A monitoring method, device and electronic equipment of a power transformer

Abstract

The application provides a monitoring method and device of a power transformer and electronic equipment. It relates to the technical field of transformers. The monitoring method comprises the following steps: acquiring all state parameters of the power transformer during operation, and screening a plurality of state input parameters from the all state parameters; inputting the plurality of state input parameters into a preset first neural network monitoring model to obtain monitoring parameters corresponding to a plurality of components of the power transformer respectively. The application can monitor the state of the power transformer, improve the state monitoring accuracy of the power transformer, and reduce the monitoring cost.

Description

Technical Field

[0001] This invention relates to the field of transformer technology, and more specifically, to a monitoring method, device, and electronic equipment for power transformers. Background Technology

[0002] Power transformers are among the main pieces of equipment in power plants and substations. In power systems, power transformers play a crucial role in voltage transformation, and their safety and service life are vital guarantees for the safe, reliable, and economical operation of the power system, as well as the power supply and consumption of the entire grid, and are essential for the stable operation of the entire power system. Therefore, it is necessary to monitor the operating status of each component within the power transformer. Typically, sensors are used to monitor the parameters of each component online, and then the transformer's fault status is determined based on these parameters. However, the status of each component within a power transformer is difficult to obtain directly from sensors, leading to problems such as monitoring difficulties and high false alarm rates in existing monitoring methods. Summary of the Invention

[0003] To address the aforementioned problems, the present invention aims to provide a method, apparatus, and electronic device for monitoring power transformers.

[0004] In a first aspect, embodiments of the present invention provide a method for monitoring a power transformer, comprising:

[0005] Obtain all state parameters of the power transformer during operation, and filter out multiple state input parameters from all state parameters;

[0006] The multiple state input parameters are input into a preset first neural network monitoring model to obtain the monitoring parameters corresponding to each of the multiple components on the power transformer.

[0007] The first neural network monitoring model is obtained by iteratively training a preset first BP neural network based on multiple state input parameters and monitoring parameters corresponding to multiple components.

[0008] Secondly, embodiments of the present invention also provide a monitoring device for a power transformer, comprising:

[0009] The status data unit is used to acquire all status parameters of the power transformer during operation and to filter out multiple status input parameters from all the status parameters.

[0010] The model monitoring unit is used to input multiple state input parameters into a preset first neural network monitoring model to obtain the monitoring parameters corresponding to each of the multiple components on the power transformer.

[0011] The first neural network monitoring model is obtained by iteratively training a preset first BP neural network based on multiple state input parameters and monitoring parameters corresponding to multiple components.

[0012] Thirdly, embodiments of the present invention also provide an electronic device, including a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the method described in the first aspect is performed.

[0013] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed by a computer, performs the method described in the first aspect.

[0014] In the solutions provided in the first to fourth aspects of the present invention, the monitoring parameters corresponding to each of the multiple components of the power system are monitored by a pre-trained first neural network monitoring model, thereby realizing the monitoring of the state of the power transformer. Compared with the method of using sensors to monitor the parameters of each component in related technologies, the accuracy of power transformer state monitoring can be improved and the monitoring cost can be reduced.

[0015] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 The diagram shows a first flowchart of a power transformer monitoring method provided in Embodiment 1 of the present invention.

[0018] Figure 2 The diagram shows a second flowchart of a power transformer monitoring method provided in Embodiment 1 of the present invention.

[0019] Figure 3 The diagram shows a third process flow of a power transformer monitoring method provided in Embodiment 1 of the present invention.

[0020] Figure 4 The diagram shows a first structural schematic of a power transformer monitoring device provided in Embodiment 2 of the present invention.

[0021] Figure 5 The diagram shows a second structural schematic of a power transformer monitoring device provided in Embodiment 2 of the present invention.

[0022] Figure 6 The diagram shows a third structural schematic of a power transformer monitoring device provided in Embodiment 2 of the present invention.

[0023] Figure 7 A schematic diagram of the structure of an electronic device provided in Embodiment 3 of the present invention is shown. Detailed Implementation

[0024] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0025] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise expressly specified. Moreover, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0026] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0027] There are many causes of power transformer failures, mainly categorized as overheating, electrical stress, localized bushing problems, main bushing issues, and on-load tap changer failures. Throughout the entire service life of a power transformer, it is constantly affected by multiple factors, including thermal stress, electrical stress, mechanical stress, and chemical stress. The failure rates of power transformers, from highest to lowest, are: winding failures, tap changer failures, bushing failures, cooling system failures, tank and accessory failures, and core failures. Winding failures, tap changer failures, and bushing failures are the three main types of power transformer failures, accounting for approximately 70% of all power transformer failures.

[0028] To minimize the occurrence of power transformer faults or accidents, condition monitoring of power transformers is necessary. Current technologies using sensors for condition monitoring still face significant challenges in comprehensively monitoring the status of all components of the power transformer. Furthermore, reliability assessments based on the power transformer's turntable and fault early warning and health management are crucial for preventing catastrophic power transformer failures.

[0029] Based on this, the present invention proposes a power transformer monitoring method, device, electronic device, and readable storage medium, which monitors the monitoring parameters corresponding to multiple components of the power system through a pre-trained first neural network monitoring model, thereby realizing the monitoring of the power transformer's status. Compared with the related technologies that use sensors to monitor the parameters of each component, this method can improve the accuracy of power transformer status monitoring and reduce monitoring costs.

[0030] Before providing a more detailed description of the embodiments of the present invention, the first neural network monitoring model and the second neural network monitoring model disclosed in the embodiments of the present invention will be described in detail.

[0031] The first neural network monitoring model disclosed in this embodiment of the invention is obtained by iteratively training a preset first BP neural network based on multiple state input parameters and monitoring parameters corresponding to multiple components, specifically including the following steps:

[0032] 1. Construct the input parameter library and output parameter library for the neural network.

[0033] The input parameter library in this step is a database consisting of multiple state input parameters for different operating stages of the power transformer (referring to each stage of the power transformer's entire life cycle, with each stage having an operating time length set according to data collection requirements), and the output parameter library is a database consisting of multiple monitoring parameters for different operating stages of the power transformer.

[0034] 2. The constructed input parameter library and output parameter library are randomly divided into training set, validation set and test set, accounting for 70%, 15% and 15% of the total data, respectively.

[0035] 3. Using the training set, train the BP neural network, with 30 hidden layers. If the error of the neural network monitoring model is less than 1% for 10 consecutive training iterations, the training is considered complete, and the required neural network monitoring model is obtained.

[0036] 4. Add the validation set to the obtained neural network monitoring model for validation. If the error of the neural network monitoring model is less than 2%, the neural network monitoring model is confirmed to meet the requirements.

[0037] 5. By testing the test set data using a neural network monitoring model, the predictive performance of the neural network monitoring model can be obtained.

[0038] The second neural network monitoring model disclosed in this embodiment of the invention is obtained by iteratively training a preset second BP neural network based on the monitoring parameters corresponding to multiple components and the instantaneous overload parameters during the operation of the power transformer. Specifically, it includes the following steps:

[0039] 1. Construct the input parameter library and output parameter library for the neural network.

[0040] The input parameter library in this step is a database consisting of multiple monitoring parameters for different operating stages of the power transformer, and the output parameter library is a database consisting of instantaneous overload parameters for different operating stages of the power transformer.

[0041] 2. The constructed input parameter library and output parameter library are randomly divided into training set, validation set and test set, accounting for 70%, 15% and 15% of the total data, respectively.

[0042] 3. Using the training set, train the BP neural network, with 30 hidden layers. If the error of the neural network monitoring model is less than 1% for 10 consecutive training iterations, the training is considered complete, and the required neural network monitoring model is obtained.

[0043] 4. Add the validation set to the obtained neural network monitoring model for validation. If the error of the neural network monitoring model is less than 2%, the neural network monitoring model is confirmed to meet the requirements.

[0044] 5. By testing the test set data using a neural network monitoring model, the predictive performance of the neural network monitoring model can be obtained.

[0045] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0046] Example 1

[0047] A schematic diagram of a power transformer monitoring method provided in an embodiment of the present invention is shown below. Figure 1 As shown, the monitoring method for this power transformer includes the following steps:

[0048] S101: Obtain all state parameters of the power transformer during operation, and filter out multiple state input parameters from all state parameters.

[0049] It is understandable that all the state parameters of a power transformer during operation include the electrical and state parameters of various parts of the power transformer.

[0050] All status parameters in this step include: oil chromatography, core grounding current, line current, transformer voltage, waveform data, core data, winding data, top oil temperature, middle oil temperature, bottom oil temperature, bushing data, casing temperature, ambient temperature, weather humidity, oil level data, noise data, and vibration data.

[0051] In this step, multiple state parameters are selected from all state parameters as multiple state input parameters for the preset first neural network monitoring model. Specifically, these are selected from all state parameters based on the target component on the power transformer to be monitored and its corresponding electrical parameters. For example, the target component can be the core, bushing, or shell; the corresponding electrical parameters are core temperature or core grounding current, bushing temperature, and shell temperature or shell leakage current, respectively. It should be noted that the electrical parameters corresponding to the target component are the monitoring parameters in this embodiment.

[0052] The methods for filtering multiple state input parameters from all state parameters in this step include the following:

[0053] Step 1: Select one state parameter from all state parameters. Changes corresponding to multiple modifications to this status parameter numerical values Each time the status parameter is changed... change After obtaining the numerical values, the corresponding values ​​for multiple monitoring parameters are detected separately.

[0054] The following example illustrates step 1 above. All state parameters in the example include oil chromatography, line current, and transformer voltage. Monitored parameters include: transformer internal temperature, voltage ratio, and load current. The measured value corresponding to the transformer internal temperature is the temperature value; the measured value corresponding to the voltage ratio is the ratio; and the measured value corresponding to the load current is the current value.

[0055] First, extract one state parameter from all state parameters as the line current, and then change the corresponding current value of the line current (this current value is the value changed in step 1). After changing the current value of the line current, it is necessary to detect the temperature value corresponding to the internal temperature of the transformer, the voltage ratio, and the load current value respectively.

[0056] Therefore, after each change of the line current value, which is a status parameter, it is necessary to check the temperature value corresponding to the internal temperature of the transformer, the voltage ratio, and the load current value.

[0057] Step 2: Perform linear fitting between the multiple changes in the state parameter and the multiple detection values ​​of a monitoring parameter to obtain the fitted line and the slope of the fitted line.

[0058] Based on the above example, we can obtain the current values ​​corresponding to multiple line currents, as well as the temperature values ​​corresponding to the transformer's internal temperature detected after each change. A coordinate system is formed by using each current value as the x-axis and the corresponding temperature value as the y-axis; the current value and temperature value constitute a coordinate point on this system. After multiple changes, multiple coordinate points are constructed on the coordinate system. Linear fitting of these multiple coordinate points yields a fitted straight line. The slope of this fitted line can be calculated using the coordinate points on the fitted line; this slope is the fitting slope.

[0059] Understandably, by performing the fitting process in step 2, we can obtain the fitted straight lines of the line current with respect to the transformer's internal temperature, voltage ratio, and load current, as well as the fitting slope of each straight line. This allows us to obtain multiple fitting slopes between the state parameter and various monitoring parameters.

[0060] By replacing the line current with the transformer voltage and repeating steps 1 and 2, we can obtain multiple fitting slopes between each state parameter and each monitoring parameter among all the monitoring parameters.

[0061] In other words, by using the above method, the correlation coefficient between each state parameter and the monitoring parameter corresponding to each component can be determined; where the correlation coefficient is the fitting slope between the state parameter and the monitoring parameter.

[0062] When filtering all state parameters, the filtering is performed based on the correlation coefficients corresponding to each state parameter, resulting in multiple state input parameters. Specifically, among all state parameters, those whose correlation coefficients are all greater than a preset correlation coefficient threshold are identified as state input parameters. In other words, if multiple fitting slopes for each state parameter are greater than the preset correlation coefficient threshold, then that state parameter is considered as a single state input parameter.

[0063] This comparison method can yield multiple state parameters that meet the comparison requirements (i.e., the fitting slopes of multiple state parameters are all greater than the preset correlation coefficient threshold). These multiple state parameters that meet the comparison requirements are the multiple state input parameters.

[0064] In one implementation, the aforementioned preset correlation coefficient threshold can be set to 0.15.

[0065] In addition, this embodiment of the invention also provides another method for filtering multiple state input parameters from all state parameters. Unlike the above method, if any one of the multiple fitting slopes corresponding to a state parameter is greater than a preset correlation coefficient threshold, then that state parameter can be determined as a state input parameter. This method can also be used to filter multiple state input parameters from all state parameters.

[0066] S102: Input the multiple state input parameters into the preset first neural network monitoring model to obtain the monitoring parameters corresponding to each of the multiple components on the power transformer.

[0067] In this step, the first neural network monitoring model is divided into three parts: the input layer, the hidden layer, and the output layer.

[0068] The input layer receives multiple state input parameters obtained from the filtering in step S101.

[0069] Hidden layers are the complex internal structure of a neural network. In this step, we set there to 30 hidden layers, with connections between them. Through learning based on the training dataset, the mapping relationship between the input and output layers is achieved. The hidden layers do not directly output the monitoring parameters of the power transformer to the outside world, nor do they directly receive the transformer's state input parameters.

[0070] The output layer outputs monitoring parameters of components on the power transformer. These monitoring parameters include the transformer's internal temperature, internal humidity, leakage current between the transformer's internal windings and casing, leakage current between the transformer casing and ground, voltage ratio, load current, bushing leakage current, eddy current in the core, fan speed, circulating oil flow rate, and arc length during opening and closing.

[0071] It should be noted that the input layer information is closely related to the state input parameters of the power transformer. By analyzing and processing the state input parameters, the state information of the power transformer can be effectively extracted. Since there is multi-parameter coupling and a strong nonlinear relationship between the state information and the state input parameters of the power transformer, a neural network is used for mapping and construction.

[0072] To implement health management for power transformers, thereby reducing manual inspections and lowering maintenance costs. See also Figure 2 Following step S102 in the above embodiment, the following specific content is included:

[0073] S201: Based on the first neural network monitoring model, obtain the winding temperature of the upper winding of the power transformer and the leakage current of the casing.

[0074] In this step, the transformer internal temperature output by the first neural network monitoring model is determined to be the winding temperature of the upper winding of the power transformer, and the leakage current between the transformer internal winding and the casing output by the first neural network monitoring model is determined to be the leakage current of the casing.

[0075] S202: Collect the cooling oil temperature of the power transformer during operation.

[0076] In this step, a temperature sensor can be built into the cooling oil of the power transformer, and the temperature of the cooling oil measured by the temperature sensor can be transmitted to the outside of the power transformer wirelessly.

[0077] S203: Adjust the fan speed of the power transformer and the flow rate of the cooling oil inside the power transformer based on the winding temperature and the cooling oil temperature.

[0078] This step includes two parts: adjusting the fan speed and adjusting the cooling oil flow rate. The condition for starting and adjusting the fan speed and cooling oil flow rate is that both the winding temperature and the cooling oil temperature reach 80% of the rated temperature of the power transformer.

[0079] In other words, when both the winding temperature and the cooling oil temperature reach 80% of the rated temperature of the power transformer, the fan speed and the cooling oil flow rate should be adjusted. Specifically, this can be done by increasing either the fan speed or the cooling oil flow rate. If this adjustment method fails to reduce the winding temperature or the cooling oil temperature, then it is necessary to increase both the fan speed and the cooling oil flow rate.

[0080] By adjusting the fan speed and cooling oil flow rate of power transformers, the operational reliability of power transformers can be improved, the equipment life of power transformers can be extended, and economic benefits can be enhanced.

[0081] S204: Adjust the operating voltage of the power transformer based on the leakage current.

[0082] This step involves adjusting the operating voltage of the power transformer based on changes in the leakage current. Specifically, the larger the leakage current, the lower the operating voltage should be. In other words, once the leakage current increases or decreases, the operating voltage of the power transformer needs to be lowered or increased.

[0083] Alternatively, a predefined correspondence table between leakage current variation thresholds and corresponding operating voltage variation ranges can be used. Once the leakage current is determined, its corresponding variation threshold can be obtained. The voltage variation range corresponding to this variation threshold can then be determined according to the correspondence table. The operating voltage of the power transformer can then be adjusted to fall within this voltage variation range.

[0084] For example, the permissible variation range of leakage current is 1μA-30μA. This range can be divided into three threshold values, each in 10μA increments: 1μA-10μA, 11μA-20μA, and 21μA-30μA. The corresponding voltage ranges for these three threshold values ​​are 1000kV-800kV, 799kV-600kV, and 599kV-400kV, respectively.

[0085] If the leakage current is determined to be within the variation threshold of 21μA-30μA, then the operating voltage of the power transformer should be within the variation range of 1000kV-800kV. If the leakage current is determined to be within the variation threshold of 1μA-10μA, then the operating voltage of the power transformer should be within the variation range of 599kV-400kV.

[0086] By adjusting the operating voltage of power transformers, insulation flashover can be avoided, thereby improving the operational reliability of power transformers.

[0087] As can be seen from the above description, the health management of power transformers refers to the active adjustment of the fan speed, cooling oil flow rate, and operating voltage of power transformers.

[0088] In conclusion, by implementing health management for power transformers, their reliability can be improved, thereby ensuring the stable operation of the power system.

[0089] In order to conduct inspections and maintenance of power transformers, and to achieve predictive maintenance and reduce maintenance costs. See also Figure 3 Following step S102 in the above embodiment, the following specific content is included:

[0090] S301: Input the monitoring parameters corresponding to each of the multiple components on the power transformer into the preset second neural network monitoring model to obtain the instantaneous overload parameters of the power transformer during operation;

[0091] In this step, the instantaneous overload parameters of the power transformer are evaluated by monitoring parameters, which characterize the instantaneous overload capacity of the power transformer. Through the above step S102, the monitoring parameters corresponding to each of the multiple components on the power transformer can be obtained.

[0092] By inputting the obtained monitoring parameters into the preset second neural network monitoring model, the instantaneous overload parameters of the power transformer during operation can be obtained.

[0093] S302: Obtain the rated instantaneous overload parameters of the power transformer at the time of manufacture.

[0094] In this step, the rated instantaneous overload parameters of the power transformer are pre-cached.

[0095] It is understandable that the rated instantaneous overload parameters of a power transformer will have a rated value when it leaves the factory. As the power transformer ages, the instantaneous overload parameters will gradually decrease. This is mainly related to the aging of the internal insulation of the power transformer, the faults and deformation of the transformer windings, the aging of the bushings, the aging of the iron core, and the condition of the tap changer.

[0096] S303: Determine the attenuation ratio of the power transformer based on the instantaneous overload parameters and the rated instantaneous overload parameters.

[0097] In this step, the attenuation ratio is the ratio of the instantaneous overload parameter to the rated instantaneous overload parameter.

[0098] S304: When the attenuation ratio is less than the preset attenuation value, a warning signal is issued.

[0099] In this step, if the attenuation rate is less than 50%, the instantaneous overload parameters of the power transformer during operation will continue to be monitored. Simultaneously, health management of the power transformer will be performed. If the attenuation rate exceeds 50%, a warning signal will be issued, notifying the transformer to conduct inspection and maintenance.

[0100] As described above, early warning and management of power transformers can effectively prevent transformer failures. This can improve transformer operating efficiency, enable predictive maintenance, and quickly identify potential hazards.

[0101] In summary, the power transformer monitoring method proposed in this embodiment monitors the monitoring parameters corresponding to multiple components of the power system through a pre-trained first neural network monitoring model, thereby achieving the monitoring of the power transformer's status. Compared with related technologies that use sensors to monitor the parameters of each component, this method can improve the accuracy of power transformer status monitoring and reduce monitoring costs.

[0102] Example 2

[0103] This embodiment discloses a monitoring device for power transformers, such as... Figure 4 As shown, the monitoring device for the power transformer includes:

[0104] The status data unit 10 is used to acquire all status parameters of the power transformer during operation and to filter out multiple status input parameters from all the status parameters.

[0105] Model monitoring unit 20 is used to input multiple state input parameters into a preset first neural network monitoring model to obtain monitoring parameters corresponding to multiple components on the power transformer.

[0106] The first neural network monitoring model is obtained by iteratively training a preset first BP neural network based on multiple state input parameters and monitoring parameters corresponding to multiple components.

[0107] In an optional embodiment, the state data unit 10 includes:

[0108] The correlation coefficient module is used to determine the correlation coefficient between each state parameter and the monitoring parameter corresponding to each component;

[0109] The filtering module is used to filter all state parameters based on all correlation coefficients corresponding to each state parameter to obtain multiple state input parameters.

[0110] In an optional embodiment, the filtering module includes:

[0111] The judgment submodule is used to determine the state parameters that meet the condition that all correlation coefficients corresponding to the state parameter are greater than the preset correlation coefficient threshold among all state parameters.

[0112] See Figure 5 Based on the above embodiments, the monitoring device for the power transformer further includes:

[0113] The monitoring unit 30 is used to obtain the winding temperature of the upper winding of the power transformer and the leakage current of the casing based on the first neural network monitoring model; it is also used to collect the cooling oil temperature of the power transformer during operation.

[0114] The regulating unit 40 is used to regulate the fan speed of the power transformer and the flow rate of the cooling oil inside the power transformer based on the winding temperature and the cooling oil temperature; it is also used to regulate the operating voltage of the power transformer based on the leakage current.

[0115] See Figure 6 Based on the above embodiments, the monitoring device for the power transformer further includes:

[0116] The instantaneous overload unit 50 is used to input the monitoring parameters corresponding to each of the multiple components on the power transformer into the preset second neural network monitoring model to obtain the instantaneous overload parameters during the operation of the power transformer.

[0117] The second neural network monitoring model is obtained by iteratively training a preset second BP neural network based on the monitoring parameters corresponding to multiple components and the instantaneous overload parameters during the operation of the power transformer.

[0118] The early warning unit 60 is used to acquire the rated instantaneous overload parameters of the power transformer at the time of manufacture, determine the attenuation ratio of the power transformer based on the instantaneous overload parameters and the rated instantaneous overload parameters, and issue an early warning signal when the attenuation ratio is less than a preset attenuation value.

[0119] In these embodiments of the invention, each module or unit is used to execute the manner disclosed in Embodiment 1 above. Their functions will not be repeated here, but can be found in the detailed description of Embodiment 1 above.

[0120] In summary, the power transformer monitoring device proposed in this embodiment monitors the monitoring parameters corresponding to multiple components of the power system through a pre-trained first neural network monitoring model, thereby achieving monitoring of the power transformer's status. Compared with related technologies that use sensors to monitor the parameters of each component, this method can improve the accuracy of power transformer status monitoring and reduce monitoring costs.

[0121] Example 3

[0122] This invention discloses an electronic device, including a processor and a memory. The memory stores computer-readable instructions. When the processor executes the computer-readable instructions, it performs the monitoring of a power transformer as described in Embodiment 1 above.

[0123] The steps of the method are described below. For a detailed implementation, please refer to Method Example 1, which will not be repeated here.

[0124] Example 4

[0125] This invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the power transformer monitoring described in Embodiment 1 above.

[0126] The steps of the method are described below. For a detailed implementation, please refer to Method Example 1, which will not be repeated here.

[0127] Furthermore, this embodiment of the invention also discloses the specific structure of the electronic device in embodiment 3 above, see [link to documentation]. Figure 7 The diagram shows the structure of an electronic device, which includes a bus 51, a processor 52, a transceiver 53, a bus interface 54, a memory 55, and a user interface 56. The electronic device includes a memory 55.

[0128] In this embodiment, the electronic device further includes: one or more programs stored in the memory 55 and executable on the processor 52, configured to be executed by the processor to perform the one or more programs for the following steps:

[0129] Obtain all state parameters of the power transformer during operation, and filter out multiple state input parameters from all state parameters;

[0130] The multiple state input parameters are input into a preset first neural network monitoring model to obtain the monitoring parameters corresponding to each of the multiple components on the power transformer.

[0131] The first neural network monitoring model is obtained by iteratively training a preset first BP neural network based on multiple state input parameters and monitoring parameters corresponding to multiple components.

[0132] Transceiver 53 is used to receive and send data under the control of processor 52.

[0133] The bus architecture (represented by bus 51) can include any number of interconnected buses and bridges, linking various circuits including one or more processors represented by processor 52 and memory represented by memory 55. Bus 51 can also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be further described in this embodiment. Bus interface 54 provides an interface between bus 51 and transceiver 53. Transceiver 53 can be a single element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. For example, transceiver 53 receives external data from other devices. Transceiver 53 is used to transmit data processed by processor 52 to other devices. Depending on the nature of the computing system, a user interface 56 may also be provided, such as a keypad, display, speaker, microphone, or joystick.

[0134] Processor 52 is responsible for managing bus 51 and general processing, such as running a general-purpose operating system as described above. Memory 55 can be used to store data used by processor 52 during operation.

[0135] Optionally, the processor 52 may be, but is not limited to, a central processing unit, a microcontroller, a microprocessor, or a programmable logic device.

[0136] It is understood that the memory 55 in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DRRAM). The memory 55 of the systems and methods described in this embodiment is intended to include, but is not limited to, these and any other suitable types of memory.

[0137] In some implementations, memory 55 stores elements such as executable modules or data structures, or subsets thereof, or extended sets thereof: operating system 551 and application programs 552.

[0138] The operating system 551 includes various system programs, such as the framework layer, core library layer, and driver layer, used to implement various basic business functions and handle hardware-based tasks. The application program 552 includes various applications, such as a media player and a browser, used to implement various application functions. The program implementing the method of this embodiment can be included in the application program 552.

[0139] While this invention provides the method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-inventive labor. The order of steps listed in the embodiments is merely one possible execution order among many and does not represent the only possible execution order. In actual device or client product execution, the methods shown in the embodiments or drawings can be executed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment).

[0140] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention is not limited to any single aspect, nor to any single embodiment, nor to any combination and / or substitution of these aspects and / or embodiments. Furthermore, each aspect and / or embodiment of the present invention can be used alone or in combination with one or more other aspects and / or embodiments.

[0141] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, but the protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A monitoring method of a power transformer, characterized by, The method comprises the following steps: acquiring all state parameters of the power transformer during operation, and screening a plurality of state input parameters from the all state parameters; inputting the plurality of state input parameters into a preset first neural network monitoring model to obtain monitoring parameters corresponding to each of a plurality of components on the power transformer; wherein the first neural network monitoring model is obtained by iteratively training a preset first BP neural network according to the plurality of state input parameters and the monitoring parameters corresponding to each of the plurality of components, the all state parameters include oil chromatogram, line current and transformer voltage, and the monitoring parameters include transformer internal temperature, voltage transformation ratio and load current; screening the plurality of state input parameters from the all state parameters comprises the following steps: selecting one state parameter from the all state parameters, changing the state parameter corresponding to the change value multiple times, and detecting the detection value of each monitoring parameter after changing the state parameter corresponding to the change value each time; linearly fitting the change value of the state parameter multiple times with the detection value of one monitoring parameter to obtain a plurality of fitting slopes between the state parameter and the plurality of monitoring parameters; repeating the selection of one state parameter from the all state parameters and the subsequent steps for each state parameter in the all state parameters to obtain a plurality of fitting slopes between each state parameter in the all state parameters and each monitoring parameter in the plurality of monitoring parameters; wherein the fitting slope is a correlation coefficient between the state parameter and the monitoring parameter; in the all state parameters, the state parameter corresponding to all correlation coefficients greater than a preset correlation coefficient threshold is determined as a state input parameter, and a plurality of state input parameters are obtained; or, as long as one fitting slope in the plurality of fitting slopes corresponding to one state parameter is greater than the preset correlation coefficient threshold, the state parameter is determined as a state input parameter; obtaining the winding temperature of the winding and the leakage current of the casing of the power transformer based on the first neural network monitoring model; collecting the cooling oil temperature of the cooling oil of the power transformer during operation; adjusting the wind speed of the fan of the power transformer and the flow rate of the cooling oil in the power transformer based on the winding temperature and the cooling oil temperature; adjusting the operating voltage of the power transformer based on the leakage current.

2. The method of claim 1, wherein, The method further comprises the following steps: inputting the monitoring parameters corresponding to each of the plurality of components on the power transformer into a preset second neural network monitoring model to obtain an instantaneous overload parameter during operation of the power transformer; wherein the second neural network monitoring model is obtained by iteratively training a preset second BP neural network according to the monitoring parameters corresponding to each of the plurality of components and the instantaneous overload parameter during operation of the power transformer.

3. The method of claim 2, wherein, The method further comprises the following steps: acquiring a rated instantaneous overload parameter of the power transformer when the power transformer is shipped; determining a decay ratio of the power transformer according to the instantaneous overload parameter and the rated instantaneous overload parameter; when the decay ratio is less than a preset decay value, issuing a warning signal.

4. A monitoring device of a power transformer, characterized in that The method comprises the following steps: a state data unit is configured to acquire all state parameters of the power transformer during operation, and screen a plurality of state input parameters from the all state parameters; A model monitoring unit is configured to input a plurality of state input parameters into a preset first neural network monitoring model to obtain monitoring parameters corresponding to a plurality of components on the power transformer respectively; The first neural network monitoring model is obtained by iteratively training a preset first BP neural network according to the plurality of state input parameters and the monitoring parameters corresponding to the plurality of components respectively, all the state parameters include oil chromatogram, line current and transformer voltage, and the monitoring parameters include transformer internal temperature, voltage transformation ratio and load current. The step of screening the plurality of state input parameters from the all state parameters includes: selecting one state parameter from the all state parameters, changing a change value corresponding to the state parameter multiple times, and detecting a detection value corresponding to each monitoring parameter after changing the change value corresponding to the state parameter each time; performing linear fitting processing on the change value corresponding to the state parameter and the detection value of the monitoring parameter to obtain a plurality of fitting slopes between the state parameter and the plurality of monitoring parameters; repeating the step of selecting one state parameter from the all state parameters and the subsequent steps for each state parameter in the all state parameters to obtain a plurality of fitting slopes between each state parameter and each monitoring parameter in the plurality of monitoring parameters; wherein the fitting slope is a correlation coefficient between the state parameter and the monitoring parameter; in the all state parameters, a state parameter with all correlation coefficients corresponding to the state parameter greater than a preset correlation coefficient threshold is determined as a state input parameter, and a plurality of state input parameters are obtained; or, as long as one fitting slope in the plurality of fitting slopes corresponding to one state parameter is greater than the preset correlation coefficient threshold, the state parameter is determined as a state input parameter; obtaining a winding temperature of a winding and a leakage current of a casing on the power transformer based on the first neural network monitoring model; collecting a cooling oil temperature of cooling oil when the power transformer is running; adjusting a wind speed of a fan of the power transformer and adjusting a flow rate of the cooling oil in the power transformer based on the winding temperature and the cooling oil temperature; adjusting a running voltage of the power transformer based on the leakage current.

5. The apparatus of claim 4, wherein, Further comprising: a transient overload unit configured to input the monitoring parameters corresponding to the plurality of components on the power transformer into a preset second neural network monitoring model to obtain a transient overload parameter when the power transformer is running; The second neural network monitoring model is obtained by iteratively training a preset second BP neural network according to the monitoring parameters corresponding to the plurality of components and the transient overload parameter when the power transformer is running.

6. An electronic device, comprising: The computer program is executed by the processor to perform the method of any one of claims 1-3.

7. A computer-readable storage medium having stored thereon a computer program, characterized in that The computer program is executed by the processor to perform the method of any one of claims 1-3.

Images