal or abnormal states learned during model training. Finally, based on the comparison result of the speech recognition model, determine the analysis result of the operating state of the transformer. If the comparison result shows a high degree of matching between the audio features and the features of the abnormal state, it may indicate that there is an abnormal operation of the transformer.

[0097] In specific embodiments, combine multiple different speech recognition models and use ensemble learning methods (such as voting, stacking) to improve the accuracy of fault diagnosis; provide an update mechanism to update the speech recognition model in real time and continuously optimize the model performance according to the latest operating data and fault cases. In addition to supervised learning models, unsupervised anomaly detection algorithms, such as clustering analysis, isolation forest, etc., can also be integrated to identify unknown abnormal patterns.

[0098] In this embodiment, by using the trained speech recognition model, the operating state of the transformer can be more accurately recognized, reducing false alarms and missed detections; data augmentation for model training and real-time model updates enable the system to adapt to new operating conditions and fault modes, improving the long-term stability of the system; multi-model fusion and cross-domain transfer learning can enhance the intelligence level of the system, enabling it to handle more complex monitoring tasks; accurate fault diagnosis and anomaly detection can reduce unnecessary maintenance work, lower maintenance costs, and improve economic efficiency; real-time model updates and efficient anomaly detection algorithms can accelerate the system's response speed to abnormal situations, take timely measures, and reduce potential losses.

[0099] Further, in some embodiments, step S3, "Based on the operating state analysis result, determine whether the target transformer has an abnormal operating state and generate a corresponding judgment result", may specifically include:

[0100] Based on the state analysis result, determine whether the target transformer has an abnormal operating state to obtain an abnormal state judgment result;

[0101] Based on the abnormal state judgment result, conduct a fault analysis on the target transformer to obtain the corresponding fault type and fault level.

[0102] Specifically, for step S3, using the operating state analysis results obtained from the speech recognition model, these results provide detailed information about the operating state of the transformer. Based on the state analysis result, the system will determine whether the transformer has an abnormal operating state. This generally involves comparing the analysis result with preset normal operating parameters. If the judgment result indicates that the transformer has an abnormal operating state, the system will further conduct a fault analysis to determine the specific fault type and fault level. Finally, a clear judgment result will be generated, indicating whether the operating state of the transformer is normal and the detailed information of any detected abnormal state.

[0103] In addition, more advanced intelligent fault diagnosis algorithms, such as machine learning-based classification algorithms, can be integrated to improve the accuracy and speed of fault judgment; in addition to audio data analysis, other operating parameters of the transformer (such as temperature, oil level, current, etc.) can be combined for comprehensive judgment to improve the comprehensiveness of diagnosis; real-time data stream processing technologies, such as Apache Kafka or Apache Flink, can be adopted to achieve real-time monitoring and rapid response to the operating state of the transformer; an adaptive threshold setting mechanism can be developed to dynamically adjust the judgment threshold according to historical data and environmental changes to improve the adaptability and accuracy of the system.

[0104] In this embodiment, through the intelligent fault diagnosis algorithm and multi-parameter fusion judgment, the system can more accurately identify the abnormal state of the transformer and reduce misjudgment; the real-time data stream processing technology enables the system to quickly respond to the operation changes of the transformer, detect and respond to potential faults in a timely manner; the adaptive threshold setting enables the system to adapt to different operating environments and conditions, improving the stability and reliability of the system.

[0105] Further, in some embodiments, step S4, "After determining that the target transformer has an abnormal operating state based on the judgment result, generate a corresponding alarm message and push it", may specifically include:

[0106] After determining that the target transformer has an abnormal operating state based on the judgment result, determine the abnormal state judgment result, fault type, and fault level corresponding to the target transformer;

[0107] Based on the abnormal state judgment result, fault type, and fault level, generate a corresponding alarm message;

[0108] Push the alarm message to the terminal device of the monitoring personnel.

[0109] Specifically, for step S4, after confirming that the transformer has an abnormal operating state, the system will determine the specific judgment result of the abnormal state, including the specific type and severity of the abnormality; based on the abnormal state judgment result, the system will further analyze and determine the fault type and fault level to provide precise guidance for subsequent maintenance and handling. According to the abnormal state judgment result, fault type, and fault level, the system will generate corresponding alarm messages. These messages should contain sufficient details so that the operation and maintenance personnel can quickly understand the situation and take actions. The generated alarm messages will be pushed to the terminal devices of the monitoring personnel, such as mobile phones, computers, or other dedicated monitoring devices, to ensure that the information can be conveyed in a timely manner.

[0110] In a specific embodiment, a multi-channel alarm system is provided to push alarm messages through multiple methods such as text messages, emails, mobile applications, etc. to ensure the timely conveyance of information; allow users to customize the content and format of the alarm messages according to specific needs, including the level of detail of the alarm, technical parameters, etc., to meet the needs of different users; in addition, this embodiment also provides an intelligent alarm suppression algorithm to reduce false alarms and repeated alarms, improving the accuracy and efficiency of the alarm system. Include guiding steps for fault response in the alarm message, such as recommended inspection procedures, spare parts that may be required, etc., to help the operation and maintenance personnel respond quickly. Establish an alarm history record system for storing and analyzing alarm messages to facilitate fault trend analysis and predictive maintenance.

[0111] In this embodiment, through timely and multi-channel alarm push, the response speed of operation and maintenance personnel to abnormal situations can be accelerated, and potential downtime can be reduced; intelligent alarm suppression and customized alarm information can reduce false alarms and irrelevant alarms, improving the accuracy and relevance of alarms; alarm information containing response guidance can help operation and maintenance personnel handle faults more effectively, reducing fault handling time; alarm history records and analysis can help management make more reasonable maintenance decisions, optimize maintenance plans and resource allocation; by reducing false alarms and improving the accuracy of alarms, the overall reliability of the system is enhanced, reducing unnecessary maintenance work caused by misjudgment.

[0112] Further, in some embodiments, the training method of the speech recognition model may specifically include:

[0113] Obtain the historical operation audio data of the transformer and perform preprocessing to obtain sample audio data.

[0114] Obtain the sample labels corresponding to the sample audio data; the sample labels include positive sample labels for characterizing normal operation states and negative sample labels for characterizing different types of abnormal operation states.

[0115] Based on the initial speech recognition model, identify the corresponding sample audio features according to the sample audio data.

[0116] Based on the initial speech recognition model, identify the sample audio features to obtain the corresponding operation state prediction results.

[0117] Based on the difference loss between the operation state prediction results and the sample labels, adjust the model parameters of the initial speech recognition model.

[0118] Based on the adjusted model parameters, continue to train the initial speech recognition model until the training conditions are met, and obtain the trained initial speech recognition model.

[0119] Specifically, for the speech recognition model in this embodiment, this embodiment also provides a specific training process. First, it is necessary to obtain the historical operation audio data of the transformer and perform preprocessing on it. These data will be used as sample audio data for training the model. Corresponding to the sample audio data are sample labels, including positive sample labels (characterizing normal operation states) and negative sample labels (characterizing different types of abnormal operation states). Use the initial speech recognition model to identify the corresponding sample audio features according to the sample audio data. According to the sample audio features identified by the initial speech recognition model, predict the corresponding operation state prediction results. Calculate the difference loss between the operation state prediction results and the sample labels according to the preset loss function, and thus adjust the model parameters of the initial speech recognition model based on the difference loss. Based on the adjusted model parameters, continue to train the initial speech recognition model until the training conditions are met, and finally obtain the trained speech recognition model.

[0120] In a specific embodiment, by customizing various faults and noises, the dataset of a normal transformer is classified. The audio is uniformly sampled in terms of sampling rate, duration, etc., and then model training is carried out. After the training is completed, the model is evaluated to verify its performance, and the model is optimized by adjusting parameters such as the model structure, optimizer, and learning rate. Finally, features are proposed. The audio recognition engine monitors the sound of the transformer in real time and extracts features; through feature classification and recognition comparison, faults are analyzed and warning alerts are issued.

[0121] Among them, the speech recognition model of this embodiment can perform AI language recognition through big data. The specific implementation method is to use the PaddleSpeech deep learning model. First, according to the analysis of the custom dataset, the format and sampling rate are unified, and the audio duration is uniformly cropped; PaddleSpeech provides a variety of speech recognition models, such as DeepSpeech, Conformer, etc. Training is carried out according to the specified training set, validation set, model structure, optimizer, learning rate, etc., and the usability of model training is determined by indicators such as loss function and accuracy; after the training is completed, the model needs to be evaluated to verify its performance. If the evaluation result is not ideal, the model can be optimized by adjusting parameters such as the model structure, optimizer, and learning rate. Finally, the audio file is recognized and classified through the PaddleSpeech features.

[0122] In addition, this embodiment integrates deep learning frameworks such as TensorFlow and PyTorch to support more complex model structures and more efficient training processes; uses reinforcement learning techniques to enable the model to continuously self-optimize during actual operation and improve the accuracy of fault recognition; combines audio data and other sensor data (such as vibration, temperature, etc.) to improve the diagnostic ability of the model through multi-modal learning; in order to deploy the model on resource-constrained devices, model compression techniques can be used to reduce the model size while maintaining its accuracy; uses models pre-trained in other fields (such as speech recognition, natural language processing), and quickly adapts to the transformer condition monitoring task through transfer learning.

[0123] Through meticulous sample data preprocessing and model parameter adjustment in this embodiment, the recognition accuracy of the model for the operating state of the transformer can be significantly improved; multi-modal data fusion and cross-domain model migration can enhance the generalization ability of the model, enabling it to adapt to different operating environments and conditions; model compression and optimization techniques can improve the operating efficiency of the model on edge devices and reduce computational resource consumption; reinforcement learning techniques enable the model to continuously learn and optimize in actual applications and adapt to new fault modes and operating conditions; the trained model can quickly and accurately identify faults, thus shortening the fault response time and reducing potential downtime and maintenance costs.

[0124] It can be seen that the transformer operation status monitoring method provided in this embodiment utilizes theories and technologies in aspects such as acoustics, AI, machine learning, and big data to develop a research system for monitoring the acoustic fingerprints of transformer faults, including the identification and alarm of transformer faults. It realizes the research and development of real-time acoustic fingerprint acquisition devices, solves the problem of truly acquiring the complex sound field during the operation of the transformer, and establishes a normal operation model of the transformer equipment based on acoustics. It realizes the monitoring of the operation status of the transformer and the accurate identification of fault types.

[0125] Compared with the prior art, this embodiment can monitor transformer faults in real time and give timely warnings. Then, by analyzing multi-dimensional characteristic signals through spectrogram analysis, relevant faults can be identified, making up for the deficiencies of traditional monitoring means, forming an online monitoring and diagnosis system that can be applied in actual production, providing fault warnings, and effectively ensuring the safe operation of equipment. Through the monitoring of acoustic technology, the life cycle management of the transformer is constructed; planned maintenance is changed to condition-based maintenance; the service life of the equipment is extended; through the recommended maintenance of the acoustic online monitoring equipment, the service life can be extended by 10%, and hundreds of thousands of yuan can be saved for the maintenance of each transformer.

[0126] By implementing this embodiment, the online monitoring ability of the transformer based on acoustic fingerprint monitoring is formed, which can significantly reduce the maintenance waiting time, improve the healthy working time of the transformer, reduce the enterprise maintenance management cost, improve the equipment maintenance efficiency, and reduce the dependence on the professional ability of maintenance personnel. At the same time, a large amount of acoustic fingerprint data accumulated during the implementation process can be used to establish an advanced acoustic fingerprint database to help enterprises trace data and track faults, achieving the full life cycle management of equipment. An acoustic fingerprint monitoring system with independent intellectual property rights is established, breaking through the complex fault acoustic fingerprint analysis technology, and assisting enterprises to establish an acoustic fingerprint monitoring team with the ability to analyze and diagnose complex faults. Moreover, it can be extended to the non-contact online monitoring and warning of other power equipment, comprehensively and real-time monitor the operation status of power equipment, improve the intelligent and digital management level of enterprises, and help enterprises with digital upgrading.

[0127] In summary, the present embodiment provides a method for monitoring the operating state of a transformer, which includes collecting audio data during the operation of the target transformer in real time and preprocessing the audio data; analyzing the preprocessed audio data through a trained speech recognition model to obtain an operating state analysis result corresponding to the target transformer; based on the operating state analysis result, determining whether the target transformer has an abnormal operating state and generating a corresponding judgment result; after determining that the target transformer has an abnormal operating state based on the judgment result, generating a corresponding warning message and pushing it. The transformer operating state monitoring solution provided by the present embodiment can accurately identify the operating state and potential faults of the transformer by using the acoustic vibration signal recognition technology, reduce false alarms and missed alarms, realize comprehensive real-time monitoring of the operating state of the monitored transformer, and immediately issue a warning when an abnormality is detected, reducing the risk of faults, thereby ensuring the stable operation of the transformer, reducing power supply interruptions caused by transformer faults, improving the operating efficiency and safety of the entire power system, and further improving the intelligent level of the power system.

[0128] To facilitate better implementation of the method for monitoring the operating state of a transformer in the embodiments of the present application, the embodiments of the present application also provide a device for monitoring the operating state of a transformer. The meanings of the terms are the same as those in the above method for monitoring the operating state of a transformer, and the specific implementation details can refer to the descriptions in the method embodiments.

[0129] Please refer to Figure 3 , Figure 3 which is a schematic structural diagram of the device for monitoring the operating state of a transformer provided by the embodiments of the present application. The device for monitoring the operating state of a transformer may specifically include a data module 201, an analysis module 202, a judgment module 203, and an alarm module 204, which are specifically as follows:

[0130] The data module 201 is used to collect audio data during the operation of the target transformer in real time and preprocess the audio data;

[0131] The analysis module 202 is used to analyze the preprocessed audio data through a trained speech recognition model to obtain an operating state analysis result corresponding to the target transformer;

[0132] The judgment module 203 is used to determine whether the target transformer has an abnormal operating state based on the operating state analysis result and generate a corresponding judgment result;

[0133] The alarm module 204 is used to generate a corresponding warning message and push it after determining that the target transformer has an abnormal operating state based on the judgment result.

[0134] Furthermore, in some embodiments, the data module 201 may specifically include:

[0135] The data acquisition unit is used to install at least one acoustic vibration signal sensor on the target transformer; the acoustic vibration signal sensor includes a microphone array and a signal amplifier; based on the acoustic vibration signal sensor, the acoustic vibration signal during the operation of the target transformer is collected in real time within a preset acquisition period to obtain corresponding audio data.

[0136] The data preprocessing unit is used to perform waveform denoising and voice endpoint detection on the audio data to obtain denoised audio data; after uniformly processing the sampling rate of the denoised audio data, it is framed to obtain a corresponding two-dimensional framed matrix; the two-dimensional framed matrix is windowed; the windowed two-dimensional framed matrix is subjected to fast Fourier transform to obtain a corresponding frequency domain signal; the audio features corresponding to the frequency domain signal are calculated.

[0137] Further, in some embodiments, the analysis module 202 may specifically include:

[0138] The acquisition unit is used to acquire the audio features corresponding to the preprocessed audio data;

[0139] The comparison unit is used to input the audio features into the trained speech recognition model and output a comparison result;

[0140] The analysis unit is used to determine the operation state analysis result corresponding to the target transformer based on the comparison result.

[0141] Further, in some embodiments, the judgment module 203 may specifically include:

[0142] The first judgment unit is used to judge whether the target transformer has an abnormal operation state based on the state analysis result to obtain an abnormal state judgment result;

[0143] The second judgment unit is used to perform fault analysis on the target transformer based on the abnormal state judgment result to obtain the corresponding fault type and fault level.

[0144] Further, in some embodiments, the alarm module 204 may specifically include:

[0145] The fault identification unit is used to determine the abnormal state judgment result, fault type and fault level corresponding to the target transformer after determining that the target transformer has an abnormal operation state based on the judgment result;

[0146] The alarm unit is used to generate corresponding alarm information based on the abnormal state judgment result, fault type and fault level;

[0147] The push unit is used to push the alarm information to the terminal device of the monitoring personnel.

[0148] Further, in some embodiments, it further includes a model training module, which may specifically be used for:

[0149] Obtain the historical operating audio data of the transformer and preprocess it as sample audio data;

[0150] Obtain the sample labels corresponding to the sample audio data; the sample labels include positive sample labels for characterizing the normal operating state and negative sample labels for characterizing different types of abnormal operating states;

[0151] Based on the initial speech recognition model, identify the corresponding sample audio features according to the sample audio data;

[0152] Based on the initial speech recognition model, identify the sample audio features to obtain the corresponding operating state prediction results;

[0153] Based on the difference loss between the operating state prediction results and the sample labels, adjust the model parameters of the initial speech recognition model;

[0154] Based on the adjusted model parameters, continue to train the initial speech recognition model until the training conditions are met, and obtain the trained initial speech recognition model.

[0155] In summary, the transformer operating state monitoring device provided in this embodiment collects the audio data during the operation of the target transformer in real time through the data module 201, and preprocesses the audio data; the analysis module 202 analyzes the preprocessed audio data according to the trained speech recognition model to obtain the operating state analysis result corresponding to the target transformer; the judgment module 203 judges whether the target transformer has an abnormal operating state based on the operating state analysis result, and generates a corresponding judgment result; after determining that the target transformer has an abnormal operating state based on the judgment result, the alarm module 204 generates a corresponding alarm message and pushes it. It can be seen that the transformer operating state monitoring device provided in this embodiment can accurately identify the operating state and potential faults of the transformer by using the acoustic vibration signal recognition technology, reduce false alarms and missed alarms, realize the comprehensive real-time monitoring of the operating state of the monitored transformer, and immediately issue a warning when an abnormality is detected, reducing the risk of faults, thereby ensuring the stable operation of the transformer, reducing the power supply interruption caused by transformer faults, improving the operation efficiency and safety of the entire power system, and further improving the intelligent level of the power system.

[0156] In addition, the embodiment of the present application also provides an electronic device, as Figure 4 shown, which shows the structural schematic diagram of the electronic device involved in the embodiment of the present application. Specifically: the electronic device may include a processor 301 with one or more processing cores, a memory 302 with one or more computer-readable storage media, a power supply 303, an input unit 304 and other components. Those skilled in the art can understand, Figure 4The structure of the electronic device shown does not limit the electronic device, and it may include more or fewer components than shown, or combine certain components, or have different component arrangements. Among them:

[0157] The processor 301 is the control center of the electronic device. It connects various parts of the entire electronic device using various interfaces and circuits. By running or executing software programs and / or modules stored in the memory 302, and by invoking the data stored in the memory 302, it executes various functions of the electronic device and processes data, thereby monitoring the electronic device as a whole. Optionally, the processor 301 may include one or more processing cores; preferably, the processor 301 may integrate an application processor and a modem processor. Among them, the application processor mainly processes the operating system, user interface, application programs, etc., and the modem processor mainly processes wireless communication. It can be understood that the above-mentioned modem processor may not be integrated into the processor 301.

[0158] The memory 302 can be used to store software programs and modules. The processor 301 executes various functional applications and the transformer operation state monitoring method by running the software programs and modules stored in the memory 302. The memory 302 mainly includes a program storage area and a data storage area. Among them, the program storage area can store the operating system, application programs required for at least one function (such as the sound playback function, image playback function, etc.); the data storage area can store data created according to the use of the electronic device. In addition, the memory 302 may include high-speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other volatile solid-state storage devices. Correspondingly, the memory 302 may also include a memory controller to provide the processor 301 with access to the memory 302.

[0159] The electronic device also includes a power supply 303 that powers each component. Preferably, the power supply 303 can be logically connected to the processor 301 through a power management system, so as to realize functions such as management of charging, discharging, and power consumption management through the power management system. The power supply 303 may also include any components such as one or more DC or AC power supplies, a recharge system, a power failure detection circuit, a power converter or inverter, and a power status indicator.

[0160] The electronic device may also include an input unit 304, which can be used to receive input digital or character information, and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function controls.

[0161] Although not shown, the electronic device may further include a display unit and the like, which will not be elaborated here. Specifically, in this embodiment, the processor 301 in the electronic device will load the executable files corresponding to the processes of one or more application programs into the memory 302 according to the following instructions, and the processor 301 will run the application programs stored in the memory 302 to implement various functions as follows:

[0162] Collect the audio data of the target transformer in real time during operation, and preprocess the audio data; analyze the preprocessed audio data through a trained speech recognition model to obtain the operation state analysis result corresponding to the target transformer; based on the operation state analysis result, judge whether the target transformer has an abnormal operation state, and generate a corresponding judgment result; after determining that the target transformer has an abnormal operation state based on the judgment result, generate a corresponding alarm message and push it.

[0163] For the specific implementation of each of the above operations, reference may be made to the previous embodiments, which will not be elaborated here.

[0164] By using the acoustic vibration signal recognition technology, the embodiments of the present application can accurately identify the operation state and potential faults of the transformer, reduce false alarms and missed alarms, realize comprehensive real-time monitoring of the operation state of the monitored transformer, and immediately issue a warning when an abnormality is detected, reducing the risk of faults, thereby ensuring the stable operation of the transformer, reducing the power supply interruption caused by transformer faults, improving the operation efficiency and safety of the entire power system, and further improving the intelligent level of the power system.

[0165] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructions, or by instructions controlling related hardware. The instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0166] Therefore, the embodiments of the present application provide a storage medium, which stores multiple instructions that can be loaded by a processor to execute the steps in any of the transformer operation state monitoring methods provided by the embodiments of the present application. For example, the instructions can execute the following steps:

[0167] Collect the audio data of the target transformer in real time during operation, and preprocess the audio data; analyze the preprocessed audio data through a trained speech recognition model to obtain the operation state analysis result corresponding to the target transformer; based on the operation state analysis result, judge whether the target transformer has an abnormal operation state, and generate a corresponding judgment result; after determining that the target transformer has an abnormal operation state based on the judgment result, generate a corresponding alarm message and push it.

[0168] For the specific implementation of each of the above operations, reference may be made to the previous embodiments, which will not be elaborated herein.

[0169] Among them, the storage medium may include: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk, etc. Since the instructions stored in the storage medium can execute the steps in any of the transformer operation state monitoring methods provided by the embodiments of the present application, the beneficial effects achievable by any of the transformer operation state monitoring methods provided by the embodiments of the present application can be realized. For details, refer to the previous embodiments, which will not be elaborated herein.

[0170] The above has introduced in detail a transformer operation state monitoring method, device, electronic device and storage medium provided by the embodiments of the present application. Specific examples are used herein to elaborate on the principle and implementation manner of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those skilled in the art, according to the idea of the present application, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A transformer operating status monitoring method, characterized in that: The steps include: Collecting audio data of the target transformer in real time during operation, and preprocessing the audio data; The preprocessed audio data is analyzed by a trained speech recognition model to obtain an operating status analysis result corresponding to the target transformer; Based on the operating status analysis result, determine whether the target transformer is in an abnormal operating state, and generate a corresponding determination result; After determining that the target transformer is in an abnormal operating state based on the judgment result, corresponding alarm information is generated and pushed.

2. The transformer operating status monitoring method according to claim 1, characterized in that: The real-time collection of audio data of the target transformer during operation includes: Installing at least one acoustic vibration signal sensor on the target transformer; the acoustic vibration signal sensor includes a microphone array and a signal amplifier; The acoustic vibration signal sensor collects the acoustic vibration signal of the target transformer during operation in real time within a preset collection period to obtain corresponding audio data.

3. The transformer operating status monitoring method according to claim 1, characterized in that: The preprocessing of the audio data comprises: Performing waveform denoising and voice endpoint detection on the audio data to obtain denoised audio data; The denoised audio data is subjected to uniform sampling rate processing and then framed to obtain a corresponding two-dimensional frame matrix; Performing windowing processing on the two-dimensional frame matrix; Perform fast Fourier transform on the two-dimensional frame matrix after windowing to obtain the corresponding frequency domain signal; The audio feature corresponding to the frequency domain signal is calculated.

4. The transformer operating status monitoring method according to claim 1, characterized in that: The pre-processed audio data is analyzed by the trained speech recognition model to obtain the operating status analysis result corresponding to the target transformer, including: Obtaining audio features corresponding to the preprocessed audio data; Input the audio features into a trained speech recognition model and output a comparison result; Based on the comparison result, an operation status analysis result corresponding to the target transformer is determined.

5. The transformer operating status monitoring method according to claim 1, characterized in that: The step of judging whether the target transformer is in an abnormal operating state based on the operating state analysis result and generating a corresponding judgment result includes: Based on the state analysis result, determine whether the target transformer is in an abnormal operating state, and obtain an abnormal state determination result; Based on the abnormal state judgment result, a fault analysis is performed on the target transformer to obtain a corresponding fault type and fault level.

6. The transformer operating status monitoring method according to claim 1, characterized in that: After determining that the target transformer is in an abnormal operating state based on the judgment result, generating corresponding alarm information and pushing it, including: After determining that the target transformer is in an abnormal operating state based on the judgment result, determining the abnormal state judgment result, fault type and fault level corresponding to the target transformer; Generate corresponding alarm information based on the abnormal state judgment result, the fault type and the fault level; The alarm information is pushed to the terminal device of the monitoring personnel.

7. The transformer operating status monitoring method according to claim 1, characterized in that: The training method of the speech recognition model includes: Acquire historical operating audio data of the transformer and pre-process it as sample audio data; Obtaining sample labels corresponding to the sample audio data; the sample labels include positive sample labels for characterizing normal operating states and negative sample labels for characterizing different types of abnormal operating states; Based on the initial speech recognition model, identifying corresponding sample audio features according to the sample audio data; Identify the sample audio features based on the initial speech recognition model to obtain a corresponding running state prediction result; Adjusting the model parameters of the initial speech recognition model based on the difference loss between the running state prediction result and the sample label; The initial speech recognition model is continuously trained based on the adjusted model parameters until the training conditions are met, thereby obtaining a trained initial speech recognition model.

8. A transformer operating status monitoring device, characterized in that: include: A data module, used for real-time acquisition of audio data of the target transformer during operation, and preprocessing the audio data; An analysis module is used to analyze the preprocessed audio data through a trained speech recognition model to obtain an analysis result of the operating status corresponding to the target transformer; A judgment module, used to judge whether the target transformer has an abnormal operating state based on the operating state analysis result, and generate a corresponding judgment result; The alarm module is used to generate and push corresponding alarm information after determining that the target transformer is in an abnormal operating state based on the judgment result.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the transformer operating status monitoring method according to any one of claims 1 to 7 are implemented.

10. A storage medium, characterized in that: A computer program is stored which can be loaded by a processor and execute the transformer operation status monitoring method according to any one of claims 1 to 7.

Images