Transformer oil temperature foreground and background synchronization consistency verification system based on image recognition algorithm technology

Abstract

The invention provides a transformer oil temperature foreground and background synchronization consistency verification system based on an image recognition algorithm technology. The system comprises a standard temperature acquisition module which adopts an oil temperature sensor to monitor the transformer oil temperature in real time and transmits the transformer oil temperature to a background server; the image acquisition and recognition module is used for acquiring an image of the oil temperature meter in real time through an image acquisition module which is deployed on a transformer site and is used for capturing the oil temperature meter in real time, transmitting the image to the background server, and recognizing the acquired image to obtain an acquisition result of the reading of the oil temperature meter on site; and the foreground and background consistency checking module is used for checking the oil temperature meter reading result and the oil temperature background data which are acquired at the same time and are identified by the image acquisition and identification module respectively by taking the acquisition result of the standard temperature acquisition module as a checking reference, and outputting a checking result.

Description

Technical Field

[0001] The present invention belongs to the technical fields of transformer oil temperature detection and foreground and background data synchronization consistency verification, and in particular, relates to a transformer oil temperature foreground and background synchronization consistency verification system based on image recognition algorithm technology. Background Art

[0002] At present, the consistency check of transformer oil temperature between front and back ends mostly adopts thermal simulation measurement and manual calibration methods. The oil temperature and winding temperature are mainly measured by thermocouples, the measurement results are read manually, and then compared and calibrated with the back-end personnel through intercom.

[0003] Its typical device consists of: constant temperature oil tank, AC current (clamp) meter, AC current source, standard thermometer, etc. Constant temperature bath: provides a stable temperature environment and places the temperature measuring device sensor in it.

[0004] AC current (clamp) meter: measure transformer winding current AC current source: provides power to the device Standard thermometer: serves as an accurate temperature reference standard.

[0005] The typical calibration steps include: Put the sensor of the pointer oil thermometer and the standard thermometer into the constant temperature bath. Wait for the temperature of each table to be fully stable. After stabilization, read the value of the standard thermometer, the display value of the pointer oil thermometer and the background value at the same time. By comparing the differences in each value, the error of the oil thermometer can be determined. If the error exceeds the allowable range, the front and back of the oil thermometer need to be adjusted or corrected to make the measurement result more accurate.

[0006] In the existing scheme, the transformer oil temperature on-site pointer and background data display are far apart. During the detection process, it is necessary to monitor the meter head, PT100 and background digital display at the same time. In the thermal simulation measurement method, more equipment is needed, and multiple people are involved. Constant temperature oil tank, AC current (clamp) meter, AC current source, standard thermometer, etc. are used at the same time. The operation steps are relatively cumbersome and need to be compared repeatedly, which introduces more uncertain factors. Pointer oil thermometers of different types and accuracy requirements may require different calibration devices and more complex calibration processes. When performing calibration operations, it is necessary to strictly follow the relevant operating procedures and standards to ensure the accuracy and reliability of the calibration results. Therefore, there will be the following problems in the manual calibration process: ① The equipment is numerous and complex, requiring high manpower costs ② It takes more time to face load changes ③ Manual measurement cannot reflect the winding temperature and transformer oil temperature under various operating external conditions in real time, and there is a certain error, it cannot reflect the temperature of the transient process, and the measurement has a certain lag ④ It is impossible to predict the future oil temperature ⑤ In the event of an accident, the cause of the accident cannot be quickly located and optimal control can be made. Summary of the invention

[0007] In view of the defects and shortcomings of the above existing technologies, the present invention proposes a transformer oil temperature front-end and back-end synchronization consistency verification system based on image recognition algorithm technology. The system can automatically and accurately monitor and verify the transformer oil temperature, and realize the synchronization consistency verification of the front-end and back-end data. In addition, the system also has fault prediction and judgment functions. It not only helps to solve the problems existing in the traditional method, but also improves the accuracy and real-time performance of transformer oil temperature detection. It is expected to bring more efficient and accurate transformer oil temperature monitoring and calibration solutions to the power industry.

[0008] The present invention specifically adopts the following technical solutions: A transformer oil temperature front-end and back-end synchronization consistency verification system based on image recognition algorithm technology, comprising: Standard temperature acquisition module: uses oil temperature sensor to monitor transformer oil temperature in real time and transmits it to the backend server; Image acquisition and recognition module: The image acquisition module deployed at the transformer site is used to capture the oil temperature gauge in real time. The oil temperature gauge image is collected in real time and transmitted to the backend server. The collected image is recognized to obtain the collection result of the on-site oil temperature gauge reading. Front-end and back-end consistency verification module: the oil temperature gauge readings and oil temperature background data obtained by the image acquisition and recognition module at the same time are verified using the acquisition results of the standard temperature acquisition module as the verification benchmark, and the verification results are output.

[0009] Furthermore, in the standard temperature acquisition module, the oil temperature of the transformer is monitored in real time by an oil temperature sensor, and the data acquisition module receives the electrical signal of the oil temperature sensor for analog-to-digital conversion, and then uploads it to the background server through the communication module.

[0010] Furthermore, in the image acquisition and recognition module, the image acquisition module uses a high-resolution camera, and the acquired image is pre-processed and then recognized by a deep learning algorithm model to obtain the position and reading of the oil temperature gauge pointer, and further converted into readable oil temperature data.

[0011] Furthermore, the background server is also deployed with a prediction model for collecting the oil temperature information uploaded by each module and obtaining the prediction result of the oil temperature through training.

[0012] Furthermore, the operating oil temperature parameters of each transformer are uploaded to the cloud platform regularly or in real time to achieve data sharing and centralized control.

[0013] Furthermore, if the verification result of the foreground and background consistency verification modules shows that the data are consistent, indicating that the oil temperature monitoring system is working normally, the monitoring will continue; if the data are inconsistent, an alarm signal will be issued.

[0014] Furthermore, when the front-end and back-end consistency check module detects a consistency failure, the predicted oil temperature of the prediction model is checked for consistency with the calibrated oil temperature value, and an error threshold is set. If the error between the two numbers is greater than the error threshold, an alarm signal exceeding the threshold is further issued.

[0015] Furthermore, the prediction model predicts the transformer oil temperature for a period of time in the future based on historical data, operating parameters, environmental parameters and load conditions, and outputs prediction results including the transformer commissioning and maintenance time and the transformer remaining life.

[0016] Compared with the prior art, the present invention and its preferred solution realize automatic detection and verification of oil temperature through image recognition algorithm, which greatly reduces the burden of manual operation. Pre-prevention: by predicting the oil temperature of the transformer and the temperature of the winding, the load capacity of the transformer can be predicted in advance, which is beneficial to load planning. In-process control: real-time monitoring of the oil temperature of the transformer, through the learning function, powerful computing power and powerful algorithm function of the AI ​​big model, the cause of the abnormality can be determined and located, the optimal solution can be made, and then a quick response can be made. In summary, the system is equipped with a database and an AI big model, which can learn and store each event, which is in line with the current development trend of power grid big data.

[0017] High accuracy: The image recognition algorithm can accurately identify the oil temperature gauge reading and transmit it remotely, reducing on-site human reading errors and improving calibration accuracy.

[0018] Strong real-time performance: Real-time monitoring and synchronous calibration can realize real-time monitoring of transformer oil temperature, and perform synchronous calibration in time according to the monitoring results to ensure that the transformer oil temperature always remains within the accurate range, which helps to ensure the safe and stable operation of the transformer.

[0019] Strong adaptability: The support based on AI models has the ability of self-learning and optimization, and can adapt to changes in different application scenarios. The impact of environmental factors (such as light, temperature, humidity, etc.) during transformer operation and factors such as transformer models and operating parameter life, oil temperature gauge type, etc. on calibration can be accurately learned and judged by AI large models through data sharing. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments: Figure 1 This is a structural diagram of the image acquisition and verification part of the system according to an embodiment of the present invention; Figure 2 This is a framework diagram of the AI ​​big model system according to an embodiment of the present invention; Figure 3 This is a flow chart of the front-end and back-end consistency check of an embodiment of the present invention; Figure 4 This is a flow chart of fault determination in an embodiment of the present invention; Figure 5 This is a diagram of data sharing between transformers according to an embodiment of the present invention. DETAILED DESCRIPTION

[0021] Hereinafter, the specific embodiments of the present application will be described in detail with reference to the accompanying drawings. According to these detailed descriptions, those skilled in the art can clearly understand the present application and can implement the present application. Without violating the principles of the present application, the features in the various embodiments can be combined to obtain new implementations, or certain features in certain embodiments can be replaced to obtain other preferred implementations.

[0022] In order to make the features and advantages of the present invention more clearly understood, the following embodiments are specifically described in detail with reference to the accompanying drawings: In the design scheme provided in the embodiment of the present invention, intelligent oil temperature data verification and comparison are mainly achieved through multi-channel data remote transmission combined with algorithm identification and verification.

[0023] Based on the existing background collection of transformer oil temperature data, this embodiment completes the oil temperature data collection through two channels, one of which is through direct collection, consisting of an oil temperature sensor, a data collection module, a communication module, etc., wherein the oil temperature sensor monitors the transformer oil temperature in real time, and preferably a PT100 temperature sensor can be used. The data collection module receives the electrical signal for analog-to-digital conversion and converts it into a digital signal. The communication module is responsible for realizing data transmission between the data collection module and the background server.

[0024] Therefore, combined with the results of the background collection of the existing transformer oil temperature data, the verification module receives the front-end display data (such as the oil temperature data displayed by the monitoring system) and the actual background measurement data (data transmitted from the data acquisition module), and then compares and analyzes them to determine the consistency of the two, thereby realizing the first background verification solution. If the data is inconsistent, the verification module may issue an alarm signal, and the display module is used to display the verification results and oil temperature data and other information.

[0025] In such Figure 1 As shown, in the second data collection solution provided by the embodiment of the present invention, the image of the on-site meter is collected and transmitted remotely, and after image recognition processing, the consistency check is performed with the oil temperature data directly collected by the above oil temperature sensor, wherein: Image acquisition device: deployed at the transformer site, used to capture the image of the oil temperature gauge in real time and transmit the image back through the network. The device uses a high-resolution camera to ensure clear and accurate images.

[0026] Image processing module: responsible for preprocessing the collected images, including denoising, contrast enhancement, image correction and other operations to improve image quality and facilitate subsequent recognition and analysis.

[0027] Image recognition algorithm: Use deep learning algorithms, such as convolutional neural networks (CNN), to train and recognize preprocessed images, accurately identify the position and reading of the oil temperature gauge pointer, and convert it into readable oil temperature data.

[0028] Communication module: It carries out data transmission between various parts. The output transmission adopts the combination of wireless transmission and wired transmission to ensure the reliability and stability of transmission.

[0029] Calibration control module: compares and analyzes the received oil temperature data with the standard oil temperature value, determines whether calibration is required, and generates corresponding calibration instructions.

[0030] Mobile terminal and monitor: can realize real-time monitoring and convenient control of oil temperature.

[0031] In order to avoid the situation where both of the above two verification methods have abnormalities and failures, the present invention also provides oil temperature prediction based on the AI ​​large model based on the above data collection, such as Figure 2 As shown, the specific design of its preferred solution includes: Data collection layer: This layer is responsible for raw data from various sources; Data preprocessing layer: In this layer, the raw data is cleaned, enhanced, and formatted to facilitate subsequent feature extraction and model training; Feature Engineering Layer: In this layer, the system extracts and selects useful features from the preprocessed data, which will be used to train the AI ​​model; Model training layer: In this layer, the system selects a suitable model for training; Prediction layer: predict and execute the results; Post-processing and optimization layer: evaluate the results and optimize the model; User interface layer: human-computer interface and data storage, which connects to the database: archives and summarizes historical data.

[0032] like Figure 5 As shown, as a preferred embodiment, the operating oil temperature parameters of each transformer are uploaded to the cloud platform regularly or in real time to achieve data sharing and centralized control.

[0033] Installation and connection: Install the image acquisition module opposite to the oil temperature gauge and connect each module to the network.

[0034] Establish connections between the communication module and the backend server and verification module to ensure smooth data transmission.

[0035] Device initialization: Turn on the device power, perform initialization settings, and ensure that each module is working properly.

[0036] Check whether the communication connection is normal to ensure that the data acquisition module can effectively exchange data with the background server.

[0037] Data collection: The oil temperature sensor monitors the transformer oil temperature in real time, and the image sensor converts the oil temperature data into electrical signals and transmits them to the data acquisition module.

[0038] The data acquisition module performs analog-to-digital conversion on the electrical signal and converts the oil temperature data into a digital signal.

[0039] Data transmission: The data acquisition module transmits the converted digital signal (i.e. the actual measurement data in the background) to the background server through the communication module.

[0040] The backend server receives and stores the data, and sends the oil temperature data to the frontend for display (i.e., the frontend displays the data) according to a preset communication protocol.

[0041] Consistency check: The verification module receives both the foreground display data and the background actual measurement data.

[0042] Using specific algorithms and comparison rules, the two sets of data are compared and analyzed to determine whether their values ​​at the same time are consistent and whether the difference is within the allowable error range.

[0043] like Figure 3 As shown in the figure, during the consistency check between the front and back ends, the oil temperature meter and the back end data are checked with the standard thermometer as the check reference to ensure the consistency of the check. First, the data are collected and processed, and the processed oil temperature data is sent to the consistency check module for consistency check. If the consistency check fails, the corresponding check procedure will be executed to operate each check actuator.

[0044] Result processing: If the verification results show that the data are consistent, it means that the oil temperature monitoring system is working properly and monitoring can continue.

[0045] If the data is inconsistent, the verification module will send out an alarm signal to inform the staff that there is a problem. At the same time, the display module will display the verification results and specific oil temperature data so that the staff can locate the problem in time and take corresponding measures, such as checking whether the sensor is faulty, whether the data transmission line is abnormal, and whether the background data processing is correct.

[0046] Records and Maintenance: Record the results of each calibration, including calibration time, oil temperature data, consistency judgment results, etc., for subsequent review and analysis.

[0047] Regularly maintain and inspect the device to ensure the stable performance of its components, such as checking the accuracy of the sensor, cleaning the sensor surface, checking the connection of the communication line, etc.

[0048] Workflow: 1. Image acquisition: The image acquisition device takes images of the transformer oil temperature gauge at regular intervals or in real time.

[0049] 2. Image processing and recognition: The image processing module processes the image, and then the image recognition algorithm analyzes the processed image to obtain the current oil temperature reading.

[0050] 3. Calibration judgment and execution: The calibration control module compares the received oil temperature data with the standard oil temperature value. If there is a deviation and it exceeds the allowable range, a calibration instruction is generated and sent to the corresponding actuator (such as the control device for adjusting the transformer cooling system, etc.) to achieve synchronous calibration of the transformer oil temperature. The calibrated oil temperature is compared with the predicted oil temperature. If there is a deviation and it exceeds the allowable range, a fault judgment is made, the cause of the fault is analyzed, and an alarm signal is issued.

[0051] like Figure 4 As shown, in fault judgment, the predicted oil temperature and the calibrated oil temperature value are checked for consistency, and an error threshold is set in the consistency check. If the error between the two numbers is greater than the error threshold, the error judgment program is started to judge the cause of the error and send out an alarm signal.

[0052] 4. As a further preferred solution, the AI ​​big model and database equipped on the server can predict the transformer oil temperature in the future based on historical data, operating parameters and load forecasts, and provide data such as the transformer commissioning and maintenance time and the remaining life of the transformer in advance to ensure the safe and efficient operation of the equipment and realize more intelligent power system control.

[0053] Based on the above design, this embodiment further provides the following preferred design: Through image acquisition and recognition, and then synchronous recognition and calibration through the network, real-time and calibration accuracy are guaranteed, and manpower is saved.

[0054] Each module is connected through a local area network, and each regional system is connected through a wide area network. The system is equipped with camera modules, image processing modules, AI large models, databases, etc.

[0055] Furthermore, all transformers in various regions and areas will be connected to the system, and all operating parameters, status and other parameters will be uploaded to the database to provide sufficient data for the AI ​​large model, thereby achieving more accurate training results.

[0056] Through the AI ​​big model, we can achieve pre-prevention, in-process control and post-event summary. It can provide a certain basis for load planning in advance, accurately locate the cause of the fault in the process, propose the best solution, and archive and summarize it afterwards.

[0057] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0058] The above is only a preferred embodiment of the present invention, and does not limit the present invention in other forms. Any technician familiar with the profession may use the above disclosed technical content to change or modify it into an equivalent embodiment with equivalent changes. However, any simple modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention without departing from the technical solution of the present invention still belongs to the protection scope of the technical solution of the present invention.

[0059] The present invention is not limited to the above-mentioned optimal implementation mode. Anyone can derive other various forms of transformer oil temperature front-end and back-end synchronization consistency verification systems based on image recognition algorithm technology under the inspiration of the present invention. All equal changes and modifications made according to the scope of the patent application of the present invention should fall within the scope of the present invention.

Claims

1. A transformer oil temperature front-end and back-end synchronization consistency verification system based on image recognition algorithm technology, characterized in that: include: Standard temperature acquisition module: uses oil temperature sensor to monitor transformer oil temperature in real time and transmits it to the backend server; Image acquisition and recognition module: The image acquisition module deployed at the transformer site is used to capture the oil temperature gauge in real time. The oil temperature gauge image is collected in real time and transmitted to the backend server. The collected image is recognized to obtain the collection result of the on-site oil temperature gauge reading. Front-end and back-end consistency verification module: the oil temperature gauge readings and oil temperature background data obtained by the image acquisition and recognition module at the same time are verified using the acquisition results of the standard temperature acquisition module as the verification benchmark, and the verification results are output.

2. According to claim 1, a transformer oil temperature front-end and back-end synchronization consistency verification system based on image recognition algorithm technology is characterized in that: In the standard temperature acquisition module, the oil temperature of the transformer is monitored in real time through the oil temperature sensor. The data acquisition module receives the electrical signal of the oil temperature sensor for analog-to-digital conversion and then uploads it to the background server through the communication module.

3. According to claim 1, a transformer oil temperature front-end and back-end synchronization consistency verification system based on image recognition algorithm technology is characterized in that: In the image acquisition and recognition module, the image acquisition module uses a high-resolution camera. After preprocessing, the acquired image is recognized by a deep learning algorithm model to obtain the position and reading of the oil temperature gauge pointer, and further converted into readable oil temperature data.

4. According to claim 1, a transformer oil temperature front-end and back-end synchronization consistency verification system based on image recognition algorithm technology is characterized in that: The background server is also deployed with a prediction model for collecting the oil temperature information uploaded by each module and obtaining the prediction result of the oil temperature through training.

5. According to the transformer oil temperature front-end and back-end synchronization consistency verification system based on image recognition algorithm technology according to claim 1, it is characterized by: The operating oil temperature parameters of each transformer are uploaded to the cloud platform regularly or in real time to achieve data sharing and centralized control.

6. According to claim 1, a transformer oil temperature front-end and back-end synchronization consistency verification system based on image recognition algorithm technology is characterized in that: If the verification result of the foreground and background consistency verification modules shows that the data are consistent, it means that the oil temperature monitoring system is working normally, and the monitoring will continue; if the data are inconsistent, an alarm signal will be issued.

7. The transformer oil temperature front-end and back-end synchronization consistency verification system based on image recognition algorithm technology according to claim 4 is characterized by: When the front-end and back-end consistency check module detects a consistency failure, the predicted oil temperature of the prediction model is checked for consistency with the calibrated oil temperature value, and an error threshold is set. If the error between the two numbers is greater than the error threshold, an alarm signal exceeding the threshold is further issued.

8. According to claim 4, a transformer oil temperature front-end and back-end synchronization consistency verification system based on image recognition algorithm technology is characterized in that: The prediction model predicts the transformer oil temperature in the future based on historical data, operating parameters, environmental parameters and load conditions, and outputs prediction results including the transformer commissioning and maintenance time and the transformer remaining life.

Images