CN115861222A - Transformer oil leakage detection method, device, equipment and storage medium - Google Patents
Transformer oil leakage detection method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN115861222A CN115861222A CN202211506257.5A CN202211506257A CN115861222A CN 115861222 A CN115861222 A CN 115861222A CN 202211506257 A CN202211506257 A CN 202211506257A CN 115861222 A CN115861222 A CN 115861222A
- Authority
- CN
- China
- Prior art keywords
- transformer
- oil leakage
- oil
- sample
- current
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Housings And Mounting Of Transformers (AREA)
Abstract
The invention discloses a method, a device, equipment and a storage medium for detecting oil leakage of a transformer, wherein the method comprises the following steps: acquiring current operating parameters of a target transformer; preprocessing the current operation parameters to obtain a first current operation characteristic; inputting the first current operation characteristic into a preset vector machine, and outputting the current oil level state of the target transformer based on the preset vector machine; and judging whether the target transformer leaks oil or not according to the current oil level state. According to the invention, the current operation parameters of the target transformer are obtained, and the preprocessed current operation parameters are input into the trained preset vector machine, so that whether the target transformer leaks oil or not can be known, therefore, the labor cost of the conventional manual oil leakage detection is avoided, and in addition, the oil leakage detection time of the invention is less than a large amount of investigation time spent in the conventional manual oil leakage detection process, so that the detection cost is reduced and the detection efficiency is improved.
Description
Technical Field
The invention relates to the technical field of transformer oil leakage monitoring, in particular to a transformer oil leakage detection method, a device, equipment and a storage medium.
Background
The large capacity power transformer of power plant mostly is oil-immersed transformer, and the transformer oil leakage phenomenon takes place occasionally, and serious seepage has not only reduced the life of transformer, but also can influence the safety of power plant, economic operation, so transformer oil leakage detection just is one of the current urgent problems that need to solve of mill.
At present, the oil leakage detection of a transformer is generally implemented by examining leakage-prone parts of a radiator, a valve, a flange interface and the like of the transformer one by one on site by detection personnel to find out leakage fault points, so that the existing manual oil leakage detection method is high in labor cost, long in time consumption and low in detection efficiency. Therefore, a method for detecting oil leakage with low cost and high detection efficiency is needed.
The above is only for the purpose of assisting understanding of the technical solution of the present invention, and does not represent an admission that the above is the prior art.
Disclosure of Invention
The invention mainly aims to provide a transformer oil leakage detection method, a transformer oil leakage detection device, transformer oil leakage detection equipment and a storage medium, and aims to solve the technical problems of high cost and low detection efficiency of the existing transformer oil leakage detection technology. In order to achieve the purpose, the invention provides a transformer oil leakage detection method. The method comprises the following steps:
acquiring current operating parameters of a target transformer;
preprocessing the current operation parameters to obtain a first current operation characteristic;
inputting the first current operation characteristic into a preset vector machine, and outputting the current oil level state of the target transformer based on the preset vector machine;
and judging whether the target transformer leaks oil or not according to the current oil level state.
Optionally, the step of preprocessing the current operation parameter to obtain a first current operation characteristic includes:
acquiring a first operation characteristic corresponding to the current operation parameter;
standardizing the first operation characteristic to obtain a second operation characteristic;
and smoothing the second operation characteristic through a preset Gaussian kernel function to obtain a first current operation characteristic.
Optionally, the step of obtaining the first operation characteristic corresponding to the current operation parameter includes:
acquiring a first time-frequency image corresponding to the current operation parameter;
and performing gray level histogram feature extraction on the first time frequency image to obtain a first operation feature.
Optionally, after determining whether the target transformer leaks oil according to the current oil level state, the method further includes:
when the target transformer is judged to leak oil, inputting the second current operation characteristic of the target transformer into a preset oil leakage classifier to obtain an oil leakage identifier of the target transformer;
and inquiring a corresponding oil leakage reason in a preset mapping relation according to the oil leakage identification.
Optionally, when it is determined that the target transformer leaks oil, before inputting the second current operating characteristic of the target transformer into a preset oil leakage classifier and obtaining an oil leakage identifier of the target transformer, the method further includes:
acquiring historical operation data of a sample transformer and a sample oil leakage reason;
constructing a sample oil leakage identifier of the sample transformer based on the sample oil leakage reason of the sample transformer;
obtaining historical operating characteristics of the sample transformer based on historical operating data of the sample transformer;
inputting the historical operating characteristics and the corresponding sample oil leakage identification into an initial oil leakage classifier for iterative training to obtain a preset oil leakage classifier.
Optionally, the step of obtaining historical operating characteristics of the sample transformer based on the historical operating data of the sample transformer includes:
acquiring a second time-frequency image corresponding to the historical operating data of the sample transformer;
and extracting the characteristics of the second time-frequency image through a ResNet18 backbone network to obtain historical operation characteristics corresponding to the historical operation data.
Optionally, the step of inputting the historical operating characteristics and the corresponding sample oil leakage identifications into an initial oil leakage classifier for iterative training to obtain a preset oil leakage classifier includes:
inputting the historical operating characteristics and the corresponding sample oil leakage identification into an initial oil leakage classifier to obtain the sample oil leakage reason of the sample transformer;
determining a loss value corresponding to the oil leakage classifier based on the sample oil leakage identification and the sample oil leakage reason;
judging whether the initial oil leakage classifier is optimized or not according to the loss value;
if yes, determining target optimization parameters through a preset optimization function, continuously optimizing the initial oil leakage classifier according to the target optimization parameters, and taking the optimized oil leakage classifier as the preset oil leakage classifier.
In addition, in order to achieve the above object, the present invention further provides a transformer oil leakage detection apparatus, including:
the parameter acquisition module is used for acquiring the current operating parameters of the target transformer;
the parameter processing module is used for preprocessing the current operation parameters to obtain a first current operation characteristic;
the state estimation module is used for inputting the first current operation characteristic into a preset vector machine and outputting the current oil level state of the target transformer based on the preset vector machine;
and the oil leakage judging module is used for judging whether the target transformer leaks oil or not according to the current oil level state.
In addition, in order to achieve the above object, the present invention further provides an oil leakage detection apparatus for a transformer, the apparatus comprising: a memory, a processor and a transformer oil leakage detection program stored on the memory and executable on the processor, the transformer oil leakage detection program being configured to implement the steps of the transformer oil leakage detection method as described above.
In addition, to achieve the above object, the present invention further provides a storage medium, wherein the storage medium stores a transformer oil leakage detection program, and the transformer oil leakage detection program, when executed by a processor, implements the steps of the transformer oil leakage detection method as described above.
The invention discloses a method, a device, equipment and a storage medium for detecting oil leakage of a transformer, wherein the method comprises the following steps: acquiring current operating parameters of a target transformer; preprocessing the current operation parameters to obtain a first current operation characteristic; inputting the first current operation characteristic into a preset vector machine, and outputting the current oil level state of the target transformer based on the preset vector machine; and judging whether the target transformer leaks oil or not according to the current oil level state. Firstly, acquiring current operation parameters of a target transformer, preprocessing the current operation parameters, and acquiring a stable first current operation characteristic; and then inputting the first current operation characteristic into a pre-set vector machine which is trained in advance, so that the current oil level state of the target transformer is obtained through the pre-set vector machine. And when the current oil level state of the target transformer is the abnormal oil level state, judging that the target transformer has oil leakage. According to the invention, whether the target transformer leaks oil can be known only by acquiring the current operating parameters of the target transformer and inputting the preprocessed current operating parameters into the trained preset vector machine, so that the labor cost of the conventional manual oil leakage detection is avoided, and in addition, the oil leakage detection time of the invention is less than a large amount of investigation time spent in the conventional manual oil leakage detection process, so that the detection cost is reduced and the detection efficiency is improved.
Drawings
Fig. 1 is a schematic structural diagram of a transformer oil leakage detection device in a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart illustrating a first embodiment of a method for detecting oil leakage of a transformer according to the present invention;
FIG. 3 is a schematic flow chart illustrating a transformer oil leakage detection method according to a second embodiment of the present invention;
FIG. 4 is a schematic flow chart illustrating a transformer oil leakage detection method according to a third embodiment of the present invention;
fig. 5 is a block diagram illustrating a first embodiment of the oil leakage detection apparatus for a transformer according to the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Referring to fig. 1, fig. 1 is a schematic structural diagram of a transformer oil leakage detection device in a hardware operating environment according to an embodiment of the present invention.
As shown in fig. 1, the transformer oil leakage detecting apparatus may include: a processor 1001, such as a Central Processing Unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a Wireless interface (e.g., a Wireless-Fidelity (WI-FI) interface). The Memory 1005 may be a Random Access Memory (RAM) or a Non-Volatile Memory (NVM), such as a disk Memory. The memory 1005 may alternatively be a storage device separate from the processor 1001 described previously.
Those skilled in the art will appreciate that the configuration shown in fig. 1 does not constitute a limitation of the transformer oil leakage detection apparatus, and may include more or less components than those shown, or some components in combination, or a different arrangement of components.
As shown in fig. 1, a memory 1005, which is a storage medium, may include therein an operating system, a data storage module, a network communication module, a user interface module, and a transformer oil leakage detection program.
In the transformer oil leakage detection apparatus shown in fig. 1, the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 of the transformer oil leakage detection apparatus according to the present invention may be disposed in the transformer oil leakage detection apparatus, and the transformer oil leakage detection apparatus calls the transformer oil leakage detection program stored in the memory 1005 through the processor 1001 and executes the transformer oil leakage detection method according to the present invention.
An embodiment of the present invention provides a method for detecting oil leakage of a transformer, and referring to fig. 2, fig. 2 is a schematic flow chart of a first embodiment of the method for detecting oil leakage of a transformer according to the present invention.
In this embodiment, the method for detecting oil leakage of a transformer includes the following steps:
step S10: acquiring current operating parameters of a target transformer;
it should be noted that the execution subject of the method of this embodiment may be a computing service device with data processing, network communication, and program running functions, such as a mobile phone, a tablet computer, a personal computer, and the like, or may be other electronic devices capable of implementing the same or similar functions. The method for detecting oil leakage of a transformer according to the present embodiment and the following embodiments will be specifically described with reference to the above-mentioned oil leakage detection device (abbreviated as oil leakage detection device).
It can be understood that the target transformer is a transformer to be detected for oil leakage.
It should be noted that the current operating parameter may be a parameter related to the current operating state of the target transformer and a name plate parameter of the target transformer, such as: the present oil level of the transformer, the present oil temperature, the operated time of the target transformer, the grade of the transformer, etc., the data content of the current operation parameter data and the number of the related data, which are not limited in this embodiment. In addition, the parameters related to the working state in the current operation parameters are time sequence data collected according to the time sequence.
Step S20: preprocessing the current operation parameters to obtain a first current operation characteristic;
it should be noted that, because the value ranges of the parameters in the data set formed by the current operating parameters of the target transformer are very different, if the current operating parameters are directly input into the preset vector machine, the model prediction of the preset vector machine becomes difficult and is easily interfered by noise, and therefore, the current operating parameters need to be preprocessed, and then the first current operating characteristics obtained after the preprocessing are input into the preset vector machine.
Further, as an implementation manner, in this embodiment, the step S20 includes:
step S201: acquiring a first operation characteristic corresponding to the current operation parameter;
it should be noted that, the first operation characteristic may be obtained by: firstly, a first time-frequency image corresponding to the current operation parameter is obtained, and then the first time-frequency image is subjected to gray histogram feature extraction to obtain the first operation feature.
In addition, since the obtained current operating parameter is time series data, the present embodiment may use a time-frequency analysis method to take the current operating parameter as the first time-frequency image, and the available time-frequency analysis methods include: short-time fourier transform (STFT), morlet Wavelet (MW), and filter-based hilbert transform (FHT), which are not limited in this embodiment, and specifically used as a time-frequency analysis method.
It should be understood that, after the first time-frequency image is obtained, a gray histogram feature extraction method may be used to obtain the time-frequency feature of the target transformer. In practical application, the first time-frequency image can be grayed to obtain a grayscale image of the first time-frequency image; then configuring a window by taking any pixel point in the gray-scale image as a center, acquiring gray values of other pixel points in the window, and comparing the gray values of the other pixel points with the gray value of the center pixel point to acquire a direction gradient value of the center pixel point; and then, a Histogram is constructed according to the direction gradient value of the central pixel point, so that the HOG (Histogram of directional Gradients) feature of the first time-frequency image can be obtained, and the HOG feature is the time-frequency feature.
Step S202: standardizing the first operating characteristic to obtain a second operating characteristic;
step S203: and smoothing the second operation characteristic through a preset Gaussian kernel function to obtain a first current operation characteristic.
It should be noted that, in order to further increase the training speed and prevent overfitting of the model, the time-frequency features described above need to be standardized in this embodiment, and certainly, normalization or regularization may also be performed, and in practical application, the time-frequency features may be processed by using a self-contained standardization, normalization and/or regularization program in Python.
It should be understood that, in this embodiment, the kernel function is used to simplify calculation of a dot product in a preset vector machine and reduce time complexity, and a low-dimensional feature space is mapped into a high-dimensional space to implement linear separable (or curve separable), where common kernel functions include: linear kernel functions, gaussian kernel functions, polynomial kernel functions, and the like, and after training, it is found that since the operation characteristics obtained after the operation parameters of the transformer are processed are more, the accuracy and score of the vector machine using the gaussian kernel functions are much higher than those of the other kernel functions, and therefore the gaussian kernel functions are preferably used in the embodiment to smooth the second operation characteristics. In practical application, a penalty coefficient C of a gaussian kernel function and a kernel function parameter γ need to be set.
Step S30: inputting the first current operation characteristic into a preset vector machine, and outputting the current oil level state of the target transformer based on the preset vector machine;
step S40: and judging whether the target transformer leaks oil or not according to the current oil level state.
It should be noted that the preset vector machine is a support vector machine trained in advance, and the preset vector machine may be a data model trained based on the operation parameters of a large number of sample transformers. In the actual training process, a training set can be established based on a part of operation parameters of a sample transformer, then the data in the training set is preprocessed, the preprocessed sample operation parameters are substituted into an SVM algorithm for training to generate an initial vector machine model, finally a test set is established based on the other part of operation parameters of the sample transformer, correct classification marks are added to the data in the test set, the data in the test set are substituted into the initial vector machine model after being preprocessed for classification prediction, finally the classification result is compared with an actual classification label to obtain the accuracy, the recall ratio and the like of the initial vector machine, the initial vector machine is iteratively updated based on the accuracy and the recall ratio, and the preset vector machine is finally obtained.
In addition, it should be understood that the output result of the preset vector machine is divided into a positive type and a negative type, in this embodiment, if the output result of the preset vector machine is the negative type, it indicates that the current oil level state of the target transformer is an abnormal state, and it can be determined that the target transformer has oil leakage, and a warning lamp can be flashed or warning information can be sent to related staff to remind the target transformer that maintenance is needed; if the output result of the preset vector machine is positive, the current oil level state of the target transformer is a normal state, and the target transformer can be judged to have no oil leakage.
The method comprises the steps of obtaining current operation parameters of a target transformer; preprocessing the current operation parameters to obtain a first current operation characteristic; inputting the first current operation characteristic into a preset vector machine, and outputting the current oil level state of the target transformer based on the preset vector machine; and judging whether the target transformer leaks oil or not according to the current oil level state. The method comprises the steps of firstly, obtaining current operation parameters of a target transformer, preprocessing the current operation parameters, and obtaining stable first current operation characteristics; and then inputting the first current operation characteristic into a pre-set vector machine trained in advance, so as to obtain the current oil level state of the target transformer through the pre-set vector machine. And when the current oil level state of the target transformer is the abnormal oil level state, judging that the target transformer has oil leakage. Therefore, in the embodiment, only the current operation parameters of the target transformer need to be acquired, and then the preprocessed current operation parameters are input into the trained preset vector machine, so that whether the target transformer leaks oil or not can be known.
Referring to fig. 3, fig. 3 is a schematic flow chart of a transformer oil leakage detection method according to a second embodiment of the present invention, and the transformer oil leakage detection method according to the second embodiment of the present invention is proposed based on the embodiment shown in fig. 2.
It can be understood that after the transformer oil leakage is detected by the preset vector machine, the specific reason for the transformer oil leakage may be further analyzed, so that the relevant personnel may find the oil leakage fault point target of the transformer as soon as possible to remedy the oil leakage fault point target, and therefore, in this embodiment, the step S40 is followed by:
step S50: when the target transformer is judged to leak oil, inputting the second current operation characteristic of the target transformer into a preset oil leakage classifier to obtain an oil leakage identifier of the target transformer;
it should be noted that the above-mentioned predetermined vector machine is generally only suitable for two-class situations, and the oil leakage of the transformer generally occurs in the following main parts: the body main cover is connected face, radiator, sample valve, sleeve pipe, transformer base kneck and buchholz relay etc. therefore the oil leak reason of transformer far exceeds two kinds, consequently this embodiment is through constructing the oil leak classifier and come the oil leak reason of analysis target transformer.
It can be understood that the preset oil leakage classifier is an oil leakage classifier which is finally trained, and therefore before the second current operation characteristic is input into the preset oil leakage classifier, the preset oil leakage classifier needs to be trained based on historical operation data of a sample transformer in which an oil leakage condition occurs and a sample oil leakage reason.
It should be understood that, in the process of constructing the preset oil leakage classifier, the historical operating data of the sample transformer is used to generate corresponding historical operating characteristics, and the sample oil leakage reason of the sample transformer may be an actual oil leakage reason of the sample transformer, and may be used to construct a sample oil leakage identifier of the sample transformer.
It should be noted that the preset historical operating data of the oil leakage classifier and the current historical operating parameters are data with the same dimension, and the content to be collected is also the same, and may be parameters related to the current working state of the sample transformer and the nameplate parameters of the sample transformer, for example: current oil level of the transformer, current oil temperature, run time of the target transformer, grade of the transformer, and the like. The second current operation characteristic of the target transformer and the historical operation characteristic of the sample transformer are obtained in the same manner, and taking the sample transformer as an example, in practical application, the method may be: acquiring a second time-frequency image corresponding to historical operating data of the sample transformer; and extracting the characteristics of the second time-frequency image through a ResNet18 backbone network to obtain historical operation characteristics corresponding to the historical operation data.
In addition, this sample oil leak sign is based on the common oil leak reason of transformer and is constructed, specifically, three kinds of classification signs have been constructed for the sample transformer based on the oil leakage volume of the common oil leak trouble of transformer and each oil leak trouble in this embodiment: the mark 1 represents a kind of interface leakage (specifically, the interface leakage can be a radiator interface, a plane butterfly valve cap, a sampling valve or a connecting flange seal, etc.); the mark 2 represents the leakage of the second type of interface (specifically, the leakage can be a gas relay (or a gas relay) interface, a transformer base interface or a body large cover connecting surface, etc.); the identifier 3 represents an inner leakage (specifically, a casing leakage or a porcelain insulator leakage), so in practical application, the sample oil leakage identifier may be selected from the identifiers 1 to 3 based on an actual oil leakage reason of the sample transformer, and then the historical operating characteristics and the sample oil leakage identifier of the sample transformer are input to an initial oil leakage classifier for iterative training, so as to obtain the preset oil leakage classifier, which is not limited by the specific name of the identifier.
It should be noted that, the preset oil leakage classifier finally outputs the corresponding probabilities of the three types of classification identifiers of the target transformer, where the classification identifier with the highest probability is the oil leakage identifier of the target transformer that is desired to be obtained in this embodiment. In practical application, if the output result of the preset oil leakage classifier of the target transformer is as follows: identification 1:0.7; and identification 2:0.2; identification 1: and 0.1, the oil leakage mark corresponding to the target transformer is marked as a mark 1.
Step S60: and inquiring a corresponding oil leakage reason in a preset mapping relation according to the oil leakage identification.
It should be noted that, a preset mapping relationship is stored in the oil leakage detection device, and the preset mapping relationship table includes mapping relationships between the oil leakage identifications and the oil leakage reasons corresponding to the oil leakage identifications. For example, if the oil leakage mark of the target transformer is mark 2, the oil leakage cause of the target transformer is: the second type of interface leakage (specifically, the interface can be a gas relay (or a gas relay) interface, a transformer base interface or a body large cover connecting surface, etc.).
In the embodiment, historical operation data of a sample transformer and the reason of sample oil leakage are obtained; constructing a sample oil leakage identifier of the sample transformer based on the sample oil leakage reason of the sample transformer; obtaining historical operating characteristics of a sample transformer based on historical operating data of the sample transformer; inputting the historical operation characteristics and the corresponding sample oil leakage identifications into an initial oil leakage classifier to perform iterative training to obtain a preset oil leakage classifier; when the target transformer is judged to leak oil, inputting the second current operation characteristic of the target transformer into a preset oil leakage classifier to obtain an oil leakage identifier of the target transformer; and inquiring the corresponding oil leakage reason in a preset mapping relation according to the oil leakage identification. Therefore, in the embodiment, iterative training is performed based on the historical operating characteristics of the sample transformer and the corresponding sample oil leakage identification to obtain the preset oil leakage classifier, and then the prediction of the specific oil leakage reason of the target transformer is realized through the pre-trained preset oil leakage classifier, so that related personnel can find the oil leakage fault point target of the transformer as soon as possible to remedy, and the oil leakage detection efficiency is further improved.
Referring to fig. 4, fig. 4 is a schematic flow chart of a transformer oil leakage detection method according to a third embodiment of the present invention, which is proposed based on the embodiment shown in fig. 2 or 3, and fig. 4 takes the embodiment proposed based on the embodiment shown in fig. 1 as an example.
It can be understood that after the initial oil leak classifier is constructed, iterative training needs to be continuously performed based on a large number of samples, necessary evaluation indexes need to be calculated according to output prediction results to evaluate the performance of the classifier, and the classifier network is corrected or updated, so that the accuracy of the output of the oil leak classifier model is improved, and the preset oil leak classifier which can be finally used for oil leak reason prediction is obtained.
Further, as an implementation manner, before step S50 in this implementation, the method further includes:
step S411: inputting the historical operating characteristics and the corresponding sample oil leakage identification into an initial oil leakage classifier to obtain the sample oil leakage reason of the sample transformer;
step S412: determining a loss value corresponding to the oil leakage classifier based on the sample oil leakage identification and the sample oil leakage reason;
it should be noted that the loss value is calculated by using the loss function corresponding to the initial oil leakage classifier, and since the oil leakage classifier constructed in the present embodiment is a Softmax classifier, the corresponding loss function is a cross-entropy loss function, and a formula of the cross-entropy loss function is as follows:
wherein L represents the loss value, i represents a historical operating characteristic of the sample transformer, N represents the total number of the historical operating characteristics of the sample transformer, and y i Representative sample leak indicator, s i Representing the probability corresponding to the reason of the oil leakage of the sample
Step S413: judging whether the initial oil leakage classifier is optimized or not according to the loss value;
step S414: if yes, determining target optimization parameters through a preset optimization function, continuously optimizing the initial oil leakage classifier according to the target optimization parameters, and taking the optimized oil leakage classifier as the preset oil leakage classifier.
It should be noted that, a loss threshold may be stored in the oil leakage detection device, and the loss value is compared with the loss threshold, and if the loss value exceeds the loss threshold, it represents that the current oil leakage classifier model is not good enough, and therefore, optimization is required for the time. The optimization process may be to configure an optimization function F (X, W, B) in the oil leakage detection apparatus, that is, the input data of the optimization function F (X, W, B) may include training data (i.e., the historical operating data of the sample transformer) X, the weight W of the current oil leakage classifier, and the deviation B. The specific optimization method may be a back propagation method based on the training data X and a gradient descent method based on the parameters (W and B) of the oil leakage classifier, and the embodiment does not limit this. However, since the training data is usually given based on the training set, in this embodiment, the model is preferably optimized based on a gradient descent method with weights W and deviations B, and in practical applications, after the oil leakage detection device detects that the loss value is greater than the loss threshold, it may continuously train based on the optimization function F (X, W, B) to find the weight W and the deviation B corresponding to the oil leakage classifier when the gradient (i.e., the derivative of F with respect to W or the derivative of F with respect to B) of the optimization function F (X, W, B) reaches the minimum, where the weight W and the deviation B are the target optimization parameters. The oil leak detection device will repeat this process until the loss value no longer exceeds the loss threshold, and the oil leak detection device will use the finally optimized oil leak classifier as the preset oil leak classifier.
In the embodiment, the sample oil leakage reason of the sample transformer is obtained by inputting the historical operating characteristics and the corresponding sample oil leakage identification into the initial oil leakage classifier; determining a loss value corresponding to the oil leakage classifier based on the sample oil leakage identification and the sample oil leakage reason; judging whether to optimize the initial oil leakage classifier or not according to the loss value; if yes, determining target optimization parameters through a preset optimization function, continuously optimizing the initial oil leakage classifier according to the target optimization parameters, and taking the optimized oil leakage classifier as a preset oil leakage classifier. According to the method and the device, the oil leakage classifier is updated through the loss function and the optimization function, and the prediction performance and the generalization capability of the classifier model are improved.
In addition, an embodiment of the present invention further provides a storage medium, where a transformer oil leakage detection program is stored on the storage medium, and when executed by a processor, the transformer oil leakage detection program implements the steps of the transformer oil leakage detection method described above.
Referring to fig. 5, fig. 5 is a block diagram illustrating a first embodiment of a transformer oil leakage detection apparatus according to the present invention.
As shown in fig. 5, a transformer oil leakage detection apparatus provided in an embodiment of the present invention includes:
the parameter acquisition module is used for acquiring the current operating parameters of the target transformer;
the parameter processing module is used for preprocessing the current operation parameters to obtain a first current operation characteristic;
the state estimation module is used for inputting the first current operation characteristic into a preset vector machine and outputting the current oil level state of the target transformer based on the preset vector machine;
and the oil leakage judging module is used for judging whether the target transformer leaks oil or not according to the current oil level state.
The embodiment obtains the current operation parameters of the target transformer; preprocessing the current operation parameters to obtain a first current operation characteristic; inputting the first current operation characteristic into a preset vector machine, and outputting the current oil level state of the target transformer based on the preset vector machine; and judging whether the target transformer leaks oil or not according to the current oil level state. The method comprises the steps of firstly, obtaining current operation parameters of a target transformer, preprocessing the current operation parameters, and obtaining stable first current operation characteristics; and then inputting the first current operation characteristic into a pre-set vector machine trained in advance, so as to obtain the current oil level state of the target transformer through the pre-set vector machine. And when the current oil level state of the target transformer is the abnormal oil level state, judging that the target transformer has oil leakage. Therefore, in the embodiment, only the current operation parameters of the target transformer need to be acquired, and then the preprocessed current operation parameters are input into the trained preset vector machine, so that whether the target transformer leaks oil or not can be known.
Other embodiments or specific implementation manners of the transformer oil leakage detection device of the present invention may refer to the above method embodiments, and are not described herein again.
Claims (10)
1. A method for detecting oil leakage of a transformer is characterized by comprising the following steps:
acquiring current operating parameters of a target transformer;
preprocessing the current operation parameters to obtain a first current operation characteristic;
inputting the first current operation characteristic into a preset vector machine, and outputting the current oil level state of the target transformer based on the preset vector machine;
and judging whether the target transformer leaks oil or not according to the current oil level state.
2. The method for detecting oil leakage of transformer according to claim 1, wherein the step of preprocessing the current operation parameters to obtain the first current operation characteristic comprises:
acquiring a first operation characteristic corresponding to a current operation parameter;
standardizing the first operation characteristic to obtain a second operation characteristic;
and smoothing the second operation characteristic through a preset Gaussian kernel function to obtain a first current operation characteristic.
3. The method for detecting oil leakage of transformer according to claim 2, wherein the step of obtaining the first operation characteristic corresponding to the current operation parameter includes:
acquiring a first time-frequency image corresponding to the current operation parameter;
and performing gray level histogram feature extraction on the first time frequency image to obtain a first operation feature.
4. The method for detecting oil leakage from transformer according to claim 3, wherein after determining whether the target transformer leaks oil according to the current oil level state, the method further comprises:
when the oil leakage of the target transformer is judged, inputting the second current operation characteristic of the target transformer into a preset oil leakage classifier to obtain an oil leakage identifier of the target transformer;
and inquiring the corresponding oil leakage reason in a preset mapping relation according to the oil leakage identification.
5. The method for detecting oil leakage of transformer according to claim 4, wherein when it is determined that the target transformer leaks oil, before inputting the second current operation characteristic of the target transformer into a preset oil leakage classifier to obtain an oil leakage identifier of the target transformer, the method further comprises:
acquiring historical operation data of a sample transformer and a sample oil leakage reason;
constructing a sample oil leakage identifier of the sample transformer based on the sample oil leakage reason of the sample transformer;
obtaining historical operation characteristics of the sample transformer based on historical operation data of the sample transformer;
inputting the historical operation characteristics and the corresponding sample oil leakage identification into an initial oil leakage classifier for iterative training to obtain a preset oil leakage classifier.
6. The method for detecting oil leakage in a transformer of claim 5, wherein the step of obtaining historical operating characteristics of the sample transformer based on historical operating data of the sample transformer comprises:
acquiring a second time-frequency image corresponding to historical operating data of the sample transformer;
and extracting the characteristics of the second time-frequency image through a ResNet18 backbone network to obtain historical operation characteristics corresponding to the historical operation data.
7. The method for detecting oil leakage of transformer according to claim 6, wherein before inputting the second current operating characteristic of the target transformer into the preset oil leakage classifier and obtaining the oil leakage identification of the target transformer when oil leakage of the target transformer is determined, the method further comprises:
inputting the historical operation characteristics and the corresponding sample oil leakage identification into an initial oil leakage classifier to obtain the sample oil leakage reason of the sample transformer;
determining a loss value corresponding to the oil leakage classifier based on the sample oil leakage identification and the sample oil leakage reason;
judging whether to optimize the initial oil leakage classifier or not according to the loss value;
if yes, determining target optimization parameters through a preset optimization function, continuously optimizing the initial oil leakage classifier according to the target optimization parameters, and taking the optimized oil leakage classifier as the preset oil leakage classifier.
8. The utility model provides a transformer oil leak detection device which characterized in that, transformer oil leak detection device includes:
the parameter acquisition module is used for acquiring the current operating parameters of the target transformer;
the parameter processing module is used for preprocessing the current operation parameters to obtain a first current operation characteristic;
the state estimation module is used for inputting the first current operation characteristic into a preset vector machine and outputting the current oil level state of the target transformer based on the preset vector machine;
and the oil leakage judging module is used for judging whether the target transformer leaks oil or not according to the current oil level state.
9. An oil leakage detection apparatus for a transformer, the apparatus comprising: a memory, a processor and a transformer oil leakage detection program stored on the memory and executable on the processor, the transformer oil leakage detection program being configured to implement the steps of the transformer oil leakage detection method according to any one of claims 1 to 7.
10. A storage medium having a transformer oil leakage detection program stored thereon, wherein the transformer oil leakage detection program, when executed by a processor, implements the steps of the transformer oil leakage detection method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211506257.5A CN115861222A (en) | 2022-11-28 | 2022-11-28 | Transformer oil leakage detection method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211506257.5A CN115861222A (en) | 2022-11-28 | 2022-11-28 | Transformer oil leakage detection method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115861222A true CN115861222A (en) | 2023-03-28 |
Family
ID=85667433
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211506257.5A Pending CN115861222A (en) | 2022-11-28 | 2022-11-28 | Transformer oil leakage detection method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115861222A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118209260A (en) * | 2024-05-21 | 2024-06-18 | 南通世睿电力科技有限公司 | Oil leakage monitoring and early warning system for transformer oil storage cabinet |
-
2022
- 2022-11-28 CN CN202211506257.5A patent/CN115861222A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118209260A (en) * | 2024-05-21 | 2024-06-18 | 南通世睿电力科技有限公司 | Oil leakage monitoring and early warning system for transformer oil storage cabinet |
| CN118209260B (en) * | 2024-05-21 | 2024-08-27 | 南通世睿电力科技有限公司 | Oil leakage monitoring and early warning system for transformer oil storage cabinet |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11556934B1 (en) | Heuristic account fraud detection engine | |
| CN110007220B (en) | Method and device for diagnosing operating state of circuit breaker mechanism | |
| CN110472675B (en) | Image classification method, image classification device, storage medium and electronic equipment | |
| CN111818198B (en) | Domain name detection method, domain name detection device, equipment and medium | |
| CN113314144B (en) | Voice recognition and power equipment fault early warning method, system, terminal and medium | |
| CN115861222A (en) | Transformer oil leakage detection method, device, equipment and storage medium | |
| CN111583592B (en) | Experimental environment safety early warning method based on multidimensional convolution neural network | |
| CN116167010B (en) | Rapid identification method for abnormal events of power system with intelligent transfer learning capability | |
| CN112069930A (en) | Vibration signal processing method and device for improving GIS equipment fault diagnosis accuracy | |
| CN112765003B (en) | Risk prediction method based on APP behavior log | |
| CN115524002A (en) | Running state early warning method and system for power plant rotating equipment and storage medium | |
| WO2023207557A1 (en) | Method and apparatus for evaluating robustness of service prediction model, and computing device | |
| US20220147777A1 (en) | Automatic Partial Discharge and Noise Signals Separation using Arithmetic Coding in Time Domain and Magnitude Distributions in Frequency Domain | |
| CN116740900B (en) | SVM-based power construction early warning method and system | |
| US12031848B2 (en) | Method and computing device for detecting anomalous sensor data | |
| CN116772122A (en) | Natural gas pipeline leakage fault diagnosis method, system, equipment and medium | |
| CN116384223A (en) | Nuclear equipment reliability assessment method and system based on intelligent degradation state identification | |
| CN116451116A (en) | Fault identification model construction method, device, computer equipment and storage medium | |
| CN116012152A (en) | Method, device and equipment for identifying abnormal transaction entity and readable storage medium | |
| CN114414951A (en) | Method and system for diagnosing insulation defect of gas insulated metal enclosed switchgear | |
| CN114970490A (en) | Text labeling data quality inspection method and device, electronic equipment and storage medium | |
| JP2019160072A (en) | Information processing device, information processing method, and program | |
| CN117633634A (en) | Non-stationary working condition fault diagnosis method, system and medium | |
| CN112307090B (en) | Detection method and system based on tunnel security data | |
| CN116304776B (en) | Power grid data value anomaly detection method and system based on k-Means algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |