Out-of-tolerance monitoring method, device and equipment for capacitor voltage transformer and storage medium

Abstract

The invention relates to an out-of-tolerance monitoring method and device for a capacitor voltage transformer, electronic equipment and a computer readable storage medium. Comprising the steps of performing kernel principal component analysis based on historical sample data, and determining an SPE statistic threshold value; calculating the SPE statistic of the current sample data according to the current sample data collected in real time; if the SPE statistical magnitude is smaller than or equal to an SPE statistical magnitude threshold value, it is determined that the operation state of the mutual inductor set is normal; and if the SPE statistical magnitude is greater than the SPE statistical magnitude threshold, calculating the contribution rate of each element in the current sample data to the square prediction error statistical magnitude, determining a fault element, and positioning out-of-tolerance transformers in the transformer group. According to the method, the nonlinear problem of the data is effectively processed by using the KPCA, and the data features are extracted more accurately; fault element separation is realized, and the out-of-tolerance CVT is accurately positioned; cVT errors are monitored on line in real time, and the stability and reliability of a power system are improved.

Description

Technical Field

[0001] The present application relates to the technical field of power systems, and particularly to a method, device, electronic device and computer-readable storage medium for out-of-tolerance monitoring of a capacitive voltage transformer. Background Art

[0002] As a key device in the power system, the capacitive voltage transformer (CVT) plays an indispensable role in the fields of electric energy metering, relay protection, and power system condition monitoring. The CVT converts high voltage into low voltage suitable for measurement and protection equipment through the principle of capacitive voltage division, ensuring the stable operation and precise control of all links in the power system.

[0003] During actual operation, the performance of the CVT is affected by various factors, such as capacitance aging and insulation performance degradation caused by long-term operation, drastic changes in environmental temperature and humidity, and external electromagnetic field interference. These factors will cause errors in the CVT. When the error exceeds the specified range (i.e., out-of-tolerance), a series of serious problems will occur. In terms of electric energy metering, out-of-tolerance will lead to inaccurate metering, causing economic disputes between power enterprises and users and affecting the fair trading of the power market; in the field of relay protection, out-of-tolerance may cause the protection device to malfunction or refuse to operate, threatening the safe and stable operation of the power system and even triggering large-scale power outages.

[0004] Currently, CVT error detection technologies are mainly divided into two categories: offline detection and online detection. In the offline detection, the standard comparison method requires using a standard transformer to compare with the CVT to be detected offline. This method not only cannot achieve long-term real-time monitoring of the CVT, but also has great limitations in portability of the standard transformer, making it difficult to meet the detection requirements in complex on-site environments. At the same time, there are differences between the operating conditions during offline detection and the actual operating conditions, resulting in the detection results being unable to truly reflect the state of the CVT during actual operation. In the online detection, the live comparison scheme can achieve real-time monitoring to a certain extent, but due to the involvement of high-voltage live operations, there are great safety hazards, with extremely high requirements for the professional skills and protective equipment of operators, and the operation process is complex, prone to safety accidents.

[0005] In addition, the output data of the CVT has complex distribution characteristics, making it difficult to meet the prerequisite conditions such as normal distribution relied on by traditional monitoring methods. Traditional monitoring methods based on linear models often cannot accurately extract key information from data with non-linear characteristics like the CVT, affecting the accuracy and reliability of monitoring results. With the continuous expansion of the scale of the power system and the increasing improvement of the degree of intelligence, higher requirements are put forward for the accuracy, real-time performance, and reliability of CVT out-of-tolerance monitoring. Traditional monitoring technologies can no longer meet these needs, and there is an urgent need for a new monitoring method to ensure the safe, stable, and efficient operation of the power system. Summary of the Invention

[0006] Based on this, in view of the above technical problems, it is necessary to provide a method, device, electronic device and computer-readable storage medium for out-of-tolerance monitoring of capacitive voltage transformers, which can effectively handle the problem of data non-linearity, accurately locate out-of-tolerance CVTs, and improve the stability and reliability of power systems.

[0007] In a first aspect, the present application provides a method for out-of-tolerance monitoring of a capacitive voltage transformer, including:

[0008] Performing kernel principal component analysis on historical sample data to determine a threshold of the squared prediction error statistic corresponding to the historical sample data; the historical sample data includes multiple sample data collected during a historical time period, and the sample data includes three-phase voltage data within a transformer bank;

[0009] Calculating the squared prediction error statistic of the current sample data for the currently collected current sample data;

[0010] When the squared prediction error statistic is less than or equal to the threshold of the squared prediction error statistic, determining that the operating state of the transformer bank is normal;

[0011] When the squared prediction error statistic is greater than the threshold of the squared prediction error statistic, calculating the contribution rate of each element in the current sample data to the squared prediction error statistic, and determining the fault element;

[0012] Locating the out-of-tolerance transformer in the transformer bank according to the fault element.

[0013] In one embodiment, performing kernel principal component analysis on historical sample data to determine a threshold of the squared prediction error statistic corresponding to the historical sample data includes:

[0014] Constructing a kernel matrix based on the historical sample data according to a preset kernel function;

[0015] Centering the kernel matrix, and performing eigenvalue decomposition on the centered kernel matrix to determine multiple eigenvalues and multiple eigenvectors;

[0016] Selecting k eigenvalues from the multiple eigenvalues to form a variance contribution diagonal matrix, and selecting k eigenvectors from the multiple eigenvectors to form a principal component direction matrix;

[0017] For the i-th sample data, calculating the principal component projection of the i-th sample data according to the i-th row vector of the centered kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix;

[0018] Based on a preset kernel function, determine the self-kernel function value of the i-th sample data, and determine the squared prediction error statistic of the i-th sample data according to the self-kernel function value and the principal component projection;

[0019] According to the squared prediction error statistics corresponding to multiple sample data respectively, determine the threshold of the squared prediction error statistic corresponding to the preset confidence level through a kernel density estimation strategy.

[0020] In one embodiment, the method further includes:

[0021] Calculate a mapping matrix according to the centralized kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix;

[0022] Determine the mapping feature of the i-th sample data according to the mapping matrix, and calculate the residual vector of the i-th sample data according to the mapping feature, the mapping matrix, and the principal component projection;

[0023] Calculate the squared prediction error statistic of the i-th sample data according to the residual vector of the i-th sample data.

[0024] In one embodiment, selecting k eigenvalues from multiple eigenvalues to form a variance contribution diagonal matrix, and selecting k eigenvectors from multiple eigenvectors to form a principal component direction matrix includes:

[0025] Select the first k largest eigenvalues from multiple eigenvalues so that the cumulative contribution rate reaches a set threshold, and form a variance contribution diagonal matrix according to the selected k eigenvalues;

[0026] Select the eigenvectors corresponding to the first k largest eigenvalues from multiple eigenvectors, and form a principal component direction matrix according to the selected k eigenvectors.

[0027] In one embodiment, calculating the squared prediction error statistic of the current sample data includes:

[0028] Based on a preset kernel function, construct a current kernel matrix of the current sample data and historical sample data;

[0029] Perform centralized processing on the current kernel matrix;

[0030] Calculate the current principal component projection of the current sample data according to the centralized current kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix;

[0031] Based on a preset kernel function, determine the current self-kernel function value of the current sample data, and determine the squared prediction error statistic of the current sample data according to the current self-kernel function value and the current principal component projection.

[0032] In one embodiment, the contribution rate of each element in the current sample data to the squared prediction error statistic is calculated by the following formula:

[0033] ;

[0034] where, represents the contribution rate of the i-th element to the squared prediction error statistic, represents the value of the i-th element, represents the j-th sample data in the historical sample data, represents the current sample data, represents the kernel function;

[0035] Correspondingly, determining the faulty element includes:

[0036] Determining the contribution rate percentage corresponding to each element by the following formula:

[0037] ;

[0038] Determining the element with the largest contribution rate percentage as the faulty element.

[0039] In one embodiment, after locating the out-of-tolerance instrument transformer in the instrument transformer group according to the faulty element, the method further includes:

[0040] Sending an alarm message according to the out-of-tolerance instrument transformer.

[0041] In a second aspect, the present application further provides an out-of-tolerance monitoring device for a capacitive voltage transformer, including:

[0042] A historical data processing module, configured to perform kernel principal component analysis based on historical sample data to determine a threshold of the squared prediction error statistic corresponding to the historical sample data; the historical sample data includes multiple sample data collected during a historical time period, and the sample data includes three-phase voltage data within the instrument transformer group;

[0043] A real-time data processing module, configured to calculate the squared prediction error statistic of the currently collected current sample data;

[0044] An out-of-tolerance monitoring module, configured to determine that the operation state of the instrument transformer group is normal when the squared prediction error statistic is less than or equal to the threshold of the squared prediction error statistic; when the squared prediction error statistic is greater than the threshold of the squared prediction error statistic, calculate the contribution rate of each element in the current sample data to the squared prediction error statistic, determine the faulty element; and locate the out-of-tolerance instrument transformer in the instrument transformer group according to the faulty element.

[0045] In a third aspect, the present application further provides an electronic device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the method described in the first aspect above are implemented.

[0046] In a fourth aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in the first aspect above are implemented.

[0047] The above-mentioned out-of-tolerance monitoring method, device, electronic device and computer-readable storage medium of the capacitive voltage transformer perform kernel principal component analysis based on historical sample data to determine the threshold of the squared prediction error statistic corresponding to the historical sample data; the historical sample data includes multiple sample data collected within a historical time period, and the sample data includes three-phase voltage data within the transformer group; for the currently collected current sample data, calculate the squared prediction error statistic of the current sample data; when the squared prediction error statistic is less than or equal to the threshold of the squared prediction error statistic, determine that the operating state of the transformer group is normal; when the squared prediction error statistic is greater than the threshold of the squared prediction error statistic, calculate the contribution rate of each element in the current sample data to the squared prediction error statistic, and determine the faulty element; locate the out-of-tolerance transformer in the transformer group according to the faulty element. By the above method, the kernel principal component analysis (KPCA) is used to mine the characteristics of the three-phase voltage data of the CVT operation, effectively handle the non-linear problem of data, break through the limitations of traditional linear methods, extract data characteristics more accurately, improve the accuracy of out-of-tolerance monitoring, and provide a reliable basis for timely discovering the out-of-tolerance situation of the CVT. By calculating the contribution rate of elements to separate faulty elements and locate the out-of-tolerance transformer in the transformer group, it provides a precise fault troubleshooting direction for maintenance personnel, improves the maintenance efficiency, reduces the power outage time and economic losses, and ensures the continuity and stability of power supply. It can perform online real-time monitoring of the CVT error, continuously track the operating state of the CVT, timely capture the error change, and improve the stability and reliability of the power system. Description of the Drawings

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for the description of the embodiments of the present application or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other related drawings can be obtained according to these drawings.

[0049] Figure 1 It is a schematic flowchart of the out-of-tolerance monitoring method of the capacitive voltage transformer in an embodiment;

[0050] Figure 2Schematic flowchart of the out-of-tolerance monitoring method for a capacitive voltage transformer in another embodiment;

[0051] Figure 3 Block diagram of the structure of the out-of-tolerance monitoring device for a capacitive voltage transformer in one embodiment;

[0052] Figure 4 Internal structure diagram of an electronic device in one embodiment. Detailed implementation manners

[0053] In order to make the objectives, technical solutions and advantages of the present application clearer and more understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0054] In an exemplary embodiment, as Figure 1 shown, a method for monitoring out-of-tolerance of a capacitive voltage transformer is provided. In this embodiment, it is exemplified that this method is applied to an electronic device. The method includes the following steps:

[0055] Step 102: Perform kernel principal component analysis on historical sample data to determine the threshold of the squared prediction error statistic corresponding to the historical sample data; the historical sample data includes multiple sample data collected during a historical time period, and the sample data includes three-phase voltage data within the transformer bank.

[0056] Among them, taking the three-phase voltage data within the transformer bank as the sample data for a single collection, the electronic device applying the out-of-tolerance monitoring method for a capacitive voltage transformer provided in this embodiment collects multiple sample data during a historical time period through voltage sensors installed at each capacitive voltage transformer (CVT) to form historical sample data.

[0057] Construct a kernel matrix for the historical sample data and centralize it. Establish a steady state through calculations such as KPCA feature decomposition and component decomposition to obtain the threshold of the squared prediction error (SPE) statistic of the historical sample data. It can be understood that traditional principal component analysis (PCA) will produce deviations when used for nonlinear processes. In this embodiment, based on KPCA, through the kernel trick, the data is mapped from the original space to a high-dimensional space, and then PCA is performed in the high-dimensional space to solve the problem of dimensionality reduction of nonlinear data.

[0058] Optionally, the threshold of the squared prediction error statistic is configured as a parameter of the monitoring model. The electronic device collects sample data in real time. If it detects that the power system is in a steady state, it continuously updates the monitoring model based on the real-time collected CVT operation data. A steady state refers to a state in which parameters such as voltage, current, and power remain relatively stable during system operation.

[0059] Step 104 : Calculate the square prediction error statistic of the current sample data collected in real time.

[0060] Step 106: When the square prediction error statistic is less than or equal to the square prediction error statistic threshold, determine that the operating state of the mutual inductor group is normal.

[0061] The monitoring model compares the real-time SPE statistic with the SPE statistic threshold. If the SPE statistic does not exceed the SPE statistic threshold, it is determined that the operating state of the mutual inductor group is normal.

[0062] Step 108 : When the square prediction error statistic is greater than the square prediction error statistic threshold, the contribution rate of each element in the current sample data to the square prediction error statistic is calculated to determine the fault element.

[0063] Step 110: locate the out-of-tolerance transformer in the transformer group according to the fault element.

[0064] If the SPE statistic exceeds the SPE statistic threshold, the transformer group's operating status is determined to be abnormal. The fault isolation strategy further traces the source, identifying the faulty element based on the contribution of each element in the current sample data to the squared prediction error statistic, and locating the out-of-tolerance transformer in the transformer group.

[0065] It's understandable that the accurate feature extraction used in calculating the SPE statistic threshold for historical sample data makes the monitoring model's assessment of CVT out-of-tolerance conditions more reliable. This ability to accurately identify subtle error changes in complex power environments significantly improves the accuracy of out-of-tolerance monitoring compared to traditional methods, reducing misjudgments and missed detections and providing stronger assurance for the stable operation of the power system.

[0066] The above-mentioned out-of-tolerance monitoring method for capacitor voltage transformers utilizes kernel principal component analysis to mine the voltage data characteristics of the three-phase CVT operating within the group. This effectively addresses the nonlinearity of the data, overcomes the limitations of traditional linear methods, and more accurately extracts data features, improving the accuracy of out-of-tolerance monitoring and providing a reliable basis for timely detection of CVT out-of-tolerance conditions. By calculating the contribution rate of each element, faulty elements are isolated and out-of-tolerance transformers within the transformer group are located. This provides maintenance personnel with precise troubleshooting guidance, improves maintenance efficiency, reduces power outages and economic losses, and ensures the continuity and stability of power supply. This method enables online, real-time monitoring of CVT errors, continuously tracking the CVT's operating status, and promptly capturing error changes, thereby improving the stability and reliability of the power system.

[0067] In an exemplary embodiment, step 102 includes: constructing a kernel matrix based on a preset kernel function according to historical sample data; performing centering processing on the kernel matrix, and performing eigenvalue decomposition on the centered kernel matrix to determine a plurality of eigenvalues and a plurality of eigenvectors; selecting k eigenvalues from the plurality of eigenvalues to form a variance contribution diagonal matrix, and selecting k eigenvectors from the plurality of eigenvectors to form a principal component direction matrix; for the i-th sample data, calculating the principal component projection of the i-th sample data according to the i-th row vector of the centered kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; determining the self-kernel function value of the i-th sample data based on the preset kernel function, and determining the squared prediction error statistic of the i-th sample data according to the self-kernel function value and the principal component projection; determining the squared prediction error statistic threshold corresponding to a preset confidence level through a kernel density estimation strategy according to the squared prediction error statistics corresponding to multiple sample data.

[0068] Wherein, assuming that the historical sample data includes n sample data, the kernel matrix is constructed through the following formula:

[0069] ;

[0070] In the formula, represents the preset kernel function, represents the i-th sample data, represents the j-th sample data.

[0071] The preset kernel function can be a Gaussian radial basis function (RBF) kernel function, a p-th order polynomial kernel function, a sigmoid kernel function, etc., and a specific kernel function can be selected according to actual needs. Taking the preset kernel function as the Gaussian kernel function as an example, it can handle complex non-linear relationships, and its expression is as follows:

[0072] ;

[0073] In the formula, is the key parameter of the Gaussian kernel function.

[0074] The centering processing method is used to transform the kernel matrix to eliminate the mean shift in the feature space. The specific transformation method is as follows:

[0075] ;

[0076] In the formula, is an n×n identity matrix.

[0077] Based on the following formula, eigenvalue decomposition is performed on the centered kernel matrix:

[0078] ;

[0079] Where, represents the eigenvalue, represents the feature vector.

[0080] Select k eigenvalues to form the variance contribution diagonal matrix And the k eigenvectors corresponding to the k eigenvalues constitute the principal component direction matrix .

[0081] For the i-th sample data, the principal component projection is calculated using the following formula:

[0082] ;

[0083] Where, express The i-th row vector of Used to standardize the eigenvalues in the variance contribution diagonal matrix to ensure that the principal component direction has unit variance. represents the principal component direction matrix.

[0084] The SPE statistic of the i-th sample data is determined by the following formula:

[0085] ;

[0086] Where, Represents the value of the self-kernel function.

[0087] Based on the method provided in this embodiment, the SPE statistics corresponding to each of the n sample data can be calculated, and the SPE statistic threshold corresponding to the preset confidence level can be determined by the kernel density estimation strategy. The preset confidence level refers to the confidence level set in advance according to actual needs. This embodiment does not limit the specific value. Taking the preset confidence level of 95% as an example, the SPE statistics of the n sample data are fitted using the kernel density estimation method to obtain the probability density function f(x) of SPE. The probability density function f(x) is integrated to obtain the cumulative distribution function F(x). The x value that makes F(x)=0.95 is searched, that is, the SPE statistic threshold corresponding to the 95% confidence level is obtained. This method is also effective for data whose original measurement points do not obey the normal distribution.

[0088] In an exemplary embodiment, the method further includes: calculating a mapping matrix based on the kernel matrix after centering, the variance contribution diagonal matrix, and the principal component direction matrix; determining the mapping characteristics of the i-th sample data based on the mapping matrix, and calculating the residual vector of the i-th sample data based on the mapping characteristics, the mapping matrix, and the principal component projection; and calculating the square prediction error statistic of the i-th sample data based on the residual vector of the i-th sample data.

[0089] It is understandable that step 102 includes: constructing a kernel matrix based on a preset kernel function according to historical sample data; performing centering processing on the kernel matrix, and performing eigenvalue decomposition on the centered kernel matrix to determine multiple eigenvalues and multiple eigenvectors; selecting k eigenvalues from the multiple eigenvalues to form a variance contribution diagonal matrix, and selecting k eigenvectors from the multiple eigenvectors to form a principal component direction matrix; calculating a mapping matrix according to the centered kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; for the i-th sample data, calculating the principal component projection of the i-th sample data according to the i-th row vector of the centered kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; determining the mapping feature of the i-th sample data according to the mapping matrix, and calculating the residual vector of the i-th sample data according to the mapping feature, the mapping matrix, and the principal component projection; calculating the squared prediction error statistic of the i-th sample data according to the residual vector of the i-th sample data; and determining the squared prediction error statistic threshold corresponding to the preset confidence level through a kernel density estimation strategy according to the squared prediction error statistics corresponding to multiple sample data.

[0090] In a specific implementation, assuming that the historical sample data includes n sample data, the kernel matrix is constructed through the following formula:

[0091] ;

[0092] In the formula, represents the preset kernel function, represents the i-th sample data, represents the j-th sample data.

[0093] The preset kernel function can be a Gaussian radial basis function (RBF) kernel function, a p-order polynomial kernel function, a sigmoid kernel function, etc., and a specific kernel function can be selected according to actual needs. Taking the preset kernel function as a Gaussian kernel function as an example, it can handle complex non-linear relationships, and its expression is as follows:

[0094] ;

[0095] In the formula, is the key parameter of the Gaussian kernel function.

[0096] The centering processing method is used to transform the kernel matrix to eliminate the mean shift in the feature space. The specific transformation method is as follows:

[0097] ;

[0098] In the formula, is an n×n identity matrix.

[0099] Perform eigenvalue decomposition on the centralized kernel matrix based on the following formula:

[0100] ;

[0101] In the formula, represents the eigenvalue, represents the eigenvector.

[0102] Select k eigenvalues to form the variance contribution diagonal matrix , and form the principal component direction matrix with the k eigenvectors corresponding to the k eigenvalues.

[0103] Calculate the mapping matrix through the following formula:

[0104] ;

[0105] In the formula, represents the mapping matrix of the first k principal components, is used to standardize the eigenvalues in the variance contribution diagonal matrix to ensure that the principal component direction has unit variance, represents the principal component direction matrix.

[0106] For the i-th sample data, calculate its principal component projection using the following formula:

[0107] ;

[0108] In the formula, represents the principal component projection, represents the i-th row vector of.

[0109] Reconstruct the original feature space and calculate the residual vector of the i-th sample data:

[0110] ;

[0111] In the formula, represents the residual vector.

[0112] Calculate the SPE statistic of the i-th sample data through the sum of squares of the residual vector:

[0113] ;

[0114] Based on the method provided in this embodiment, the SPE statistics corresponding to n sample data can be calculated respectively, and the SPE statistic threshold corresponding to the preset confidence level can be determined through the kernel density estimation strategy. This method is also effective for data where the original measurement points do not follow a normal distribution.

[0115] In an exemplary embodiment, selecting k eigenvalue from multiple eigenvalues to form a variance contribution diagonal matrix, and selecting k eigenvectors from multiple eigenvectors to form a principal component direction matrix, includes: selecting the first k largest eigenvalues from multiple eigenvalues, such that the cumulative contribution rate reaches a set threshold, and forming a variance contribution diagonal matrix according to the selected k eigenvalues; selecting the eigenvectors corresponding to the first k largest eigenvalues from multiple eigenvectors, and forming a principal component direction matrix according to the selected k eigenvectors.

[0116] Wherein, using the cumulative contribution rate strategy, select the first k largest eigenvalues, such that the cumulative contribution rate reaches a set threshold, and form a variance contribution diagonal matrix according to the selected k eigenvalues and form a principal component direction matrix with the corresponding k eigenvectors The cumulative contribution rate refers to the ratio of the sum of the first k largest eigenvalues to the sum of all eigenvalues, and the preset threshold is a parameter set in advance, generally any value selected from 85% to 95%.

[0117] In an exemplary embodiment, calculating the squared prediction error statistic of the current sample data, includes: based on a preset kernel function, constructing a current kernel matrix of the current sample data and historical sample data; performing a centering process on the current kernel matrix; according to the centered current kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix, calculating the current principal component projection of the current sample data; based on the preset kernel function, determining the current self-kernel function value of the current sample data, and determining the squared prediction error statistic of the current sample data according to the current self-kernel function value and the current principal component projection.

[0118] It can be understood that the statistics in the PCA process include the T2 statistic and the Q statistic. The T2 statistic is sensitive to the error in the principal component space and can reflect the systematic change in the measurement process; the Q statistic is the total change in the residual space, also known as the squared prediction error (SPE), which reflects the change related to the measurement noise.

[0119] Wherein, assuming the current sample data is calculate its current kernel matrix with the historical sample data using the same kernel function as the historical sample data The elements in the matrix are:

[0120] ;

[0121] Use the centering process method to transform the current kernel matrix to eliminate the mean shift in the feature space. The specific transformation method is as follows:

[0122] ;

[0123] Map the current sample data to the principal component direction of the high-dimensional feature space and calculate the current principal component projection:

[0124] ;

[0125] Simplify the calculation of the SPE statistic of the current sample data through the kernel function:

[0126] ;

[0127] Optionally, in addition to simplifying the calculation of the SPE statistic based on the kernel function, the current mapping matrix can also be calculated according to the current kernel matrix, variance contribution diagonal matrix, and principal component direction matrix after centering processing; determine the current mapping features of the current sample data according to the current mapping matrix, and calculate the current residual vector of the current sample data according to the current mapping features, the current mapping matrix, and the current principal component projection; calculate the SPE statistic of the current sample data according to the current residual vector of the current sample data.

[0128] In an exemplary embodiment, calculate the contribution rate of each element in the current sample data to the squared prediction error statistic through the following formula:

[0129] ;

[0130] where, represents the contribution rate of the i-th element to the squared prediction error statistic, represents the value of the i-th element, represents the j-th sample data in the historical sample data, represents the current sample data, represents the kernel function;

[0131] Correspondingly, determining the faulty element includes:

[0132] Determine the contribution rate percentage corresponding to each element through the following formula:

[0133] ;

[0134] Determine the element with the largest contribution rate percentage as the faulty element.

[0135] It can be understood that since the explicit representation cannot be adopted between the input variables (i.e., each element in the current sample data) and the kernel principal components, the contribution rate of the input variables to the SPE statistic can be solved through the partial derivative of the kernel function. Assume that the current sample data V = [v1, …, v m T ∈R m×1 , where the single element is v i ​When (i = 1, …, m) and the selected kernel function is the Gaussian kernel function, the contribution rates of each variable in the input space to the statistic T2 and SPE are derived as follows (hereafter, i represents the i-th element in the current sample data, and j represents the j-th measurement value in the historical sample data):

[0136] (1);

[0137] (2);

[0138] In the formula, is the contribution rate of the i-th element in the current sample data to the T2 statistic. tr(*) represents the trace of matrix *, α is the standardized eigenvector, is the current kernel matrix after centering processing, and any one of its elements is (p, q = 1, …, n), is a diagonal matrix; is the contribution rate of the i-th element in the current sample data to the SPE statistic, is the parameter of the Gaussian kernel function, (j = 1, …, n) is the historical sample data, is the current sample data, is the SPE value of.

[0139] Specifically, taking the partial derivative of can be decomposed into taking the partial derivative of each element, that is:[[]]

[0140] ;

[0141] Similarly, taking the partial derivative of can also be decomposed into taking the partial derivative of each of its elements, that is:[[]]

[0142] ;

[0143] In the formula:[[]]

[0144] ;

[0145] ;

[0146] ;

[0147] From the derivation results, it can be seen that the calculation process of obtaining the contribution rates of each variable is extremely complex, and the main computational effort comes from taking the partial derivative of the matrix with respect to the variable. Assume that the matrix is:[[]]

[0148] ;

[0149] Then Equation (1) represents finding the trace of the matrix , that is, the contribution rate of the i-th element in the input space to the change in the principal component space is the trace of the matrix . Without loss of generality, Equation (2) can be split, and the following derivation can be performed on the second term:

[0150] ;

[0151] Since is a diagonal matrix, then represents 's linear transformation. Also, since represents the contribution rate of the i-th element in the input space to the change in the principal component space, then , that is, the second term in Equation (2) also represents the influence of the i-th element in the input space on the change in the principal component space. Therefore, the first term in Equation (2) after removing this term can be expressed as the influence of the i-th element in the input space on the change in the residual space. Based on the above derivation, the contribution rate of an element to the residual space can be defined as:

[0152] ;

[0153] In practical applications, normalization processing is used to calculate the specific contribution percentage of each element in the current sample data:

[0154] ;

[0155] The element with the largest contribution rate percentage is determined as the faulty element. In this way, the contribution rate of the input variables (i.e., each element in the current sample data) to the SPE statistic is solved through the partial derivative of the kernel function. When multiple current transformers fail simultaneously, the influence degree of each variable on the fault can be clarified, so as to accurately separate the fault variables and quickly locate the faulty current transformer. This characteristic provides a precise fault troubleshooting direction for maintenance personnel, greatly shortening the fault troubleshooting time, improving the maintenance efficiency, reducing the power outage time and economic losses caused by power system faults, and ensuring the continuity and stability of power supply.

[0156] In an exemplary embodiment, after step 110, the method further includes: sending an alarm message according to the out-of-tolerance current transformer.

[0157] It can be understood that during actual operation, the dynamic changes of the CVT error are captured in a timely manner. Once an out-of-tolerance situation occurs, an alarm message is quickly sent out to feedback the alarm signal. Through this real-time monitoring function, the operation and maintenance personnel can take measures immediately to avoid the impact on power metering and the safe and stable operation of the power system caused by the out-of-tolerance of the CVT error. It is of great significance for ensuring the efficient operation of the power system, improving the accuracy of power metering, and maintaining the fair trading in the power market, and can meet the development needs of the modern intelligent and efficient power system.

[0158] In an optional implementation manner, referring to Figure 2 , the out-of-tolerance monitoring method for a capacitive voltage transformer includes the following steps:

[0159] Taking the three-phase voltage data within the same transformer group as a single acquisition sample, during the establishment of the system steady state, historical data is collected, and KPCA feature decomposition is performed on the historical data to determine the SPE statistic threshold;

[0160] Collect real-time data, and calculate its SPE statistic for the collected data through high-dimensional projection;

[0161] Compare the obtained SPE statistic with the SPE statistic threshold to determine whether there is an out-of-tolerance. If there is no out-of-tolerance, it indicates that the operation state of the transformer group is normal;

[0162] If there is an out-of-tolerance in the three-phase voltage within the transformer group, the contribution rate of the input variable to the SPE statistic is solved through the partial derivative of the kernel function to separate the fault variable, locate the out-of-tolerance transformer (i.e., the faulty transformer) within the transformer group and give an alarm.

[0163] It should be understood that although each step in the flowcharts involved in the above-described embodiments is shown in sequence according to the indication of the arrow, these steps are not necessarily executed in the order indicated by the arrow. Unless there is a clear description in this article, the execution of these steps does not have a strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same moment, but can be executed at different moments. The execution order of these steps or stages is not necessarily sequential either, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0164] Based on the same inventive concept, an embodiment of the present application further provides a device for monitoring an out-of-tolerance capacitor voltage transformer for implementing the above-mentioned method for monitoring an out-of-tolerance capacitor voltage transformer. The solution provided by this device is similar to the solution described in the above-mentioned method. Therefore, the specific limitations of the embodiments of one or more devices for monitoring an out-of-tolerance capacitor voltage transformer provided below can be found in the above-mentioned limitations of the method for monitoring an out-of-tolerance capacitor voltage transformer, and will not be repeated here.

[0165] In an exemplary embodiment, Figure 3 As shown, a device for monitoring an over-tolerance of a capacitor voltage transformer is provided, comprising:

[0166] The historical data processing module 302 is used to perform kernel principal component analysis based on historical sample data to determine the square prediction error statistic threshold corresponding to the historical sample data; the historical sample data includes multiple sample data collected within a historical time period, and the sample data includes three-phase voltage data in the transformer group.

[0167] The real-time data processing module 304 is used to calculate the square prediction error statistics of the current sample data collected in real time.

[0168] The out-of-tolerance monitoring module 306 is used to determine that the operating status of the transformer group is normal when the square prediction error statistic is less than or equal to the square prediction error statistic threshold; when the square prediction error statistic is greater than the square prediction error statistic threshold, calculate the contribution rate of each element in the current sample data to the square prediction error statistic and determine the fault element; and locate the out-of-tolerance transformer in the transformer group based on the fault element.

[0169] The aforementioned out-of-tolerance monitoring device for capacitor voltage transformers utilizes kernel principal component analysis (KPCA) to identify the characteristics of three-phase voltage data during CVT operation. This effectively addresses nonlinear data issues, overcomes the limitations of traditional linear methods, and more precisely extracts data features, improving the accuracy of out-of-tolerance monitoring and providing a reliable basis for timely detection of CVT out-of-tolerance conditions. By calculating the contribution rate of these elements, faulty elements are isolated and out-of-tolerance transformers within the transformer group are located. This provides maintenance personnel with precise troubleshooting guidance, improves maintenance efficiency, reduces outages and economic losses, and ensures the continuity and stability of power supply. The system enables online, real-time monitoring of CVT errors, continuously tracking the CVT's operating status and promptly capturing error changes, thereby enhancing the stability and reliability of the power system.

[0170] In an exemplary embodiment, the historical data processing module 302 is further configured to construct a kernel matrix based on a preset kernel function according to historical sample data; perform centering processing on the kernel matrix, and perform eigenvalue decomposition on the centered kernel matrix to determine a plurality of eigenvalues and a plurality of eigenvectors; select k eigenvalues from the plurality of eigenvalues to form a variance contribution diagonal matrix, and select k eigenvectors from the plurality of eigenvectors to form a principal component direction matrix; for the i-th sample data, calculate the principal component projection of the i-th sample data according to the i-th row vector of the centered kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; determine the self-kernel function value of the i-th sample data based on the preset kernel function, and determine the squared prediction error statistic of the i-th sample data according to the self-kernel function value and the principal component projection; determine the squared prediction error statistic threshold corresponding to a preset confidence level through a kernel density estimation strategy according to the squared prediction error statistics corresponding to multiple sample data.

[0171] In an exemplary embodiment, the historical data processing module 302 is further configured to calculate a mapping matrix according to the centered kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; determine the mapping feature of the i-th sample data according to the mapping matrix, and calculate the residual vector of the i-th sample data according to the mapping feature, the mapping matrix, and the principal component projection; calculate the squared prediction error statistic of the i-th sample data according to the residual vector of the i-th sample data.

[0172] In an exemplary embodiment, the historical data processing module 302 is further configured to select the first k largest eigenvalues from the plurality of eigenvalues so that the cumulative contribution rate reaches a set threshold, and form a variance contribution diagonal matrix according to the selected k eigenvalues; select the eigenvectors corresponding to the first k largest eigenvalues from the plurality of eigenvectors, and form a principal component direction matrix according to the selected k eigenvectors.

[0173] In an exemplary embodiment, the real-time data processing module 304 is further configured to construct a current kernel matrix of the current sample data and the historical sample data based on a preset kernel function; perform centering processing on the current kernel matrix; calculate the current principal component projection of the current sample data according to the centered current kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; determine the current self-kernel function value of the current sample data based on the preset kernel function, and determine the squared prediction error statistic of the current sample data according to the current self-kernel function value and the current principal component projection.

[0174] In an exemplary embodiment, the out-of-tolerance monitoring module 306 is further configured to calculate the contribution rate of each element in the current sample data to the squared prediction error statistic through the following formula:

[0175] ;

[0176] Among them, represents the contribution rate of the i-th element to the square prediction error statistic, represents the value of the i-th element, represents the j-th sample data in the historical sample data, represents the current sample data, represents the kernel function;

[0177] The contribution rate percentage corresponding to each element is determined by the following formula:

[0178] ;

[0179] The element with the largest contribution rate percentage is determined as the faulty element.

[0180] In an exemplary embodiment, the out-of-tolerance monitoring module 306 is further configured to issue an alarm message according to the out-of-tolerance mutual inductor.

[0181] Each module in the above out-of-tolerance monitoring device of the capacitive voltage transformer can be implemented in whole or in part by software, hardware, and their combination. The above modules can be embedded in or independent of the processor in the electronic device in the form of hardware, or stored in the memory of the electronic device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0182] In an exemplary embodiment, an electronic device is provided. The electronic device can be a server, and its internal structure diagram can be as Figure 4 shown. The electronic device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the electronic device is used to exchange information between the processor and external devices. The communication interface of the electronic device is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, it implements a method for monitoring out-of-tolerance of a capacitive voltage transformer.

[0183] Those skilled in the art can understand, Figure 4The structure shown is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the electronic device to which the solution of this application is applied. The specific electronic device may include more or fewer components than those shown in the figure, or combine some components, or have a different component layout.

[0184] In an exemplary embodiment, an electronic device is provided, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, the following steps are implemented: performing kernel principal component analysis based on historical sample data to determine a threshold of the squared prediction error statistic corresponding to the historical sample data; the historical sample data includes multiple sample data collected during a historical time period, and the sample data includes three-phase voltage data within a transformer bank; for currently collected current sample data, calculating the squared prediction error statistic of the current sample data; in the case where the squared prediction error statistic is less than or equal to the threshold of the squared prediction error statistic, determining that the operating state of the transformer bank is normal; in the case where the squared prediction error statistic is greater than the threshold of the squared prediction error statistic, calculating the contribution rate of each element in the current sample data to the squared prediction error statistic, and determining a faulty element; and locating an out-of-tolerance transformer in the transformer bank according to the faulty element.

[0185] In an embodiment, when the processor executes the computer program, the following steps are further implemented: constructing a kernel matrix based on the historical sample data according to a preset kernel function; performing centering processing on the kernel matrix, and performing eigenvalue decomposition on the centered kernel matrix to determine multiple eigenvalues and multiple eigenvectors; selecting k eigenvalues from the multiple eigenvalues to form a variance contribution diagonal matrix, and selecting k eigenvectors from the multiple eigenvectors to form a principal component direction matrix; for the i-th sample data, calculating the principal component projection of the i-th sample data according to the i-th row vector of the centered kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; determining the self-kernel function value of the i-th sample data based on the preset kernel function, and determining the squared prediction error statistic of the i-th sample data according to the self-kernel function value and the principal component projection; and determining the threshold of the squared prediction error statistic corresponding to a preset confidence level through a kernel density estimation strategy according to the squared prediction error statistics corresponding to the multiple sample data.

[0186] In an embodiment, when the processor executes the computer program, the following steps are further implemented: calculating a mapping matrix according to the centered kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; determining the mapping feature of the i-th sample data according to the mapping matrix, and calculating the residual vector of the i-th sample data according to the mapping feature, the mapping matrix, and the principal component projection; and calculating the squared prediction error statistic of the i-th sample data according to the residual vector of the i-th sample data.

[0187] In one embodiment, when the processor executes the computer program, the following steps are further implemented: select the top k largest eigenvalue from a plurality of eigenvalues such that the cumulative contribution rate reaches a set threshold, and form a variance contribution diagonal matrix according to the selected k eigenvalues; select the eigenvectors corresponding to the top k largest eigenvalues from a plurality of eigenvectors, and form a principal component direction matrix according to the selected k eigenvectors.

[0188] In one embodiment, when the processor executes the computer program, the following steps are further implemented: based on a preset kernel function, construct a current kernel matrix of the current sample data and the historical sample data; perform centering processing on the current kernel matrix; calculate the current principal component projection of the current sample data according to the centered current kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; based on the preset kernel function, determine the current self-kernel function value of the current sample data, and determine the squared prediction error statistic of the current sample data according to the current self-kernel function value and the current principal component projection.

[0189] In one embodiment, when the processor executes the computer program, the following steps are further implemented: calculate the contribution rate of each element in the current sample data to the squared prediction error statistic through the following formula:

[0190] ;

[0191] where, represents the contribution rate of the i-th element to the squared prediction error statistic, represents the value of the i-th element, represents the j-th sample data in the historical sample data, represents the current sample data, represents the kernel function;

[0192] Correspondingly, to determine the faulty element, it includes: determining the contribution rate percentage corresponding to each element through the following formula:

[0193] ;

[0194] Determine the element with the largest contribution rate percentage as the faulty element.

[0195] In one embodiment, when the processor executes the computer program, the following steps are further implemented: send an alarm message according to the out-of-tolerance mutual inductor.

[0196] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: performing kernel principal component analysis based on historical sample data to determine a threshold of the squared prediction error statistic corresponding to the historical sample data; the historical sample data includes multiple sample data collected within a historical time period, and the sample data includes three-phase voltage data within a transformer bank; for currently collected current sample data, calculating the squared prediction error statistic of the current sample data; in the case where the squared prediction error statistic is less than or equal to the threshold of the squared prediction error statistic, determining that the operating state of the transformer bank is normal; in the case where the squared prediction error statistic is greater than the threshold of the squared prediction error statistic, calculating the contribution rate of each element in the current sample data to the squared prediction error statistic to determine a faulty element; and locating an out-of-tolerance transformer in the transformer bank according to the faulty element.

[0197] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented: constructing a kernel matrix according to the historical sample data based on a preset kernel function; performing centering processing on the kernel matrix, and performing eigenvalue decomposition on the centered kernel matrix to determine multiple eigenvalues and multiple eigenvectors; selecting k eigenvalues from the multiple eigenvalues to form a variance contribution diagonal matrix, and selecting k eigenvectors from the multiple eigenvectors to form a principal component direction matrix; for the i-th sample data, calculating the principal component projection of the i-th sample data according to the i-th row vector of the centered kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; determining the self-kernel function value of the i-th sample data based on the preset kernel function, and determining the squared prediction error statistic of the i-th sample data according to the self-kernel function value and the principal component projection; and determining the threshold of the squared prediction error statistic corresponding to a preset confidence level through a kernel density estimation strategy according to the squared prediction error statistics respectively corresponding to the multiple sample data.

[0198] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented: calculating a mapping matrix according to the centered kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; determining the mapping feature of the i-th sample data according to the mapping matrix, and calculating the residual vector of the i-th sample data according to the mapping feature, the mapping matrix, and the principal component projection; and calculating the squared prediction error statistic of the i-th sample data according to the residual vector of the i-th sample data.

[0199] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented: selecting the first k largest eigenvalues from the multiple eigenvalues such that the cumulative contribution rate reaches a set threshold, and forming a variance contribution diagonal matrix according to the selected k eigenvalues; and selecting the eigenvectors corresponding to the first k largest eigenvalues from the multiple eigenvectors, and forming a principal component direction matrix according to the selected k eigenvectors.

[0200] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented: constructing a current kernel matrix of the current sample data and the historical sample data based on a preset kernel function; performing centering processing on the current kernel matrix; calculating a current principal component projection of the current sample data according to the centered current kernel matrix, a variance contribution diagonal matrix, and a principal component direction matrix; determining a current self-kernel function value of the current sample data based on the preset kernel function, and determining a squared prediction error statistic of the current sample data according to the current self-kernel function value and the current principal component projection.

[0201] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented: calculating a contribution rate of each element in the current sample data to the squared prediction error statistic through the following formula:

[0202] ;

[0203] wherein, [[]] represents the contribution rate of the i-th element to the squared prediction error statistic, [[]] represents the value of the i-th element, [[]] represents the j-th sample data in the historical sample data, [[]] represents the current sample data, [[]] represents the kernel function;

[0204] Correspondingly, determining a faulty element includes: determining a contribution rate percentage corresponding to each element through the following formula:

[0205] ;

[0206] Determining the element with the largest contribution rate percentage as the faulty element.

[0207] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented: sending an alarm message according to an out-of-tolerance instrument transformer.

[0208] In one embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0209] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, the memories, databases, or other media mentioned in the embodiments provided in this application can all include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in this application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in this application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logics based on quantum computing, artificial intelligence (AI) processors, etc., without limitation.

[0210] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope recorded in this application.

[0211] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A method for monitoring the out-of-tolerance of a capacitive voltage transformer, characterized in that, The method includes: Performing kernel principal component analysis based on historical sample data to determine the threshold of the squared prediction error statistic corresponding to the historical sample data; the historical sample data includes multiple sample data collected within a historical time period, and the sample data includes three-phase voltage data within a transformer group; Calculating the squared prediction error statistic of the currently collected current sample data; Determining that the operating state of the transformer group is normal when the squared prediction error statistic is less than or equal to the threshold of the squared prediction error statistic; Calculating the contribution rate of each element in the current sample data to the squared prediction error statistic and determining the fault elements when the squared prediction error statistic is greater than the threshold of the squared prediction error statistic; Locating the out-of-tolerance transformers in the transformer group according to the fault elements; 2. The method according to claim 1, characterized in that, The performing kernel principal component analysis based on historical sample data to determine the threshold of the squared prediction error statistic corresponding to the historical sample data includes: Constructing a kernel matrix based on the historical sample data according to a preset kernel function; Performing centering processing on the kernel matrix, and performing eigenvalue decomposition on the centered kernel matrix to determine multiple eigenvalues and multiple eigenvectors; Selecting k eigenvalues from the multiple eigenvalues to form a variance contribution diagonal matrix, and selecting k eigenvectors from the multiple eigenvectors to form a principal component direction matrix; For the i-th sample data, calculating the principal component projection of the i-th sample data according to the i-th row vector of the centered kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; Based on the preset kernel function, determining the self-kernel function value of the i-th sample data, and determining the squared prediction error statistic of the i-th sample data according to the self-kernel function value and the principal component projection; Determining the threshold of the squared prediction error statistic corresponding to a preset confidence level through a kernel density estimation strategy according to the squared prediction error statistics corresponding to the multiple sample data; 3. The method according to claim 2, wherein The method further includes: Calculating a mapping matrix according to the centered kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; Determining the mapping feature of the i-th sample data according to the mapping matrix, and calculating the residual vector of the i-th sample data according to the mapping feature, the mapping matrix, and the principal component projection; Calculating the squared prediction error statistic of the i-th sample data according to the residual vector of the i-th sample data; 4. The method according to claim 2, wherein The selecting k eigenvalues from the multiple eigenvalues to form a variance contribution diagonal matrix, and selecting k eigenvectors from the multiple eigenvectors to form a principal component direction matrix includes: Selecting the first k largest eigenvalues from the multiple eigenvalues so that the cumulative contribution rate reaches a set threshold, and forming a variance contribution diagonal matrix according to the selected k eigenvalues; Selecting the eigenvectors corresponding to the first k largest eigenvalues from the multiple eigenvectors, and forming a principal component direction matrix according to the selected k eigenvectors; 5. The method according to claim 2, wherein The calculating the squared prediction error statistic of the current sample data includes: Construct a current kernel matrix of the current sample data and the historical sample data based on the preset kernel function; Perform centering processing on the current kernel matrix; Calculate the current principal component projection of the current sample data according to the centered current kernel matrix, the variance contribution diagonal matrix, and the principal component direction matrix; Based on the preset kernel function, determine the current self-kernel function value of the current sample data, and determine the squared prediction error statistic of the current sample data according to the current self-kernel function value and the current principal component projection; 6. The method according to any one of claims 1 to 5, characterized in that Calculate the contribution rate of each element in the current sample data to the squared prediction error statistic through the following formula: ； Among them, represents the contribution rate of the i-th element to the squared prediction error statistic, represents the value of the i-th element, represents the j-th sample data in the historical sample data, represents the current sample data, represents the kernel function; Correspondingly, the determining the faulty element includes: Determine the contribution rate percentage corresponding to each element through the following formula: ； Determine the element with the largest contribution rate percentage as the faulty element; 7. The method according to any one of claims 1 to 5, characterized in that After locating the out-of-tolerance instrument transformer in the instrument transformer group according to the faulty element, the method further includes: Send an alarm message according to the out-of-tolerance instrument transformer; 8. An out-of-tolerance monitoring device for a capacitive voltage transformer, characterized in that, The device includes: A historical data processing module, configured to perform kernel principal component analysis based on historical sample data to determine a threshold of the squared prediction error statistic corresponding to the historical sample data; the historical sample data includes multiple sample data collected during a historical time period, and the sample data includes three-phase voltage data in the instrument transformer group; A real-time data processing module, configured to calculate the squared prediction error statistic of the currently collected current sample data; An out-of-tolerance monitoring module, configured to determine that the operation state of the instrument transformer group is normal when the squared prediction error statistic is less than or equal to the threshold of the squared prediction error statistic; when the squared prediction error statistic is greater than the threshold of the squared prediction error statistic, calculate the contribution rate of each element in the current sample data to the squared prediction error statistic, determine the faulty element, and locate the out-of-tolerance instrument transformer in the instrument transformer group according to the faulty element; 9. An electronic device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented; 10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, the steps of the method according to any one of claims 1 to 7 are implemented;

Images