Transformer operation state detection method and device

Abstract

The invention discloses a transformer operation state detection method and device, and the method comprises the steps: obtaining the current operation data of a transformer, and enabling the current operation data to be used for representing the operation condition of the transformer in a current processing period; determining a current feature matrix according to the current operation data, wherein the current feature matrix is used for representing key operation features of the transformer in the current processing period; and determining a state evaluation result according to the current characteristic matrix and a preset target state analysis model, wherein the state evaluation result is used for indicating the operation state of the transformer in the current operation period. The current characteristic matrix representing the key operation characteristics of the transformer in the current processing period is determined according to the current operation data, and then the state evaluation result is determined according to the current characteristic matrix and the target state analysis model, so that the influence of redundant information on the output result is reduced; and the target state analysis model is suitable for diversified transformers, so that the stability and accuracy of evaluation and fault diagnosis of the transformers are improved.

Description

Technical Field

[0001] The present application relates to the field of information processing technology, and in particular to a method and device for detecting the operating status of a transformer. Background Art

[0002] At present, transformers are key equipment to ensure the safe operation of power grids, so it is necessary to evaluate the operating status of transformers and diagnose faults. In the prior art, the collected transformer data is evaluated and faults are diagnosed through artificial intelligence algorithms that require a large amount of data support. Since such algorithms are affected by redundant information in the input data and cannot accurately and effectively capture the dependencies between the transformer's operating data, the accuracy and stability of transformer operating status evaluation and fault diagnosis are low. Therefore, how to improve the stability and accuracy of transformer evaluation and fault diagnosis has become a technical problem that needs to be further solved. Summary of the invention

[0003] The present application proposes a transformer operating status detection method and device to solve the problem of low accuracy in transformer evaluation and fault diagnosis, and to improve the stability and accuracy of transformer evaluation and fault diagnosis.

[0004] In a first aspect, an embodiment of the present application provides a transformer operation status detection method, which is applied to a server in a transformer operation status detection system, wherein the transformer operation status detection system includes a transformer and the server, and the method includes:

[0005] Acquiring current operating data of the transformer, wherein the current operating data is used to characterize the operating status of the transformer in a current processing cycle;

[0006] Determine a current feature matrix according to the current operation data, wherein the current feature matrix is ​​used to characterize key operation characteristics of the transformer in the current processing cycle;

[0007] A state evaluation result is determined according to the current feature matrix and a preset target state analysis model, wherein the target state analysis model is used to analyze the before-after dependencies of each key operating feature, and the state evaluation result is used to indicate the operating state of the transformer in the current operating cycle.

[0008] In a second aspect, an embodiment of the present application provides a transformer operation status detection device, which is applied to a server in a transformer operation status detection system. The transformer operation status detection system includes a transformer and the server. The device includes:

[0009] A first receiving unit, used for acquiring current operation data of the transformer, wherein the current operation data is used for representing the operation status of the transformer in a current processing cycle;

[0010] A first processing unit is used to determine a current feature matrix based on the current operating data, wherein the current feature matrix is ​​used to characterize key operating features of the transformer in the current processing cycle; determine a state evaluation result based on the current feature matrix and a preset target state analysis model, wherein the target state analysis model is used to analyze the before-and-after dependencies of each key operating feature, and the state evaluation result is used to indicate the operating state of the transformer in the current operating cycle.

[0011] In a third aspect, an embodiment of the present application provides a server, comprising a processor, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the program includes instructions for executing the steps in the method described in any one of the first aspects.

[0012] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program / instruction stored thereon, wherein the computer program / instruction, when executed by a processor, implements the steps of the method described in any one of the first aspects.

[0013] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements part or all of the steps of the method described in any one of the first aspects of the embodiments of the present application.

[0014] It can be seen that in this application, the server obtains the current operating data of the transformer, and the current operating data is used to characterize the operating status of the transformer in the current processing cycle; the current feature matrix is ​​determined based on the current operating data, and the current feature matrix is ​​used to characterize the key operating characteristics of the transformer in the current processing cycle; the state evaluation result is determined based on the current feature matrix and the preset target state analysis model, and the target state analysis model is used to analyze the front-end dependency of each key operating characteristic, and the state evaluation result is used to indicate the operating state of the transformer in the current operating cycle. Since the current feature matrix characterizing the key operating characteristics of the transformer in the current processing cycle is determined based on the current operating data, and then the state evaluation result is determined based on the current feature matrix and the target state analysis model, not only the influence of redundant information on the output result of the target state analysis model is reduced, but also the target state analysis model can capture the front-end dependency in the operating data of the transformer, thereby improving the stability and accuracy of transformer evaluation and fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 It is a structural schematic diagram of a transformer operation status detection system provided in an embodiment of the present application;

[0017] Figure 2 It is a structural diagram of a server in a transformer operation status detection system provided in an embodiment of the present application;

[0018] Figure 3 It is a structural diagram of a fault diagnosis module provided in an embodiment of the present application;

[0019] Figure 4 It is a structural diagram of a server in another transformer operation status detection system provided in an embodiment of the present application;

[0020] Figure 5 It is a flow chart of a transformer operating status detection method provided in an embodiment of the present application;

[0021] Figure 6 It is a flow chart of another transformer operating status detection method provided in an embodiment of the present application;

[0022] Figure 7 It is a structural diagram of a data enhancement model provided in an embodiment of the present application;

[0023] Figure 8 is a structural diagram of a target state classification model provided in an embodiment of the present application;

[0024] Fig. 9 This is a functional unit composition block diagram of a transformer operating status detection device provided in an embodiment of the present application;

[0025] Fig.10 It is a functional unit composition block diagram of another transformer operation status detection device provided in an embodiment of the present application;

[0026] Fig.11 It is a structural block diagram of a server provided in an embodiment of the present application. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0028] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.

[0029] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0030] In the embodiments of the present application, "and / or" describes the association relationship of the associated objects, indicating that three relationships may exist. For example, A and / or B can represent the following three situations: A exists alone; A and B exist at the same time; B exists alone. Among them, A and B can be singular or plural.

[0031] In the embodiment of the present application, the symbol " / " can indicate that the objects associated with each other are in an "or" relationship. In addition, the symbol " / " can also indicate a division sign, that is, performing a division operation. For example, A / B can indicate A divided by B.

[0032] In the embodiments of the present application, "at least one item" or similar expressions refer to any combination of these items, including any combination of single items or plural items, and refer to one or more, and multiple refers to two or more. For example, at least one item of a, b, or c can represent the following seven situations: a, b, c, a and b, a and c, b and c, a, b, and c. Among them, each of a, b, and c can be an element or a set containing one or more elements.

[0033] In the embodiments of the present application, "equal to" can be used in conjunction with greater than, and is applicable to the technical solution adopted when greater than, and can also be used in conjunction with less than, and is applicable to the technical solution adopted when less than. When equal to is used in conjunction with greater than, it is not used in conjunction with less than; when equal to is used in conjunction with less than, it is not used in conjunction with greater than.

[0034] In order to better understand the solutions of the embodiments of the present application, the terminal devices, related concepts and backgrounds that may be involved in the embodiments of the present application are first introduced below.

[0035] As a key equipment for energy conversion, transmission and distribution, the reliability of transformers plays a vital role in the stability of power supply. Transformer defects and failures may be caused by factors such as overload, overvoltage, internal insulation aging, and natural environment, which will not only cause huge economic losses and affect the daily life of power consumers, but also affect public safety and cause serious social losses. Continuously detecting and evaluating the operating status of power transformers and accurately grasping the operating status of transformers are of great significance to reducing failures, reducing operating costs, and improving reliability performance. In the prior art, artificial intelligence algorithms that require a large amount of data support are used to evaluate and diagnose faults in the collected transformer data. Since such algorithms are interfered by redundant information in the input data and cannot accurately and effectively capture the dependencies between the previous and next operating data of the transformer, the accuracy and stability of transformer operating status evaluation and fault diagnosis are low. Therefore, how to improve the stability and accuracy of transformer evaluation and fault diagnosis has become a technical problem that needs to be further solved.

[0036] To solve the above problems, an embodiment of the present application provides a transformer operating status detection method and device. The method determines a current feature matrix that characterizes the key operating characteristics of the transformer in the current processing cycle based on current operating data, and then determines the status evaluation result based on the current feature matrix and the target state analysis model. This not only reduces the impact of redundant information on the output results of the target state analysis model, but also enables the target state analysis model to capture the forward and backward dependencies in the transformer operating data, thereby improving the stability and accuracy of transformer evaluation and fault diagnosis.

[0037] See also Figure 1 , Figure 1 Schematic diagram of a transformer operation status detection system provided in an embodiment of the present application. Figure 1As shown, the transformer operation status detection system 100 includes a transformer 110 and a server 120, wherein the transformer 110 and the server 120 are connected in communication, and the transformer 110 can be a transformer, or a transformer group consisting of multiple transformers. The server 120 can be a server, or a server group consisting of multiple servers. The transformer operation status detection system 100 also includes a terminal device 130, wherein the terminal device 130 is connected in communication with the server 120, and the terminal device 130 can be a mobile terminal, a tablet computer, or a laptop computer, etc.

[0038] In daily use of the transformer operation status detection system 100, the server 120 obtains the current operation data of the transformer, and the current operation data is used to characterize the operation status of the transformer in the current processing cycle; the current feature matrix is ​​determined based on the current operation data, and the current feature matrix is ​​used to characterize the key operation characteristics of the transformer in the current processing cycle; the state evaluation result is determined based on the current feature matrix and a preset target state analysis model, and the target state analysis model is used to analyze the before-and-after dependencies of each key operation characteristic, and the state evaluation result is used to indicate the operation status of the transformer in the current operation cycle.

[0039] See also Figure 2 , Figure 2 Schematic diagram of the structure of a server in a transformer operation status detection system provided in an embodiment of the present application. Figure 2 As shown, the server 120 includes a data preprocessing module 210, a data enhancement module 220, a feature extraction module 230 and a fault diagnosis module 240. The data preprocessing module 210 is respectively communicated with the data enhancement module 220 and the feature extraction module 230, the feature extraction module 230 is respectively communicated with the data enhancement module 220 and the fault diagnosis module 240, the server 120 is communicated with the transformer 110 through the data preprocessing module 210, and the server 120 is communicated with the terminal device 130 through the fault diagnosis module 240.

[0040] The data preprocessing module 210 is used to obtain the current operating data of the transformer, and the current operating data is used to characterize the operating status of the transformer in the current processing cycle.

[0041] The data enhancement module 220 is used to perform at least one expansion and judgment operation on the historical operation data of the transformer.

[0042] The feature extraction module 230 is used to determine a current feature matrix according to the current operation data.

[0043] The fault diagnosis module 240 is used to determine the state evaluation result according to the current feature matrix and a preset target state analysis model.

[0044] See also Figure 3 , Figure 3 Schematic diagram of a fault diagnosis module provided in an embodiment of the present application. Figure 3 As shown, the fault diagnosis module 240 includes a bidirectional long short-term memory neural network layer 301, a fully connected layer 302 and a classification function layer 303. The fault diagnosis module 240 is communicated with the feature extraction module 230 through the bidirectional long short-term memory neural network layer 301, the bidirectional long short-term memory neural network layer 301 is communicated with the fully connected layer 302, the fully connected layer 302 is communicated with the classification function layer 303, and the fault diagnosis module is communicated with the terminal device 130 through the classification function layer.

[0045] See also Figure 4 , Figure 4 Schematic diagram of the structure of another server in the transformer operation status detection system provided by the embodiment of the present application. Figure 4 As shown, the server 120 includes a processor 410 and a memory 420, and the processor 410 is communicatively connected to the memory 420. Among them, one or more programs are stored in the memory 420, and the one or more programs are configured to be executed by the processor 410. The function of the one or more programs is to obtain the current operating data of the transformer, and the current operating data is used to characterize the operating status of the transformer in the current processing cycle; determine the current feature matrix according to the current operating data, and the current feature matrix is ​​used to characterize the key operating characteristics of the transformer in the current processing cycle; determine the state evaluation result according to the current feature matrix and the preset target state analysis model, and the target state analysis model is used to analyze the front-end dependency of each key operating characteristic, and the state evaluation result is used to indicate the operating status of the transformer in the current operating cycle.

[0046] A transformer operating status detection method provided by an embodiment of the present application is introduced below.

[0047] See also Figure 5 , Figure 5 is a flow chart of a transformer operating status detection method provided by an embodiment of the present application, which is applied to Figure 1 The server 120 in the transformer operation status detection system 100 shown in the figure, the transformer operation status detection system 100 includes a transformer 110 and a server 120, the transformer 110 and the server 120 are communicatively connected, and the transformer operation status detection system 100 also includes a terminal device 130, and the terminal device 130 is communicatively connected to the server 120; Figure 5As shown, the method comprises the following steps:

[0048] Step S501, obtaining current operating data of the transformer.

[0049] The current operating data is used to characterize the operating status of the transformer in the current processing cycle.

[0050] Among them, the method for obtaining the current operating data includes the following steps: the data preprocessing module in the server obtains the dissolved gas analysis data of the transformer, and the dissolved gas analysis data is used to indicate the content data of compounds in the transformer oil of the transformer; multiple characteristic quantities of the dissolved gas analysis data are constructed by IEC three-ratio method and Rogers four-ratio method; and the multiple characteristic quantities of the dissolved gas analysis data are normalized to obtain the current operating data.

[0051] The IEC three-ratio method includes calculating C2H2 / C2H4, CH4 / H2 and C2H4 / C2H6 according to the dissolved gas analysis data to obtain three comparison values. The Rogers four-ratio method includes calculating H2 / CH4, CH4 / C2H6, C2H2 / C2H4 and CO / CO2 according to the dissolved gas analysis data to obtain four comparison values. The construction of multiple feature quantities by the IEC three-ratio method and the Rogers four-ratio method includes: determining the feature quantities of the three comparison values ​​and the four comparison values ​​to obtain the multiple feature quantities.

[0052] The data preprocessing module normalizes the multiple characteristic quantities of the dissolved gas analysis data according to the following formula:

[0053]

[0054] Among them, x is a plurality of characteristic quantities of the dissolved gas analysis data, μ is the mean value of x, σ is the standard deviation of the characteristic quantities of the dissolved gas analysis data, and x′ is the result of normalization processing of x, that is, the current operating data.

[0055] Step S502: determining a current feature matrix according to the current operation data.

[0056] The current feature matrix is ​​used to characterize key operating characteristics of the transformer in the current processing cycle.

[0057] Wherein, the feature extraction module determines the current feature matrix according to the current operation data.

[0058] In a possible embodiment, determining the current feature matrix according to the current operating data includes: performing feature extraction on the current operating data to obtain an initial feature matrix, where the dimension of the initial feature matrix is ​​the initial dimension; mapping the initial feature matrix to a first dimension to obtain a first feature matrix, where the first dimension is higher than the initial dimension; and determining the current feature matrix according to the first feature matrix, where the dimension of the current feature matrix is ​​lower than the first dimension.

[0059] The step of mapping the initial feature matrix to a first dimension to obtain a first feature matrix includes: obtaining a preset mapping function, where the mapping function is used to map the initial feature matrix to the first dimension; and determining the first feature matrix according to the mapping function and the initial feature matrix.

[0060] The mapping function may be a kernel function, which may be selected and adjusted according to actual conditions. A linear kernel, a polynomial kernel, a Gaussian kernel, etc. may be selected according to different conditions.

[0061] It can be seen that in this example, after obtaining the initial feature matrix by performing preliminary feature extraction on the current operating data, the initial feature matrix is ​​mapped to a higher dimension to obtain a first feature matrix with stronger data expression ability, so as to obtain the current feature matrix that characterizes the key operating characteristics of the transformer in the current processing cycle according to the first feature matrix, and then determine the state assessment result according to the current feature matrix and the target state analysis model. This not only reduces the impact of redundant information on the output results of the target state analysis model, but also enables the target state analysis model to capture the forward and backward dependencies in the operating data of the transformer, thereby improving the stability and accuracy of transformer assessment and fault diagnosis.

[0062] In a possible embodiment, determining the current feature matrix based on the first feature matrix includes: determining the covariance matrix of the first feature matrix to obtain a second feature matrix; determining the eigenvalues ​​and eigenvectors of each feature based on the second feature matrix to obtain multiple eigenvalues ​​and multiple eigenvectors; determining multiple cumulative contribution rates based on the multiple eigenvalues ​​and the multiple eigenvectors; determining at least one fusion feature based on the multiple cumulative contribution rates; obtaining a preset kernel matrix; and determining the current feature matrix based on the at least one fusion feature and the kernel matrix.

[0063] The feature extraction module determines the covariance matrix of the first feature matrix according to the following formula to obtain a second feature matrix:

[0064]

[0065] in, is the second characteristic matrix, Φ is the mapping function, and d is the first dimension.

[0066] The feature extraction module determines the eigenvalue and eigenvector of each feature according to the second feature matrix, and the formula for obtaining multiple eigenvalues ​​and multiple eigenvectors is:

[0067]

[0068]

[0069] Where k = 1, 2, …, d, λ is the eigenvalue of each feature, and ν is the eigenvector of each feature.

[0070] The feature extraction module determines the multiple cumulative contribution rates according to the following formula:

[0071]

[0072] Among them, λ i is the ith eigenvalue, and c is the cumulative contribution rate.

[0073] The determining of at least one fused feature according to the multiple cumulative contribution rates may specifically be determining the first t features with cumulative contribution rates greater than 80% as fused features to obtain the at least one fused feature.

[0074] The core matrix may be, for example:

[0075] K=Φ(x i )·Φ(x j );

[0076] Wherein, K is the kernel matrix.

[0077] Wherein, the formula for determining the current feature matrix according to the at least one fusion feature and the kernel matrix is:

[0078] X = K·α′;

[0079] Where X is the current feature matrix, α′ is the extracted t fusion features α1′,…,α t ′.

[0080] Among them, see Figure 6 , Figure 6 FIG. 1 is a flow chart of another transformer operation status detection method provided in an embodiment of the present application. Figure 6 As shown, in the transformer operating state detection method, the data processing operation performed by the feature extraction module includes the following steps:

[0081] Step S601, extracting features from the current running data to obtain an initial feature matrix.

[0082] Step S602: Map the initial feature matrix to a first dimension to obtain a first feature matrix.

[0083] Step S603: determine the covariance matrix of the first characteristic matrix to obtain a second characteristic matrix.

[0084] Step S604: determine the eigenvalue and eigenvector of each feature according to the second feature matrix to obtain multiple eigenvalues ​​and multiple eigenvectors.

[0085] Step S605: determining a plurality of cumulative contribution rates according to the plurality of eigenvalues ​​and the plurality of eigenvectors.

[0086] Step S606: determining at least one fusion feature according to the multiple cumulative contribution rates.

[0087] Step S607, obtaining a preset kernel matrix.

[0088] Step S608: determining the current feature matrix according to the at least one fused feature and the kernel matrix.

[0089] It can be seen that in this example, by determining the eigenvalue and eigenvector of each feature in the high-dimensional feature matrix, the features with higher cumulative contribution rates are determined as fusion features, and their dimensions are reduced to obtain the current feature matrix, so as to determine the state assessment results based on the current feature matrix and the target state analysis model. This not only reduces the impact of redundant information on the output results of the target state analysis model, but also enables the target state analysis model to capture the forward and backward dependencies in the transformer's operating data, thereby improving the stability and accuracy of transformer assessment and fault diagnosis.

[0090] Step S503: determining a state assessment result according to the current feature matrix and a preset target state analysis model.

[0091] The target state analysis model is used to analyze the before-after dependencies of each key operating feature.

[0092] The status evaluation result is used to indicate the operating status of the transformer in the current operating cycle.

[0093] Wherein, in determining the state evaluation result according to the current feature matrix and the preset target state analysis model, the method further includes: sending the state evaluation result to the terminal device so that the terminal device displays the state evaluation result in a visualization module of the terminal device.

[0094] In a possible embodiment, the training method of the target state analysis model includes the following steps: acquiring the historical operation data, wherein the historical operation data is used to characterize the operation status of the transformer before the current processing cycle; determining the target operation data based on the historical operation data and a preset data enhancement model, wherein the data volume of the target operation data is greater than the data volume of the historical operation data; determining a target feature matrix based on the target operation data, wherein the target feature matrix is ​​used to characterize the key operating features of the transformer before the current processing cycle; determining training data based on the target feature matrix; and training the initial state analysis model based on the training data to obtain the target state analysis model.

[0095] Among them, the data preprocessing module obtains the historical operating data, and the specific steps of obtaining the historical operating data are the same as the steps of obtaining the current operating data mentioned above, which will not be repeated here; it is also possible to determine the operating data of the transformer obtained in the processing cycle before the current processing cycle as the historical operating data. For example, after the end of the current processing cycle, in the next processing cycle, the current operating data of the current processing cycle is determined as the historical operating data.

[0096] Wherein, the data enhancement module determines the target operation data according to the historical operation data and a preset data enhancement model.

[0097] Among them, the feature extraction module determines the target feature matrix according to the target operation data. The specific step of determining the target feature matrix according to the target operation data is the same as the step of determining the current feature matrix according to the current operation data, and will not be repeated here.

[0098] The fault diagnosis module includes the target state analysis model, and the target state analysis model is in the long short-term bidirectional memory neural network layer. Before the current processing cycle, the fault diagnosis module performs offline training locally based on the latest historical operation data to obtain the target state analysis model, so that the target state analysis model used in the current processing cycle is the latest version.

[0099] The fault diagnosis module determines the training data according to the target feature matrix, for example, 80% of the features in the target feature matrix are determined as the training set, and 20% of the features are determined as the test set.

[0100] Wherein, the initial state analysis model includes a long short-term bidirectional memory neural network model, and the fault diagnosis module trains the initial state analysis model according to the training data to obtain the target state analysis model, including: S1. Initializing the model parameters of the optimization model; S2. Determining the activation values ​​of the input gate, the forgetting gate, and the output gate according to the training data; S3. Updating the cell state and the hidden state of the long short-term bidirectional memory neural network model according to the activation values ​​of the input gate, the forgetting gate, and the output gate; S4. Determining the loss function value according to the hidden state of the long short-term bidirectional memory neural network model and a preset loss function; S5. Update the model parameters of the optimization model according to the loss function value; Repeat S2-S5 until the loss function converges or reaches a preset maximum number of iterations.

[0101] The loss function can be set as:

[0102]

[0103] Among them, L is the loss function value, y i is the actual operating status of the transformer, Diagnose the running status of the model.

[0104] The optimization model may be, for example, an Adam optimizer, and the formula for updating the model parameters of the optimization model according to the loss function value is:

[0105]

[0106] Among them, θ is the model parameter, α is the learning rate, is the gradient of the loss function with respect to the parameters.

[0107] It can be seen that in this example, the target feature matrix characterizing the key operating characteristics of the transformer before the current processing cycle is determined based on the historical operating data of the transformer, and then the initial state analysis model is trained based on the target feature matrix to obtain a target analysis state model adapted to the transformer, and then the state evaluation result is determined based on the current feature matrix and the target state analysis model. This not only reduces the impact of redundant information on the output results of the target state analysis model, but also enables the target state analysis model to capture the forward and backward dependencies in the operating data of the transformer, thereby improving the stability and accuracy of transformer evaluation and fault diagnosis.

[0108] In a possible embodiment, the data enhancement model includes a generator and a discriminator, the generator is used to generate virtual operating data, the discriminator is used to output an identification result, the identification result is used to indicate whether the virtual operating data generated by the generator is real operating data, the identification result is yes or no, and the target operating data is determined according to the historical operating data and the preset data enhancement model, including: performing at least one expansion and judgment operation until the data volume of the target operating data is equal to the preset data volume, the expansion and judgment operation includes the following steps: performing at least one expansion data operation until the identification result is yes, the expansion data operation includes the following steps: obtaining a noise vector; inputting the noise vector into the generator to obtain initial virtual operating data; inputting the initial virtual operating data and the historical operating data into the discriminator so that the discriminator outputs the identification result; when the identification result is no, performing the next expansion data operation; adding the virtual operating data to the historical operating data to obtain the target operating data; judging whether the data volume of the target operating data is equal to the preset data volume; if not, performing the next expansion and judgment operation.

[0109] The data enhancement model is stored in the data enhancement module, and the data enhancement model may specifically be a generative adversarial network.

[0110] The noise vector is randomly generated, the three-layer neural network of the generator generates initial virtual operation data similar to the distribution of real operation data, hidden layer 1 and hidden layer 2 in the three-layer neural network of the generator use ReLU activation function, output layer 1 uses Tanh activation function, and the dimension of the generated initial virtual operation data is equal to the dimension of the real data. The discriminator receives the initial virtual operation data output by the generator as input, and performs binary classification through the three-layer neural network of the discriminator, hidden layer 3 and hidden layer 4 use Leaky ReLU activation function, and output layer 2 uses Sigmoid activation function.

[0111] Among them, the optimization objectives of the generative adversarial network are as follows:

[0112]

[0113] Where D is the generator, G is the discriminator, V(D,G) represents the optimization goal of the generative adversarial network, is the mathematical expectation of the value in brackets, indicating that x′ obeys the actual running data P data The distribution of Indicates that z obeys the noise vector P zThe optimization goal of the discriminator is to maximize logD(x'), and the optimization goal of the generator is to minimize 1-D(G(z). In this way, the discriminator and the generator form a dynamic game process, and the final result is to reach a Nash equilibrium. The generator can generate data that is indistinguishable from the real running data. At this time, the discriminator cannot distinguish between the initial virtual running data and the real running data.

[0114] Among them, see Figure 7 , Figure 7 Schematic diagram of a data enhancement model provided in an embodiment of the present application. Figure 7 As shown, in the data enhancement model 700, the noise vector z is input into the generator G, and the initial virtual operation data G(z) is output through the first hidden layer 1, the second hidden layer 2 and the first output layer 3 in the generator G. The initial virtual operation data G(z) and the historical operation data x are input into the discriminator D, and the identification result is output through the third hidden layer 4, the fourth hidden layer 5 and the second output layer 6 of the discriminator D. When the identification result is yes, the virtual operation data is added to the historical operation data to obtain the target operation data. When the identification result is no, the noise vector z is regenerated.

[0115] It can be seen that in this example, the historical operating data is expanded through the data enhancement model to obtain target operating data with a larger amount of data, which can solve the problem of lack of operating data of transformer failures in the training data, and improve the accuracy of the trained target state analysis model in transformer operating state detection, thereby improving the stability and accuracy of transformer evaluation and fault diagnosis.

[0116] In a possible embodiment, determining the state assessment result based on the current feature matrix and a preset target state analysis model includes: inputting the current feature matrix into the target state analysis model to obtain a target analysis result; and determining the state assessment result based on the target analysis result and a preset state assessment model.

[0117] Among them, the target state analysis model can specifically be a bidirectional long short-term memory neural network, and the target analysis result can specifically be the final hidden state output by the bidirectional long short-term memory neural network, and the final hidden state is the concatenation of the forward hidden state and the backward hidden state.

[0118] Among them, see Figure 8 , Figure 8 Schematic diagram of a target state classification model provided in an embodiment of the present application. Figure 8As shown, the target state classification model 800 is a bidirectional long short-term memory neural network model, which includes an output layer, a forward LSTM, a backward LSTM, and an input layer. The input layer inputs the feature x in the current feature matrix, and after passing through the corresponding forward LSTM and backward LSTM, the output layer outputs the hidden state h.

[0119] Among them, the forward LSTM is calculated from time step 1 to time step T. For each time step t, the calculation formula of the forward LSTM is as follows:

[0120]

[0121] in, is the hidden state of the forward LSTM at time t, is the hidden state at the previous moment. The calculation formulas for the input gate, output gate, and forget gate of the forward LSTM unit are as follows:

[0122]

[0123] in, and are the weight matrices of the input gate, output gate, and forget gate respectively. and is the weight matrix of the hidden state, and They are the bias items of the input gate, output gate, and forget gate respectively.

[0124] Among them, the forward hidden state output by the forward LSTM unit is as follows:

[0125]

[0126] in, is the cell state at time step t, which can be calculated by the following formula:

[0127]

[0128] in, and is the weight matrix of candidate cell states, is the bias term.

[0129] Among them, the backward LSTM is calculated from time step T to time step 1. For each time step t, the calculation formula of the backward LSTM is as follows:

[0130]

[0131] The calculation formula of the backward LSTM is similar to that of the forward LSTM, except that the input order is reversed, and will not be repeated here.

[0132] Among them, in the bidirectional long short-term memory neural network model, the final hidden state is the concatenation of the forward hidden state and the backward hidden state, that is:

[0133]

[0134] It can be seen that in this example, the target analysis results are obtained by inputting the current feature matrix into the target state analysis model, and then the state assessment results are determined according to the target analysis structure and the preset state assessment model. This not only reduces the impact of redundant information on the output results of the target state analysis model, but also enables the target state analysis model to capture the forward and backward dependencies in the transformer's operating data, thereby improving the stability and accuracy of transformer evaluation and fault diagnosis.

[0135] In a possible embodiment, determining the state assessment result based on the target analysis result and a preset state assessment model includes: performing a linear transformation on the target analysis result to obtain target linear data; importing the target linear data into the state assessment model to obtain multiple target probabilities, wherein the multiple target probabilities correspond one-to-one to multiple states; and determining the state assessment result based on the multiple target probabilities.

[0136] Among them, the fully connected layer in the fault diagnosis module performs a linear transformation on the target analysis result to obtain target linear data. The formula for performing a linear transformation on the target analysis result to obtain the target linear data is:

[0137] z t =W y ·h t +b y ;

[0138] Among them, Z t is the target linear data, W y is the weight matrix, b y is the bias term.

[0139] Wherein, the classification function layer imports the target linear data into the state assessment model to obtain multiple target probabilities.

[0140] The state assessment model may specifically be:

[0141]

[0142] Among them, z t,i is the value corresponding to state i in the target linear data, K is the number of state labels, is the probability of the i-th state, and the sum of the probabilities corresponding to all states is 1.

[0143] Among them, determining the state evaluation result according to the multiple target probabilities can specifically be determining the state evaluation result according to the state corresponding to the largest probability among the multiple target probabilities. The state evaluation result includes three categories: normal, warning, and fault. The fault also includes overload, overvoltage, and internal insulation aging.

[0144] It can be seen that in this example, by performing a linear transformation on the target analysis results to obtain target linear data, multiple target probabilities are determined based on the target linear data, and then the state assessment results are determined based on the multiple target probabilities. This not only reduces the impact of redundant information on the output results of the target state analysis model, but also enables the target state analysis model to capture the forward and backward dependencies in the transformer's operating data, thereby improving the stability and accuracy of transformer assessment and fault diagnosis.

[0145] It can be seen that in this application, the server obtains the current operating data of the transformer, and the current operating data is used to characterize the operating status of the transformer in the current processing cycle; the current feature matrix is ​​determined based on the current operating data, and the current feature matrix is ​​used to characterize the key operating characteristics of the transformer in the current processing cycle; the state evaluation result is determined based on the current feature matrix and the preset target state analysis model, and the target state analysis model is used to analyze the front-end dependency of each key operating characteristic, and the state evaluation result is used to indicate the operating state of the transformer in the current operating cycle. Since the current feature matrix characterizing the key operating characteristics of the transformer in the current processing cycle is determined based on the current operating data, and then the state evaluation result is determined based on the current feature matrix and the target state analysis model, not only the influence of redundant information on the output result of the target state analysis model is reduced, but also the target state analysis model can capture the front-end dependency in the operating data of the transformer, thereby improving the stability and accuracy of transformer evaluation and fault diagnosis.

[0146] The above mainly introduces the scheme of the embodiment of the present application from the perspective of the execution process on the method side. It is understandable that, in order to realize the above functions, the controller includes a hardware structure and / or software module corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the units and algorithm steps of each example described in the embodiment provided herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present application.

[0147] In accordance with the above-mentioned embodiments, please refer to Fig. 9 , Fig. 9is a functional unit block diagram of a transformer operation status detection device provided by an embodiment of the present application, such as Fig. 9 As shown, the transformer operation status detection device 900 includes: a first receiving unit 901, used to obtain the current operation data of the transformer, and the current operation data is used to characterize the operation status of the transformer in the current processing cycle; a first processing unit 902, used to determine the current feature matrix according to the current operation data, and the current feature matrix is ​​used to characterize the key operation characteristics of the transformer in the current processing cycle; determine the state evaluation result according to the current feature matrix and a preset target state analysis model, and the target state analysis model is used to analyze the before-and-after dependencies of each key operation characteristic, and the state evaluation result is used to indicate the operation state of the transformer in the current operation cycle.

[0148] In a possible embodiment, in terms of determining the current feature matrix based on the current operating data, the first processing unit 902 is specifically used to: perform feature extraction on the current operating data to obtain an initial feature matrix, where the dimension of the initial feature matrix is ​​the initial dimension; map the initial feature matrix to a first dimension to obtain a first feature matrix, where the first dimension is higher than the initial dimension; determine the current feature matrix based on the first feature matrix, where the dimension of the current feature matrix is ​​lower than the first dimension.

[0149] In a possible embodiment, in terms of determining the current feature matrix based on the first feature matrix, the first processing unit 902 is specifically used to: determine the covariance matrix of the first feature matrix to obtain a second feature matrix; determine the eigenvalues ​​and eigenvectors of each feature based on the second feature matrix to obtain multiple eigenvalues ​​and multiple eigenvectors; determine multiple cumulative contribution rates based on the multiple eigenvalues ​​and the multiple eigenvectors; determine at least one fusion feature based on the multiple cumulative contribution rates; obtain a preset kernel matrix; and determine the current feature matrix based on the at least one fusion feature and the kernel matrix.

[0150] In a possible embodiment, in terms of the training method of the target state analysis model, the transformer operating state detection device 900 is further specifically used to: obtain the historical operating data, the historical operating data is used to characterize the operating status of the transformer before the current processing cycle; determine the target operating data according to the historical operating data and a preset data enhancement model, the data volume of the target operating data is greater than the data volume of the historical operating data; determine a target feature matrix according to the target operating data, the target feature matrix is ​​used to characterize the key operating features of the transformer before the current processing cycle; determine training data according to the target feature matrix; train the initial state analysis model according to the training data to obtain the target state analysis model.

[0151] In a possible embodiment, the data enhancement model includes a generator and a discriminator, the generator is used to generate virtual operating data, the discriminator is used to output an identification result, the identification result is used to indicate whether the virtual operating data generated by the generator is real operating data, and the identification result is yes or no. In terms of determining the target operating data based on the historical operating data and the preset data enhancement model, the first processing unit 902 is specifically used to: perform at least one expansion and judgment operation until the data volume of the target operating data is equal to the preset data volume, and the expansion and judgment operation includes the following steps: perform at least one expansion data operation until the identification result is yes, and the expansion data operation includes the following steps: obtain a noise vector; input the noise vector into the generator to obtain initial virtual operating data; input the initial virtual operating data and the historical operating data into the discriminator so that the discriminator outputs the identification result; when the identification result is no, perform the next expansion data operation; add the virtual operating data to the historical operating data to obtain the target operating data; judge whether the data volume of the target operating data is equal to the preset data volume; if not, perform the next expansion and judgment operation.

[0152] In a possible embodiment, in terms of determining the state assessment result based on the current feature matrix and a preset target state analysis model, the first processing unit 902 is specifically used to: input the current feature matrix into the target state analysis model to obtain a target analysis result; and determine the state assessment result based on the target analysis result and a preset state assessment model.

[0153] In a possible embodiment, in terms of determining the state assessment result based on the target analysis result and a preset state assessment model, the first processing unit 902 is specifically used to: perform a linear transformation on the target analysis result to obtain target linear data; import the target linear data into the state assessment model to obtain multiple target probabilities, and the multiple target probabilities correspond one-to-one to multiple states; determine the state assessment result based on the multiple target probabilities.

[0154] It can be understood that since the method embodiment and the device embodiment are different presentation forms of the same technical concept, the content of the method embodiment part in this application should be synchronously adapted to the device embodiment part and will not be repeated here.

[0155] In the case of an integrated unit, such as Fig.10 As shown, Fig.10 This is a functional unit block diagram of another transformer operating status detection device provided by an embodiment of the present application. Fig.10 In the embodiment, the transformer operation status detection device 900 includes: a processing module 1012 and a communication module 1011. The processing module 1012 is used to control and manage the actions of the transformer operation status detection device 900, for example, to execute the steps of the first receiving unit 901 and the first processing unit 902, and / or to execute other processes of the technology described in this article. The communication module 1011 is used to support the interaction between the transformer operation status detection device 900 and other devices. Fig.10 As shown, the transformer operation status detection device 900 may further include a storage module 1013 , and the storage module 1013 is used to store program codes and data of the transformer operation status detection device 900 .

[0156] Among them, the processing module 1012 can be a processor or a controller, for example, a central processing unit (CPU), a general processor, a digital signal processor (DSP), an ASIC, an FPGA or other programmable logic device, a transistor logic device, a hardware component or any combination thereof. It can implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of this application. The processor can also be a combination that implements a computing function, for example, a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like. The communication module 1011 can be a transceiver, an RF circuit or a communication interface, and the like. The storage module 1013 can be a memory.

[0157] Among them, all relevant contents of each scenario involved in the above method embodiment can be referred to the functional description of the corresponding functional module, and will not be repeated here. The above transformer operation status detection device 900 can execute the above Figure 5 The transformer operating status detection method is shown.

[0158] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0159] Fig.11 is a structural block diagram of a server provided in an embodiment of the present application. Fig.11 As shown, the server 120 may include one or more of the following components: a processor 410, and a memory 420 coupled to the processor 410, wherein the memory 420 may store one or more computer programs 421, and the one or more computer programs 421 may be configured to implement the methods described in the above embodiments when executed by one or more processors 410.

[0160] The processor 410 may include one or more processing cores. The processor 410 uses various interfaces and lines to connect the various parts of the entire server 120, and executes various functions and processes data of the server 120 by running or executing instructions, programs, code sets or instruction sets stored in the memory 420, and calling data stored in the memory 420. Optionally, the processor 410 can be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 410 can integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem is used to process wireless communications. It is understandable that the above-mentioned modem may not be integrated into the processor 410, but may be implemented separately through a communication chip.

[0161] The memory 420 may include a random access memory (RAM) or a read-only memory (ROM). The memory 420 may be used to store instructions, programs, codes, code sets or instruction sets. The memory 420 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc. The data storage area may also store data created by the server 120 during use, etc.

[0162] It is understandable that the server 120 may include more or fewer structural elements than those in the above structural block diagram, which is not limited here. The present application embodiment provides a computer-readable storage medium on which a computer program / instruction is stored, and when the computer program / instruction is executed by a processor, the steps of the method described in any possible embodiment are implemented.

[0163] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0164] In the several embodiments provided in the present application, it should be understood that the disclosed methods, devices and systems can be implemented in other ways. For example, the device embodiments described above are merely schematic; for example, the division of the unit is only a logical function division, and there may be other division methods in actual implementation; for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0165] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0166] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may be physically included separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0167] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform some steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a volatile memory or a non-volatile memory. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM) or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (double data rate SDRAM, DDR SDRAM), enhanced synchronous dynamic random access memory (enhanced SDRAM, ESDRAM), synchronous link dynamic random access memory (synchlink DRAM, SLDRAM) and direct memory bus random access memory (directrambus RAM, DR RAM) and other media that can store program code.

[0168] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions without departing from the spirit and scope of the present invention, and can make various changes and modifications, including the combination of the above-mentioned different functions and implementation steps, including software and hardware implementation methods, all of which are within the scope of protection of the present invention.

Claims

1. A transformer operating status detection method, characterized in that: A server applied to a transformer operation status detection system, the transformer operation status detection system comprising a transformer and the server, the method comprising: Acquiring current operating data of the transformer, wherein the current operating data is used to characterize the operating status of the transformer in a current processing cycle; Determine a current feature matrix according to the current operation data, wherein the current feature matrix is ​​used to characterize key operation characteristics of the transformer in the current processing cycle; A state evaluation result is determined according to the current feature matrix and a preset target state analysis model, wherein the target state analysis model is used to analyze the before-after dependencies of each key operating feature, and the state evaluation result is used to indicate the operating state of the transformer in the current operating cycle.

2. The method according to claim 1, characterized in that The training method of the target state analysis model comprises the following steps: Acquire the historical operation data, where the historical operation data is used to characterize the operation status of the transformer before the current processing cycle; Determine target operating data according to the historical operating data and a preset data enhancement model, wherein the data volume of the target operating data is greater than the data volume of the historical operating data; Determine a target feature matrix according to the target operation data, wherein the target feature matrix is ​​used to characterize key operation features of the transformer before the current processing cycle; Determining training data according to the target feature matrix; The initial state analysis model is trained according to the training data to obtain the target state analysis model.

3. The method according to claim 2, characterized in that The data enhancement model includes a generator and a discriminator, wherein the generator is used to generate virtual operation data, and the discriminator is used to output a discrimination result, wherein the discrimination result is used to indicate whether the virtual operation data generated by the generator is real operation data, and the discrimination result is yes or no. The target operation data is determined according to the historical operation data and the preset data enhancement model, including: Perform at least one expansion and judgment operation until the data volume of the target operation data is equal to the preset data volume, and the expansion and judgment operation includes the following steps: Perform at least one extended data operation until the identification result is yes, the extended data operation comprising the following steps: obtaining a noise vector; inputting the noise vector into the generator to obtain initial virtual operation data; inputting the initial virtual operation data and the historical operation data into the discriminator so that the discriminator outputs the identification result; when the identification result is no, perform the next extended data operation; Adding the virtual operation data to the historical operation data to obtain the target operation data; Determining whether the data volume of the target operation data is equal to the preset data volume; If not, the next expansion and judgment operation is performed.

4. The method according to claim 3, characterized in that The determining the current feature matrix according to the current operation data includes: Performing feature extraction on the current running data to obtain an initial feature matrix, where the dimension of the initial feature matrix is ​​the initial dimension; Mapping the initial feature matrix to a first dimension to obtain a first feature matrix, wherein the first dimension is higher than the initial dimension; The current feature matrix is ​​determined according to the first feature matrix, and the dimension of the current feature matrix is ​​lower than the first dimension.

5. The method according to claim 4, characterized in that The determining the current feature matrix according to the first feature matrix includes: Determine the covariance matrix of the first feature matrix to obtain a second feature matrix; Determine the eigenvalue and eigenvector of each feature according to the second feature matrix to obtain multiple eigenvalues ​​and multiple eigenvectors; Determine a plurality of cumulative contribution rates according to the plurality of eigenvalues ​​and the plurality of eigenvectors; determining at least one fusion feature according to the multiple cumulative contribution rates; Get the preset kernel matrix; The current feature matrix is ​​determined according to the at least one fused feature and the kernel matrix.

6. The method according to claim 5, characterized in that The determining of the state assessment result according to the current feature matrix and the preset target state analysis model includes: Inputting the current feature matrix into the target state analysis model to obtain a target analysis result; The state assessment result is determined according to the target analysis result and a preset state assessment model.

7. The method according to claim 6, characterized in that Determining the state assessment result according to the target analysis result and a preset state assessment model includes: Performing a linear transformation on the target analysis result to obtain target linear data; Importing the target linear data into the state assessment model to obtain a plurality of target probabilities, wherein the plurality of target probabilities correspond one to one to a plurality of states; The state assessment result is determined according to the multiple target probabilities.

8. A transformer operating status detection device, characterized in that: A server applied to a transformer operation status detection system, the transformer operation status detection system comprising a transformer and the server, the device comprising: A first receiving unit, used for acquiring current operation data of the transformer, wherein the current operation data is used for representing the operation status of the transformer in a current processing cycle; A first processing unit is used to determine a current feature matrix based on the current operating data, wherein the current feature matrix is ​​used to characterize key operating features of the transformer in the current processing cycle; determine a state evaluation result based on the current feature matrix and a preset target state analysis model, wherein the target state analysis model is used to analyze the before-and-after dependencies of each key operating feature, and the state evaluation result is used to indicate the operating state of the transformer in the current operating cycle.

9. A server, characterized in that: The method comprises a processor, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the programs include instructions for executing the steps in the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: A computer program / instruction is stored thereon, and when the computer program / instruction is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images