A small sample multi-modal electronic voltage transformer fault diagnosis method and system

Abstract

The present invention provides a small sample multimodal electronic voltage transformer fault diagnosis method and system, which relates to the field of fault diagnosis. The method comprises the following steps: acquiring multimodal data and related data; constructing a fault diagnosis model, including a feature extraction unit for modal feature extraction, to obtain corresponding modal features; and connecting a secondary voltage signal and all modal features and converting them into a two-dimensional image to obtain a support set and input it into an embedding network unit to obtain an embedded support set and a local descriptor of the embedded support set; acquiring multimodal data to be detected and inputting it into the fault diagnosis model to obtain an embedded query set and a local descriptor of the embedded query set; and calculating a normal similarity score and an abnormal similarity score, comparing the scores of the normal similarity score and the abnormal similarity score, to obtain a fault diagnosis result.

Description

Technical Field

[0001] The present application relates to the field of fault diagnosis, and primarily to a small sample multi-modal electronic voltage transformer fault diagnosis method and system. Background Art

[0002] Electronic voltage transformers (EVTs) are devices designed to replace traditional voltage transformers and are widely used in power system measurement, protection, and monitoring. Unlike traditional VTs, EVTs utilize digital and intelligent technologies to sample, process, and transmit voltage. Their output is typically a digital signal, offering greater accuracy, stability, and interference immunity. Due to the complexity of power systems and the high degree of integration of electronic equipment, fault diagnosis of EVTs is challenging.

[0003] At present, the fault diagnosis methods for electronic voltage transformers mainly include fault diagnosis based on signal analysis, but this method has low accuracy for complex nonlinear faults and situations where multiple faults are superimposed; fault diagnosis based on artificial intelligence has high requirements for the quality and quantity of training data, and the model has poor adaptability when the system environment undergoes major changes; traditional methods based on signal analysis have limited feature extraction capabilities in scenarios where nonlinear faults and multiple faults are superimposed, resulting in reduced diagnostic accuracy, and insufficient detection sensitivity for weak faults.

[0004] The Chinese invention patent application with publication number CN119249341A discloses a fault diagnosis method based on multimodal and multiscale feature fusion, which specifically discloses "collecting the signal source of the aircraft engine bearing at different positions when a fault occurs by a sensor, the signal source includes but is not limited to: vibration signal, sound signal, temperature, humidity, current; performing unified denoising on the vibration signal to obtain an initial one-dimensional signal X0; performing one-dimensional feature extraction on the initial one-dimensional signal X0 through the MCNN module; performing image conversion on the initial one-dimensional signal X0 to obtain an initial two-dimensional image T1, performing two-dimensional feature extraction on the initial two-dimensional image T1; fusing the extracted one-dimensional features with the two-dimensional features to generate a multimodal and multiscale feature fusion. The fused multimodal feature H1 is input into the LSTM network for fault diagnosis and classification. Although this method collects multi-source signals, the subsequent processing only focuses on the vibration signal, and other modal data are not explicitly fused and utilized, resulting in a lack of feature dimensions. It is difficult to fully reflect the multi-physical field coupling characteristics of bearing faults, which affects the accuracy of diagnosis. In addition, the one-dimensional and two-dimensional features in this method are only simply spliced ​​and fused, lacking cross-modal correlation analysis, resulting in the complementarity of different modal information not being fully explored, and the discriminative ability of the fused features is limited. At the same time, although the LSTM network of this method is good at processing time series data, it has a weak ability to classify the fused high-dimensional features, and is prone to overfitting or underfitting problems. Summary of the Invention

[0005] In order to solve the above-mentioned problems existing in the prior art, the present application provides a small sample multi-modal electronic voltage transformer fault diagnosis method and system.

[0006] The technical solution of this application is as follows:

[0007] In one aspect, the present invention provides a small sample multi-modal electronic voltage transformer fault diagnosis method, the method comprising:

[0008] Acquiring multimodal data and related data of an electronic voltage transformer, wherein the multimodal data includes grid frequency, secondary voltage signal, and primary current signal; and the related data includes power data;

[0009] Construct a fault diagnosis model, including a feature extraction unit and an embedding network unit; the feature extraction unit is used to extract modal features from multimodal data to obtain corresponding modal features; the secondary voltage signal and the modal features are connected and converted into a two-dimensional image to obtain a support set; the support set is input into the embedding network unit to obtain an embedded support set and a local descriptor of the embedded support set; based on the embedded support set, the local descriptor is divided into a local descriptor of a normal support set and a local descriptor of an abnormal support set; wherein the modal feature extraction specifically comprises inputting the grid frequency into an improved long short-term memory network (LSTM) network, performing fast Fourier frequency domain analysis on the secondary voltage signal, and performing overshoot calculation on the primary current signal;

[0010] Acquire multimodal data to be detected of a target electronic voltage transformer, input the multimodal data to be detected into the fault diagnosis model, and obtain an embedded query set and a local descriptor of the embedded query set;

[0011] Calculate the similarity between the local descriptors embedded in the query set and the local descriptors of the normal support set to obtain a normal similarity score; calculate the similarity between the local descriptors embedded in the query set and the local descriptors of the abnormal support set to obtain an abnormal similarity score; compare the normal similarity score and the abnormal similarity score to obtain the fault diagnosis result of the target detection electronic voltage transformer;

[0012] Calculate the normal similarity score and abnormal similarity score, specifically:

[0013] The embedded query set is represented as ,in Represents an embedded query set; A local descriptor representing the embedded query set; Indicates the A local descriptor that embeds the query set; Indicates the number of local descriptors embedded in the query set;

[0014] Traverse the local descriptors in the embedded query set, search for the nearest neighbors in the local descriptors of the normal support set and the local descriptors of the abnormal support set, and calculate the normal similarity score and the abnormal similarity score, which can be expressed as follows:

[0015] ;

[0016] ;

[0017] Where, represents the normal similarity score; Indicates the weight is Similarity measurement function of ; Indicates the The local descriptor of the embedded query set is The value of the dimension; Indicates the The local descriptor of the normal support set is The value of the dimension; represents the anomaly similarity score; Indicates the The local descriptor of the abnormal support set is The value of the dimension; Represents the dimension of the embedded query set; Indicates the The index value of the embedded query set dimension; Indicates a preset constant;

[0018] like , the target electronic voltage transformer is normal; otherwise, the target electronic voltage transformer is abnormal.

[0019] Preferably, the method further comprises performing data cleaning on the multimodal data and related data, wherein the data cleaning comprises processing missing values, outliers and standardizing data formats.

[0020] Preferably, the grid frequency is input into the improved LSTM network and is expressed as follows:

[0021] ;

[0022] ;

[0023] ;

[0024] ;

[0025] ;

[0026] ;

[0027] Where, express Always forget the output of the gate; represents a logical function; Represents the weight of the forget gate; express The hidden state of the moment; express The grid frequency at the moment; express The primary current signal at the moment; Represents the bias of the forget gate; express The output of the input gate at all times; Represents the weight of the input gate; represents the bias of the input gate; express Candidate cell states at time instant; represents the weight of the candidate cell state; represents the bias of the candidate cell state; represents the first activation function; express The cell state at a moment in time; express The cell state at a moment in time; express Output gate at the moment; Represents the weight of the output gate; Represents the bias of the output gate; express Power of the moment; Represents the modal characteristics of the power grid frequency; Indicates time.

[0028] Preferably, the secondary voltage signal is subjected to fast Fourier frequency domain analysis, which can be expressed as follows:

[0029] ;

[0030] ;

[0031] Where, represents the fast Fourier function; Indicates the preset frequency; Indicates the sign of the imaginary part; express Secondary voltage signal at time ; represents the real part function; represents the imaginary part function;

[0032] Choose the largest As the modal characteristic of the secondary voltage signal, it is expressed as .

[0033] Preferably, the overshoot calculation is performed on the primary current signal, which is expressed as follows:

[0034] , ;

[0035] and ;

[0036] ;

[0037] ;

[0038] Where, Indicates the average value of the primary current signal; Indicates the maximum value of the local primary current signal; Indicates the total number of primary current signals; Indicates the The index value of a primary current signal; Indicates the acquisition period of the primary current signal; Indicates the A primary current signal; Indicates the The maximum value of a local primary current signal; Indicates the The index value of the maximum value of the local primary current signal; Indicates the total number of maximum values ​​of local primary current signals; Indicates the The overshoot value of the maximum value of the local primary current signal; Represents the modal characteristics of the primary current signal; Indicates the The current signal before the maximum value of a local primary current signal; Indicates the The current signal after the maximum value of the local primary current signal.

[0039] Preferably, all modal features and secondary voltage signals are connected and converted into a two-dimensional image to obtain a support set, which is expressed as:

[0040] , ;

[0041] ;

[0042] Where, represents the two-dimensional image matrix, i.e., the support set; Represents a two-dimensional image matrix Rank Column pixels; Represents the dimension of the two-dimensional image matrix; Represents a connection sequence; Indicates the A connection sequence value; Indicates the The index value of the connected sequence; Indicates the Secondary voltage signal value; Indicates the number of secondary voltage signals; Represents a two-dimensional image matrix The index value of the row; Represents a two-dimensional image matrix The index value of the row; Indicates the The index value of the secondary voltage signal value.

[0043] Preferably, the support set is input into the embedding network unit, and the embedding network unit includes a Gaussian noise layer, a layer normalization layer, a convolution layer and a Dropout layer, specifically:

[0044] The Gaussian noise layer receives the support set and adds noise, which is expressed as:

[0045] ;

[0046] Where, represents the noise support set; Indicates Select the Gaussian distribution of noise value; represents the variance of the Gaussian distribution;

[0047] The layer normalization layer receives the noise support set and performs layer normalization to obtain an embedded support set, which is expressed as:

[0048] ;

[0049] ;

[0050] ;

[0051] Where, represents the mean of the noise support set; represents the variance of the noise support set; The number of two-dimensional images representing the noise support set; Indicates the The index value of a two-dimensional image; represents the noise support set A two-dimensional image of represents the embedding support set; Indicates a preset constant; Indicates the preset scaling parameters; Indicates the preset translation parameters;

[0052] The convolution layer convolves the embedded support set and further inputs the Dropout layer for regularization, which can be expressed as follows:

[0053] ;

[0054] ⊙ ;

[0055] Where, Indicates the feature maps, which are the outputs of the convolutional layers; Indicates the The index value of the neuron; represents the total number of neurons; represents the second activation function; Indicates the Tier The connection weights of neurons; Indicates the Tier The bias of each neuron; Represents the embedding support set A two-dimensional image; Indicates the The index value of the convolution layer; Indicates the number of convolutional layers; Indicates the A regularized feature map; Represents the preset Bernoulli mask matrix; ⊙ represents element-by-element multiplication;

[0056] Based on the regularized feature map, the local descriptor of the embedded support set is obtained, which is expressed as:

[0057] ;

[0058] ;

[0059] Where, local descriptors representing the embedded support set; Indicates the A local descriptor embedded in the support set; represents the weight of the fully connected layer; represents the bias of the fully connected layer; Indicates the The feature vector corresponding to the regularized feature map;

[0060] Divide the local descriptors of the embedded support set into local descriptors of the normal support set and local descriptors of the exception support set .

[0061] Preferably, the method also includes continuously monitoring modal characteristics, collecting diagnostic experience of each mode, and establishing a modal fault association knowledge base; evaluating the diagnostic results using preset performance evaluation indicators to obtain evaluation results; and optimizing the fault diagnosis process in combination with the modal fault association knowledge base and the evaluation results.

[0062] On the other hand, the present invention also proposes a small sample multi-modal electronic voltage transformer fault diagnosis system, which includes a data acquisition module, a local descriptor generation module, a fault diagnosis module and a result output module, wherein:

[0063] The data acquisition module is used to acquire multimodal data and related data of the electronic voltage transformer, the multimodal data including grid frequency, secondary voltage signal and primary current signal; the related data including power data; and transmit the multimodal data and related data to the local descriptor generation module;

[0064] The local descriptor generation module is used to construct a fault diagnosis model, including a feature extraction unit and an embedding network unit; the feature extraction unit is used to extract modal features from multimodal data to obtain corresponding modal features; the secondary voltage signal and the modal features are connected and converted into a two-dimensional image to obtain a support set; the support set is input into the embedding network unit to obtain an embedded support set and a local descriptor of the embedded support set; based on the embedded support set, the local descriptor is divided into a local descriptor of a normal support set and a local descriptor of an abnormal support set; wherein, the modal feature extraction is specifically to input the grid frequency into an improved long short-term memory network LSTM network, perform fast Fourier frequency domain analysis on the secondary voltage signal, and perform overshoot calculation on the primary current signal;

[0065] The fault diagnosis module is used to obtain multimodal data to be detected of the target electronic voltage transformer, input the multimodal data to be detected into the fault diagnosis model, and obtain an embedded query set and a local descriptor of the embedded query set;

[0066] Calculate the similarity between the local descriptors embedded in the query set and the local descriptors of the normal support set to obtain a normal similarity score; calculate the similarity between the local descriptors embedded in the query set and the local descriptors of the abnormal support set to obtain an abnormal similarity score; compare the normal similarity score and the abnormal similarity score to obtain the fault diagnosis result of the target detection electronic voltage transformer;

[0067] Calculate the normal similarity score and abnormal similarity score, specifically:

[0068] The embedded query set is represented as ,in Represents an embedded query set; A local descriptor representing the embedded query set; Indicates the A local descriptor that embeds the query set; Indicates the number of local descriptors embedded in the query set;

[0069] Traverse the local descriptors in the embedded query set, search for the nearest neighbors in the local descriptors of the normal support set and the local descriptors of the abnormal support set, and calculate the normal similarity score and the abnormal similarity score, which can be expressed as follows:

[0070] ;

[0071] ;

[0072] Where, represents the normal similarity score; Indicates the weight is Similarity measurement function of ; Indicates the The local descriptor of the embedded query set is The value of the dimension; Indicates the The local descriptor of the normal support set is The value of the dimension; represents the anomaly similarity score; Indicates the The local descriptor of the abnormal support set is The value of the dimension; Represents the dimension of the embedded query set; Indicates the The index value of the embedded query set dimension; Indicates a preset constant;

[0073] like , the target electronic voltage transformer is normal; otherwise, the target electronic voltage transformer is abnormal;

[0074] The result output module is used to display the fault diagnosis result of the target detection electronic voltage transformer.

[0075] Compared with the prior art, the present invention has the following beneficial effects:

[0076] 1) This invention provides a small-sample multi-modal electronic voltage transformer fault diagnosis method and system. By combining grid frequency, secondary voltage signals, and primary current signals, it improves the multi-dimensional characterization capability of fault characteristics and enhances the accuracy of identifying multi-physical field coupling faults within the transformer.

[0077] 2) This invention provides a small-sample multimodal electronic voltage transformer fault diagnosis method and system. The improved LSTM network enhances the ability to capture time-varying features in power grid frequency signals and increases detection sensitivity. The embedded network unit enhances the model's anti-interference ability through Gaussian noise layers and Dropout layers. Combined with layer normalization technology, it improves diagnostic robustness and enhances the model's adaptability.

[0078] 3) The present invention provides a small-sample multimodal electronic voltage transformer fault diagnosis method and system. By supporting set partitioning and local descriptor similarity measurement, fault diagnosis is transformed into a traceable feature matching process, which improves the transparency of decision-making basis and enhances the model's generalization ability for rare fault types. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 It is a flow chart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0080] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0081] The present invention provides the following technical solution: a small sample multi-modal electronic voltage transformer fault diagnosis method and system.

[0082] Example 1:

[0083] See Figure 1 This embodiment provides a small sample multi-mode electronic voltage transformer fault diagnosis method, the specific steps include:

[0084] S1. Acquire multimodal data and related data of an electronic voltage transformer, wherein the multimodal data includes grid frequency, secondary voltage signal, and primary current signal; and the related data includes power data;

[0085] The method further includes performing data cleaning on the multimodal data, wherein the data cleaning includes processing missing values, outliers, and standardizing data formats;

[0086] S2. Construct a fault diagnosis model, including a feature extraction unit and an embedding network unit;

[0087] S21, the feature extraction unit is used to extract modal features from the multimodal data to obtain corresponding modal features; wherein the modal feature extraction specifically comprises inputting the grid frequency into an improved long short-term memory network (LSTM) network, performing fast Fourier frequency domain analysis on the secondary voltage signal, and performing overshoot calculation on the primary current signal;

[0088] S211. Input the grid frequency into the improved LSTM network. The improved LSTM network specifically imports the primary current signal and power data, which is expressed as follows:

[0089] ;

[0090] ;

[0091] ;

[0092] ;

[0093] ;

[0094] ;

[0095] Where, express Always forget the output of the gate; represents a logical function; Represents the weight of the forget gate; express The hidden state of the moment; express The grid frequency at the moment; express The primary current signal at the moment; Represents the bias of the forget gate; express The output of the input gate at all times; Represents the weight of the input gate; represents the bias of the input gate; express Candidate cell states at time instant; represents the weight of the candidate cell state; represents the bias of the candidate cell state; represents the first activation function; express The cell state at a moment in time; express The cell state at a moment in time; express Output gate at the moment; Represents the weight of the output gate; Represents the bias of the output gate; express Power of the moment; Represents the modal characteristics of the power grid frequency; Indicates the moment;

[0096] S212, performing fast Fourier frequency domain analysis on the secondary voltage signal, which is expressed as:

[0097] ;

[0098] ;

[0099] Where, represents the fast Fourier function; Indicates the preset frequency; Indicates the sign of the imaginary part; express Secondary voltage signal at time ; represents the real part function; represents the imaginary part function;

[0100] Choose the largest As the modal characteristic of the secondary voltage signal, it is expressed as ;

[0101] S213, performing overshoot calculation on the primary current signal, which is expressed as:

[0102] , ;

[0103] and ;

[0104] ;

[0105] ;

[0106] Where, Indicates the average value of the primary current signal; Indicates the maximum value of the local primary current signal; Indicates the total number of primary current signals; Indicates the The index value of a primary current signal; Indicates the acquisition period of the primary current signal; Indicates the A primary current signal; Indicates the The maximum value of a local primary current signal; Indicates the The index value of the maximum value of the local primary current signal; Indicates the total number of maximum values ​​of local primary current signals; Indicates the The overshoot value of the maximum value of the local primary current signal; Represents the modal characteristics of the primary current signal; Indicates the The current signal before the maximum value of a local primary current signal; Indicates the A current signal after the maximum value of a local primary current signal;

[0107] S22. Connect the secondary voltage signal and all modal features and convert them into a two-dimensional image to obtain the support set, which can be expressed as:

[0108] , ;

[0109] ;

[0110] Where, represents the two-dimensional image matrix, i.e., the support set; Represents a two-dimensional image matrix Rank Column pixels; Represents the dimension of the two-dimensional image matrix; Represents a connection sequence; Indicates the A connection sequence value; Indicates the The index value of the connected sequence; Indicates the Secondary voltage signal value; Indicates the number of secondary voltage signals; Represents a two-dimensional image matrix The index value of the row; Represents a two-dimensional image matrix The index value of the row; Indicates the The index value of the secondary voltage signal value;

[0111] S23, inputting the support set into the embedding network unit to obtain an embedded support set and a local descriptor of the embedded support set; wherein the embedding network unit includes a Gaussian noise layer, a layer normalization layer, a convolution layer and a Dropout layer;

[0112] The Gaussian noise layer receives the support set and adds noise, which is expressed as:

[0113] ;

[0114] Where, represents the noise support set; Indicates Select the Gaussian distribution of noise value; represents the variance of the Gaussian distribution;

[0115] The layer normalization layer receives the noise support set and performs layer normalization, which can be expressed as:

[0116] ;

[0117] ;

[0118] ;

[0119] Where, represents the mean of the noise support set; represents the variance of the noise support set; The number of two-dimensional images representing the noise support set; Indicates the The index value of a two-dimensional image; represents the noise support set A two-dimensional image of represents the embedding support set; Indicates a preset constant; Indicates the preset scaling parameters; Indicates the preset translation parameters;

[0120] The convolution layer convolves the embedded support set and further inputs the Dropout layer for regularization, which can be expressed as follows:

[0121] ;

[0122] ⊙ ;

[0123] Where, Indicates the feature maps, which are the outputs of the convolutional layers; Indicates the The index value of the neuron; represents the total number of neurons; represents the second activation function; Indicates the Tier The connection weights of neurons; Indicates the Tier The bias of each neuron; Represents the embedding support set A two-dimensional image; Indicates the The index value of the convolution layer; Indicates the number of convolutional layers; Indicates the A regularized feature map; Represents the preset Bernoulli mask matrix; ⊙ represents element-by-element multiplication;

[0124] Based on the regularized feature map, the local descriptor of the embedded support set is obtained, which is expressed as:

[0125] ;

[0126] ;

[0127] Where, local descriptors representing the embedded support set; Indicates the A local descriptor embedded in the support set; represents the weight of the fully connected layer; represents the bias of the fully connected layer; Indicates the The feature vector corresponding to the regularized feature map;

[0128] Divide the local descriptors of the embedded support set into local descriptors of the normal support set and local descriptors of the exception support set ;

[0129] S3. Acquire multimodal data to be detected of the target electronic voltage transformer, input the multimodal data to be detected into the fault diagnosis model, and obtain an embedded query set and a local descriptor of the embedded query set;

[0130] S4. Calculate the similarity between the local descriptors embedded in the query set and the local descriptors of the normal support set to obtain a normal similarity score; calculate the similarity between the local descriptors embedded in the query set and the local descriptors of the abnormal support set to obtain an abnormal similarity score;

[0131] The embedded query set is represented as ,in Represents an embedded query set; A local descriptor representing the embedded query set; Indicates the A local descriptor that embeds the query set; Indicates the number of local descriptors embedded in the query set;

[0132] Traverse the local descriptors in the embedded query set, search for the nearest neighbors in the local descriptors of the normal support set and the local descriptors of the abnormal support set, and calculate the normal similarity score and the abnormal similarity score, which can be expressed as follows:

[0133] ;

[0134] ;

[0135] Where, represents the normal similarity score; Indicates the weight is Similarity measurement function of ; Indicates the The local descriptor of the embedded query set is The value of the dimension; Indicates the The local descriptor of the normal support set is The value of the dimension; represents the anomaly similarity score; Indicates the The local descriptor of the abnormal support set is The value of the dimension; Represents the dimension of the embedded query set; Indicates the The index value of the embedded query set dimension; Indicates a preset constant;

[0136] S5. Compare the normal similarity score and the abnormal similarity score to obtain a fault diagnosis result of the target detected electronic voltage transformer;

[0137] like , the target electronic voltage transformer is normal; otherwise, the target electronic voltage transformer is abnormal;

[0138] S6. The method also includes continuously monitoring modal characteristics, collecting diagnostic experience of each mode, and establishing a modal fault association knowledge base; evaluating the diagnostic results using preset performance evaluation indicators to obtain evaluation results; and optimizing the fault diagnosis process in combination with the modal fault association knowledge base and the evaluation results.

[0139] Example 2:

[0140] This embodiment provides a small sample multi-modal electronic voltage transformer fault diagnosis system, which includes a data acquisition module, a local descriptor generation module, a fault diagnosis module, and a result output module, wherein:

[0141] The data acquisition module is used to acquire multimodal data and related data of the electronic voltage transformer, the multimodal data including grid frequency, secondary voltage signal and primary current signal; the related data including power data; and transmit the multimodal data and related data to the local descriptor generation module;

[0142] The local descriptor generation module is used to construct a fault diagnosis model, including a feature extraction unit and an embedding network unit; the feature extraction unit is used to extract modal features from multimodal data to obtain corresponding modal features; the secondary voltage signal and the modal features are connected and converted into a two-dimensional image to obtain a support set; the support set is input into the embedding network unit to obtain an embedded support set and a local descriptor of the embedded support set; based on the embedded support set, the local descriptor is divided into a local descriptor of a normal support set and a local descriptor of an abnormal support set; wherein, the modal feature extraction is specifically to input the grid frequency into an improved long short-term memory network LSTM network, perform fast Fourier frequency domain analysis on the secondary voltage signal, and perform overshoot calculation on the primary current signal;

[0143] The fault diagnosis module is used to obtain multimodal data to be detected of the target electronic voltage transformer, input the multimodal data to be detected into the fault diagnosis model, and obtain an embedded query set and a local descriptor of the embedded query set;

[0144] Calculate the similarity between the local descriptors embedded in the query set and the local descriptors of the normal support set to obtain a normal similarity score; calculate the similarity between the local descriptors embedded in the query set and the local descriptors of the abnormal support set to obtain an abnormal similarity score; compare the normal similarity score and the abnormal similarity score to obtain the fault diagnosis result of the target detection electronic voltage transformer;

[0145] The result output module is used to display the fault diagnosis result of the target detection electronic voltage transformer.

[0146] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention's description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A small sample multi-modal electronic voltage transformer fault diagnosis method, characterized in that: The method comprises: Acquiring multimodal data and related data of an electronic voltage transformer, wherein the multimodal data includes grid frequency, secondary voltage signal, and primary current signal; and the related data includes power data; Construct a fault diagnosis model, including a feature extraction unit and an embedding network unit; the feature extraction unit is used to extract modal features from multimodal data to obtain corresponding modal features; the secondary voltage signal and the modal features are connected and converted into a two-dimensional image to obtain a support set; the support set is input into the embedding network unit to obtain an embedded support set and a local descriptor of the embedded support set; based on the embedded support set, the local descriptor is divided into a local descriptor of a normal support set and a local descriptor of an abnormal support set; wherein the modal feature extraction specifically comprises inputting the grid frequency into an improved long short-term memory network (LSTM) network, performing fast Fourier frequency domain analysis on the secondary voltage signal, and performing overshoot calculation on the primary current signal; Acquire multimodal data to be detected of a target electronic voltage transformer, input the multimodal data to be detected into the fault diagnosis model, and obtain an embedded query set and a local descriptor of the embedded query set; Calculate the similarity between the local descriptors embedded in the query set and the local descriptors of the normal support set to obtain a normal similarity score; calculate the similarity between the local descriptors embedded in the query set and the local descriptors of the abnormal support set to obtain an abnormal similarity score; compare the normal similarity score and the abnormal similarity score to obtain the fault diagnosis result of the target detection electronic voltage transformer; Calculate the normal similarity score and abnormal similarity score, specifically: The embedded query set is represented as ,in Represents an embedded query set; A local descriptor representing the embedded query set; Indicates the A local descriptor that embeds the query set; Indicates the number of local descriptors embedded in the query set; Traverse the local descriptors in the embedded query set, search for the nearest neighbors in the local descriptors of the normal support set and the local descriptors of the abnormal support set, and calculate the normal similarity score and the abnormal similarity score, which can be expressed as follows: ； ； Where, represents the normal similarity score; Indicates the weight is Similarity measurement function of ; Indicates the The local descriptor of the embedded query set is The value of the dimension; Indicates the The local descriptor of the normal support set is The value of the dimension; represents the anomaly similarity score; Indicates the The local descriptor of the abnormal support set is The value of the dimension; Represents the dimension of the embedded query set; Indicates the The index value of the embedded query set dimension; Indicates a preset constant; like , the target electronic voltage transformer is normal; otherwise, the target electronic voltage transformer is abnormal.

2. A small sample multi-mode electronic voltage transformer fault diagnosis method according to claim 1, characterized in that: The method further includes performing data cleaning on the multimodal data and related data, wherein the data cleaning includes processing missing values, outliers and standardizing data formats.

3. The small sample multi-mode electronic voltage transformer fault diagnosis method according to claim 1, characterized in that: The grid frequency is input into the improved LSTM network and is expressed as follows: ； ； ； ； ； ； Where, express Always forget the output of the gate; represents a logical function; Represents the weight of the forget gate; express The hidden state of the moment; express The grid frequency at the moment; express The primary current signal at the moment; Represents the bias of the forget gate; express The output of the input gate at all times; Represents the weight of the input gate; represents the bias of the input gate; express Candidate cell states at time instant; represents the weight of the candidate cell state; represents the bias of the candidate cell state; represents the first activation function; express The cell state at a moment in time; express The cell state at a moment in time; express Output gate at the moment; Represents the weight of the output gate; Represents the bias of the output gate; express Power of the moment; Represents the modal characteristics of the power grid frequency; Indicates time.

4. A small sample multi-mode electronic voltage transformer fault diagnosis method according to claim 3, characterized in that: The fast Fourier frequency domain analysis is performed on the secondary voltage signal, which is expressed as follows: ； ； Where, represents the fast Fourier function; Indicates the preset frequency; Indicates the sign of the imaginary part; express Secondary voltage signal at time ; represents the real part function; represents the imaginary part function; Choose the largest As the modal characteristic of the secondary voltage signal, it is expressed as .

5. The small sample multi-mode electronic voltage transformer fault diagnosis method according to claim 4, characterized in that: The overshoot calculation of the primary current signal is expressed as follows: ； and ; ； ； Where, Indicates the average value of the primary current signal; Indicates the maximum value of the local primary current signal; Indicates the total number of primary current signals; Indicates the The index value of a primary current signal; Indicates the A primary current signal; Indicates the The maximum value of a local primary current signal; Indicates the The index value of the maximum value of the local primary current signal; Indicates the total number of maximum values ​​of local primary current signals; Indicates the The overshoot value of the maximum value of the local primary current signal; Represents the modal characteristics of the primary current signal; Indicates the The current signal before the maximum value of a local primary current signal; Indicates the The current signal after the maximum value of the local primary current signal.

6. A small sample multi-mode electronic voltage transformer fault diagnosis method according to claim 5, characterized in that: The secondary voltage signal and all modal features are connected and converted into a two-dimensional image to obtain the support set, which is expressed as: ， ； ； Where, represents the two-dimensional image matrix, i.e., the support set; Represents a two-dimensional image matrix Rank Column pixels; Represents the dimension of the two-dimensional image matrix; Represents a connection sequence; Indicates the A connection sequence value; Indicates the The index value of the connected sequence; Indicates the Secondary voltage signal value; Indicates the number of secondary voltage signals; Represents a two-dimensional image matrix The index value of the row; Represents a two-dimensional image matrix The index value of the row; Indicates the The index value of the secondary voltage signal value.

7. A small sample multi-mode electronic voltage transformer fault diagnosis method according to claim 6, characterized in that: The support set is input into the embedding network unit, which includes a Gaussian noise layer, a layer normalization layer, a convolution layer, and a Dropout layer, specifically: The Gaussian noise layer receives the support set and adds noise, which is expressed as: ； Where, represents the noise support set; Indicates Select the Gaussian distribution of noise value; represents the variance of the Gaussian distribution; The layer normalization layer receives the noise support set and performs layer normalization to obtain an embedded support set, which is expressed as: ； ； ； Where, represents the mean of the noise support set; represents the variance of the noise support set; The number of two-dimensional images representing the noise support set; Indicates the The index value of a two-dimensional image; represents the noise support set A two-dimensional image of represents the embedding support set; Indicates a preset constant; Indicates the preset scaling parameters; Indicates the preset translation parameters; The convolution layer convolves the embedded support set and further inputs the Dropout layer for regularization, which can be expressed as follows: ； ⊙ ； Where, Indicates the feature maps, which are the outputs of the convolutional layers; Indicates the The index value of the neuron; represents the total number of neurons; represents the second activation function; Indicates the Tier The connection weights of neurons; Indicates the Tier The bias of each neuron; Represents the embedding support set A two-dimensional image; Indicates the The index value of the convolution layer; Indicates the number of convolutional layers; Indicates the A regularized feature map; Represents the preset Bernoulli mask matrix; ⊙ represents element-by-element multiplication; Based on the regularized feature map, the local descriptor of the embedded support set is obtained, which is expressed as: ； ； Where, local descriptors representing the embedded support set; Indicates the A local descriptor embedded in the support set; represents the weight of the fully connected layer; represents the bias of the fully connected layer; Indicates the The feature vector corresponding to the regularized feature map; Divide the local descriptors of the embedded support set into local descriptors of the normal support set and local descriptors of the exception support set .

8. The small sample multi-mode electronic voltage transformer fault diagnosis method according to claim 1, characterized in that: The method further includes continuously monitoring modal characteristics, collecting diagnostic experience of each modality, and establishing a modal fault association knowledge base; evaluating the diagnostic results using a preset performance evaluation index to obtain an evaluation result; The fault diagnosis process is optimized by combining the modal fault association knowledge base and the evaluation results.

9. A small sample multi-modal electronic voltage transformer fault diagnosis system, characterized in that: The system includes a data acquisition module, a local descriptor generation module, a fault diagnosis module and a result output module, wherein: The data acquisition module is used to acquire multimodal data and related data of the electronic voltage transformer, the multimodal data including grid frequency, secondary voltage signal and primary current signal; the related data including power data; and transmit the multimodal data and related data to the local descriptor generation module; The local descriptor generation module is used to construct a fault diagnosis model, including a feature extraction unit and an embedding network unit; the feature extraction unit is used to extract modal features from multimodal data to obtain corresponding modal features; the secondary voltage signal and the modal features are connected and converted into a two-dimensional image to obtain a support set; the support set is input into the embedding network unit to obtain an embedded support set and a local descriptor of the embedded support set; based on the embedded support set, the local descriptor is divided into a local descriptor of a normal support set and a local descriptor of an abnormal support set; wherein, the modal feature extraction is specifically to input the grid frequency into an improved long short-term memory network LSTM network, perform fast Fourier frequency domain analysis on the secondary voltage signal, and perform overshoot calculation on the primary current signal; The fault diagnosis module is used to obtain multimodal data to be detected of the target electronic voltage transformer, input the multimodal data to be detected into the fault diagnosis model, and obtain an embedded query set and a local descriptor of the embedded query set; Calculate the similarity between the local descriptors embedded in the query set and the local descriptors of the normal support set to obtain a normal similarity score; calculate the similarity between the local descriptors embedded in the query set and the local descriptors of the abnormal support set to obtain an abnormal similarity score; compare the normal similarity score and the abnormal similarity score to obtain the fault diagnosis result of the target detection electronic voltage transformer; Calculate the normal similarity score and abnormal similarity score, specifically: The embedded query set is represented as ,in Represents an embedded query set; A local descriptor representing the embedded query set; Indicates the A local descriptor that embeds the query set; Indicates the number of local descriptors embedded in the query set; Traverse the local descriptors in the embedded query set, search for the nearest neighbors in the local descriptors of the normal support set and the local descriptors of the abnormal support set, and calculate the normal similarity score and the abnormal similarity score, which can be expressed as follows: ； ； Where, represents the normal similarity score; Indicates the weight is Similarity measurement function of ; Indicates the The local descriptor of the embedded query set is The value of the dimension; Indicates the The local descriptor of the normal support set is The value of the dimension; represents the anomaly similarity score; Indicates the The local descriptor of the abnormal support set is The value of the dimension; Represents the dimension of the embedded query set; Indicates the The index value of the embedded query set dimension; Indicates a preset constant; like , the target electronic voltage transformer is normal; otherwise, the target electronic voltage transformer is abnormal; The result output module is used to display the fault diagnosis result of the target detection electronic voltage transformer.

Images