Current transformer error online detection method and system based on data secondary correction

Abstract

The invention discloses a data secondary correction-based current transformer error state online detection method and system. The method comprises the following steps of: acquiring historical data and real-time test data of a current transformer to form current data; performing primary correction on the test data to obtain primary correction data of the current data; performing secondary correction on the primary correction data to obtain secondary correction data of the current data; calculating a Q statistic by using the secondary correction data, comparing the Q statistic with a statistic control limit Qc, and judging whether the error state of the current transformer is abnormal or not; the method has the advantage that the online detection of the error state of the current transformer can be realized under the condition that the fluctuation of the unbalance degree is relatively large.

Description

Technical Field

[0001] The present invention relates to the field of current transformer error detection, and particularly to an online detection method and system for current transformer errors based on data secondary correction. Background Art

[0002] A current transformer is a sensing device used to sense large current signals in a power system. Its working principle is to convert the large current signal on the primary side into a small current signal on the secondary side in proportion, providing accurate primary current information for measurement, control, and protection devices. As a signal sensing device, error is a key indicator to measure the working performance of a current transformer. Under the influence of various complex factors such as harmonics, DC bias, remanence, leakage current, and temperature, there is a risk of error degradation in current transformers, which affects the safe, stable, and reliable operation of the power system. Therefore, it is necessary to timely detect the error state of in-service current transformers.

[0003] The traditional method for detecting the error of a current transformer is to compare and calibrate it with a standard current transformer regularly under offline or live conditions. However, the error detection method of regular comparison with a standard current transformer lacks real-time performance and cannot timely detect the error state of the current transformer. At the same time, there are many problems such as a large amount of test work and high safety hazards in on-site error verification work based on such methods. Therefore, there is a need to explore an online detection method for current transformer errors under the condition of being separated from the standard current transformer.

[0004] Chinese Patent Publication No. CN107255792A discloses an online monitoring method and system for the error of an electronic voltage transformer. The method includes: collecting measurement data of three-phase electronic transformers in the operating state, establishing a residual model to calculate its statistical characteristic quantities, and comparing them with statistical control limits to determine whether the error state of the electronic voltage transformer is normal. The online monitoring method for the error state of the electronic voltage transformer provided by this solution can realize real-time online monitoring of the error state of the electronic voltage transformer by means of a method that is separated from the traditional standard device for regular detection. However, this solution is difficult to analyze current data with relatively large unbalance fluctuations, and thus it is difficult to realize the online detection of the error state of the current transformer. Summary of the Invention

[0005] The technical problem to be solved by the present invention is that the existing online detection method for current transformer errors is difficult to analyze current data with relatively large unbalance fluctuations, and thus it is difficult to realize the online detection of the error state of the current transformer.

[0006] The present invention solves the above technical problems by the following technical means: An online detection method for current transformer errors based on data secondary correction, including:

[0007] S1. Collect the historical data and real-time test data of the current transformer to form current data;

[0008] S2. Calculate the three-phase unbalance degree sequence based on the current data, calculate the current standard deviation using the three-phase unbalance degree sequence, and compare it with the data correction standard. If the current standard deviation is greater than the data correction standard, perform a first correction on the current data to obtain the first corrected data;

[0009] S3. Calculate the historical current deviation of each phase of the historical data, perform empirical wavelet transform on it, and establish an autoregressive moving average prediction model to predict the theoretical prediction value of the historical current deviation of each phase; perform inverse calculation on the theoretical prediction value of the historical current deviation of each phase to obtain the theoretical prediction value of the historical current data; perform a second correction on the first corrected data of the current data based on the theoretical prediction value of the historical current data to obtain the second corrected data;

[0010] S4. Calculate the Q statistic using the second corrected data, compare the Q statistic with the statistical control limit Q c and determine whether the error state of the current transformer is abnormal.

[0011] Further, S2 includes:

[0012] S2.1. Divide the maximum value obtained by taking the absolute value of the difference between the A, B, and C phase data of the current data and the average value of the three-phase current of the current data by the average value of the three-phase current of the current data to obtain the three-phase unbalance degree sequence CUF;

[0013] S2.2. Calculate the current standard deviation using the three-phase unbalance degree sequence CUF and the average value of the three-phase unbalance degree sequence CUF;

[0014] S2.3. Determine the data correction standard;

[0015] S2.4. Compare the current standard deviation with the data correction standard. If the current standard deviation is less than or equal to the data correction standard, there is no need to correct the current data. If the current standard deviation is greater than the data correction standard, perform a first correction on the current data to obtain the first corrected data.

[0016] Even further, S2.4 includes:

[0017] S2.4.1. If the current standard deviation is less than the data correction standard, determine that the current data does not need to be corrected, and execute S2.4.3; if the current standard deviation is greater than the data correction standard, determine that the current data needs to be corrected, and execute S2.4.2;

[0018] S2.4.2. Perform an affine transformation on the current data, use the result of the affine transformation as the current data, and return to S2.1. The affine transformation is to multiply the current data by a linear transformation parameter and then add a translation transformation parameter.

[0019] S2.4.3. Output the current current data as the first corrected data of the current data.

[0020] Further, S3 includes:

[0021] S3.1. Subtract the three-phase current average value of the historical data from each-phase data of the historical data and then divide by the three-phase current average value of the historical data to obtain the historical current deviation of each phase of the historical data.

[0022] S3.2. Perform an empirical wavelet transform on the historical current deviation of each phase of the historical data to obtain the modal components of the historical current deviation of each phase.

[0023] S3.3. Establish an autoregressive moving average prediction model, use the modal components of the historical current deviation of each phase as the past observations of the autoregressive moving average prediction model, and predict the theoretical prediction values of the historical current deviation of each phase.

[0024] S3.4. Perform an inversion calculation on the theoretical prediction values of the historical current deviation of each phase. The inversion calculation formula is that the theoretical prediction value of the historical current data minus the three-phase current average value of the historical data and then divided by the three-phase current average value of the historical data is equal to the theoretical prediction value of the historical current deviation of each phase. Substitute the theoretical prediction values of the historical current deviation of each phase into the inversion calculation formula to solve for the theoretical prediction value of the historical current data.

[0025] S3.5. Based on the theoretical prediction value of the historical current data, perform a second correction on the first corrected data of the current data to obtain the second corrected data.

[0026] Even further, S3.5 includes:

[0027] S3.5.1. Divide the theoretical prediction value of the historical current data by the test data, and calculate the three-phase average value of the quotient value as the mean feature of the quotient value.

[0028] S3.5.2. Multiply the mean feature of the quotient value by the first corrected data to obtain the second corrected data of the current data.

[0029] Further, S4 includes:

[0030] S4.1. Perform a standardization process on the second corrected data to obtain a standardization matrix.

[0031] S4.2. Decompose the standardized matrix using the principal component analysis method into a principal component model and a residual model. The principal component model is the product of the principal component score matrix and the transpose of the principal component loading matrix, and the residual model is the product of the residual score matrix and the transpose of the residual loading matrix;

[0032] S4.3. Perform singular value decomposition on the covariance matrix of the standardized matrix to obtain multiple eigenvalues. Sort the eigenvalues from largest to smallest, calculate the cumulative percentage of the first p1 eigenvalues, such that the cumulative percentage calculated from the first p1 eigenvalues is just greater than a set limit value, and obtain the corresponding p1. Then construct the principal component loading matrix based on the first p1 eigenvalues, and construct the residual loading matrix from the remaining eigenvalues;

[0033] S4.4. Calculate the projection of the standardized matrix on the residual loading matrix to obtain the residual score matrix;

[0034] S4.5. Multiply the residual score matrix, the transpose of the residual loading matrix, and the transpose of the standardized matrix to obtain the Q statistic of the standardized matrix; If Q < Q c , then it is judged that the three-phase current transformer is operating normally at this time. If Q > Q c , then it is judged that the error state of the three-phase current transformer is in an abnormal state.

[0035] Furthermore, the on-line current transformer error detection method based on data secondary correction further includes S5 for model training. S5 includes:

[0036] When the current round of b test data is detected, count the number of times m that the Q statistic in the current round of b test data exceeds the statistic control limit Q c . If m < 30%b, then discard the earliest b test data of the current data, that is, the b test data of the first round, and update the next round of test data of the current data, and continue to execute S2; If m ≥ 30%b, then discard the b test data of the current round and update the next round of b test data of the current data, and continue to execute S2.

[0037] The present invention also provides an on-line current transformer error detection system based on data secondary correction, including:

[0038] A data acquisition module for acquiring historical data and real-time test data of the current transformer to form current data;

[0039] A first correction module for calculating a three-phase unbalance degree sequence based on the current data, calculating the current standard deviation using the three-phase unbalance degree sequence and comparing it with the data correction standard. If the current standard deviation is greater than the data correction standard, then perform a first correction on the current data to obtain first-corrected data;

[0040] A second correction module, configured to calculate the historical current deviation of each phase of historical data, perform empirical wavelet transform on the historical current deviation of each phase and establish an autoregressive moving average prediction model to predict the theoretical prediction value of the historical current deviation of each phase; perform inverse calculation on the theoretical prediction value of the historical current deviation of each phase to obtain the theoretical prediction value of the historical current data; perform secondary correction on the primary correction data of the current data based on the theoretical prediction value of the historical current data to obtain secondary correction data;

[0041] An on-line detection module, configured to calculate a Q statistic using the secondary correction data, and compare the Q statistic with a statistic control limit Q c to determine whether the error state of the current transformer is abnormal.

[0042] Further, the first correction module is further configured to:

[0043] S2.1: Divide the maximum value obtained by taking the absolute value of the difference between the A, B, and C phase data of the current data and the average value of the three-phase current of the current data by the average value of the three-phase current of the current data to obtain a three-phase unbalance degree sequence CUF;

[0044] S2.2: Calculate the current standard deviation using the three-phase unbalance degree sequence CUF and the average value of the three-phase unbalance degree sequence CUF;

[0045] S2.3: Determine the data correction standard;

[0046] S2.4: Compare the current standard deviation with the data correction standard. If the current standard deviation is less than or equal to the data correction standard, there is no need to correct the current data. If the current standard deviation is greater than the data correction standard, perform primary correction on the current data to obtain primary correction data.

[0047] Furthermore, S2.4 includes:

[0048] S2.4.1: If the current standard deviation is less than the data correction standard, determine that the current data does not need to be corrected, and execute S2.4.3; if the current standard deviation is greater than the data correction standard, determine that the current data needs to be corrected, and execute S2.4.2;

[0049] S2.4.2: Perform an affine transformation on the current data, use the result of the affine transformation as the current data, and return to S2.1; the affine transformation is to multiply the current data by a linear transformation parameter and then add a translation transformation parameter;

[0050] S2.4.3: Output the current current data as the primary correction data of the current data.

[0051] Further, the second correction module is further configured to:

[0052] S3.1. Subtract the average value of the three-phase current of the historical data from each phase data of the historical data, and then divide by the average value of the three-phase current of the historical data to obtain the historical current deviation of each phase of the historical data;

[0053] S3.2. Perform empirical wavelet transform on the historical current deviation of each phase of the historical data to obtain the modal components of the historical current deviation of each phase;

[0054] S3.3. Establish an autoregressive moving average prediction model, use the modal components of the historical current deviation of each phase as the past observations of the autoregressive moving average prediction model, and predict the theoretical prediction values of the historical current deviation of each phase;

[0055] S3.4. Perform inversion calculation on the theoretical prediction values of the historical current deviation of each phase. The inversion calculation formula is that the theoretical prediction value of the historical current data minus the average value of the three-phase current of the historical data and then divided by the average value of the three-phase current of the historical data is equal to the theoretical prediction value of the historical current deviation of each phase. Substitute the theoretical prediction values of the historical current deviation of each phase into the inversion calculation formula to solve for the theoretical prediction values of the historical current data;

[0056] S3.5. Based on the theoretical prediction values of the historical current data, perform secondary correction on the primary corrected data of the current data to obtain the secondary corrected data.

[0057] Further, S3.5 includes:

[0058] S3.5.1. Divide the theoretical prediction values of the historical current data by the test data, and calculate the three-phase average value of the quotient as the mean feature of the quotient;

[0059] S3.5.2. Multiply the mean feature of the quotient by the primary corrected data to obtain the secondary corrected data of the current data.

[0060] Further, the online detection module is also used for:

[0061] S4.1. Perform standardization processing on the secondary corrected data to obtain a standardized matrix;

[0062] S4.2. Decompose the standardized matrix by the principal component analysis method, which is decomposed into a principal component model and a residual model. The principal component model is the product of the principal component score matrix and the transpose of the principal component loading matrix, and the residual model is the product of the residual score matrix and the transpose of the residual loading matrix;

[0063] S4.3. Perform singular value decomposition on the covariance matrix of the standardized matrix to obtain multiple eigenvalues. Sort the eigenvalues from large to small, calculate the cumulative percentage of the first p1 eigenvalues, so that the cumulative percentage calculated by the first p1 eigenvalues is just greater than the set limit value, obtain the corresponding p1, and then construct the principal component loading matrix based on the first p1 eigenvalues, and construct the residual loading matrix for the remaining eigenvalues;

[0064] S4.4. Calculate the projection of the standardized matrix on the residual load matrix to obtain the residual score matrix;

[0065] S4.5. Multiply the residual score matrix, the transpose of the residual load matrix, and the transpose of the standardized matrix to obtain the Q statistic of the standardized matrix; if Q < Q c , it is determined that the three-phase current transformer is operating normally at this time. If Q > Q c , it is determined that the error state of the three-phase current transformer is in an abnormal state.

[0066] Furthermore, the on-line current transformer error detection system based on data secondary correction further includes a training module for model training. The training module is used for:

[0067] When the current round of b test data is detected, count the number m of times that the Q statistic in the current round of b test data exceeds the statistic control limit Q c . If m < 30%b, discard the b test data in the first round of current data, update the next round of test data of the current data, and continue to execute S2; if m ≥ 30%b, discard the b test data in the current round, update the next round of b test data of the current data, and continue to execute the first correction module.

[0068] The advantages of the present invention are as follows:

[0069] (1) The present invention calculates the three-phase unbalance degree sequence based on the current data, calculates the current standard deviation using the three-phase unbalance degree sequence and compares it with the data correction standard. If the current standard deviation is greater than the data correction standard, the current data is corrected once to obtain the first correction data, so as to analyze the current data with relatively large unbalance degree fluctuations and correct it. It also performs secondary correction on the first correction data of the current data based on the theoretical prediction value of the historical current data, then performs principal component analysis on the secondary correction data and establishes a residual model, and calculates the Q statistic to judge the deviation degree of the non-principal component. The calculated Q statistic not only considers the influence of the three-phase unbalance degree but also uses secondary correction to enhance the error characteristics in the data, thus finally realizing the accurate on-line detection of the error state of the current transformer.

[0070] (2) The on-line current transformer error state detection method based on data secondary correction provided by the present invention analyzes the error state of the current transformer based on the measurement data of the current transformer itself, without using a standard current transformer, and has the advantages of low detection cost and strong real-time performance compared with the traditional current transformer error state detection method;

[0071] (3) The on-line detection method for the error state of current transformers based on data secondary correction provided by the present invention performs on-line detection of the error state of current transformers based on the measurement data of the current transformers themselves, and can effectively reduce the off-line test workload and test safety hazards of power grid operation and maintenance personnel compared with traditional current transformer error state detection methods. Description of the Drawings

[0072] Figure 1 It is a flowchart of the on-line detection method for the error state of current transformers based on data secondary correction disclosed in the embodiments of the present invention;

[0073] Figure 2 It is a structural schematic diagram of the on-line detection system for the error state of current transformers based on data secondary correction disclosed in the embodiments of the present invention. Detailed Embodiments

[0074] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0075] The on-line detection method for the error state of current transformers based on data secondary correction provided by the present invention has the three-phase current transformers in a specific interval as the detection object. The specific idea is as follows: First, affine transformation is used to preliminarily correct the three-phase current data of the nodes collected synchronously, and a new comparison standard with correlation constraints is established; Second, the empirical wavelet transform-autoregressive moving average method is used to extract the residual components containing error information, and the data is secondarily corrected in combination with the daily cycle characteristics of the system current to enhance its error characteristics; Finally, the improved moving window principal component analysis method is used to dynamically evaluate the data after secondary correction to achieve long-term on-line detection of the measurement error of current transformers. The present invention can effectively reduce the cost of current transformer error detection and the safety hazards of staff, and also provides a new idea for the on-line detection method of current transformers. The flow of the on-line detection method for the error state of current transformers based on data secondary correction proposed by the present invention is as Figure 1 shown, and specifically includes the following steps:

[0076] S1: Obtain a historical data measured by each of the three phases of the three-phase current transformer and collect in real time b test data measured by each of the three phases of the three-phase current transformer The historical data and the test data together constitute the current data Among them, x takes the A-phase, B-phase or C-phase of the three-phase current transformer.

[0077] S2: Since the fluctuation of the current unbalance degree in the same line is much larger than that of the voltage unbalance degree, the current data can be corrected once. Based on the smaller fluctuation of the corrected data, a new comparison standard for on-line detection is established, and the first corrected data of the current data is obtained. The specific steps are as follows:

[0078] S2.1: Based on the current data I x measured by the three-phase current transformer respectively, calculate the three-phase unbalance degree sequence

[0079]

[0080] Among them, I A 、I B 、I C are the A-phase, B-phase and C-phase currents in the current data I x respectively. I ave is the average value of I A 、I B 、I C . S2.2: Calculate the current standard deviation

[0081]

[0082] of the three-phase unbalance degree calculation result based on the current data. Among them, i refers to the i-th data point, and CUF ave is the average value of the three-phase unbalance degree sequence CUF.

[0083] S2.3: Determine the data correction standard RS based on the time-domain characteristics of the standard deviation of the three-phase voltage data unbalance degree under the stable operation of the power system; this data correction standard is a data set according to the short-term invariance characteristics of the three-phase voltage data in the actual power system, and can be adjusted according to needs in actual applications.

[0084] S2.4: Compare the result of the current standard deviation S CUF with the data correction standard RS to judge whether data correction is needed. The specific steps are as follows:

[0085] S2.4.1: If the current standard deviation S CUF is less than the data correction standard RS, it is judged that the current data does not need to be corrected, and S2.4.3 is executed; if the current standard deviation S CUF is greater than the data correction standard RS, it is judged that the current data needs to be corrected, and S2.4.2 is executed;

[0086] S2.4.2: Perform an affine transformation on the current data Take the affine transformation result as the current data I x , return to S2.1; where, A1 represents the linear transformation parameter, and the identity matrix E can be taken; b1 represents the translation transformation parameter, and b1 = [1, 1,..., 1] can be taken T ;

[0087] S2.4.3: Output the current current data as the first correction data I of the said current data x aff1 ;

[0088] S3: The S2 may weaken the original error information in the first correction data I x aff1 . It is necessary to combine the daily cycle time domain characteristics existing under the normal operation of the power system, perform secondary correction on the said current data, and enhance the error characteristics in the data to obtain the secondary correction data of the current data; specifically, it includes the following sub-steps:

[0089] S3.1: Calculate the historical current deviation of each phase of the said historical data represents the average value of the three-phase current of the historical data, represents the x-phase data of the historical data;

[0090] S3.2: In practice, in order to eliminate the influence of the non-stationarity of the historical current deviation CUF x his on the result of S3, it is necessary to preprocess the historical current deviation CUF of each phase x his to obtain the modal components of the historical current deviation of each phase. The preprocessing means can be empirical wavelet transform. The empirical wavelet transform belongs to the prior art, and its technical principle is as follows:

[0091] I0(t) = W x (0, t) * φ1(t),

[0092] where, in the formula, W x (0, t) is the approximation coefficient; W x (n, t) is the detail coefficient; φ1(t) is the empirical scaling function; is the empirical wavelet function; I(t) is the time series of the input historical current data I of each phase x .

[0093] In this embodiment, the empirical wavelet transform can adopt the empirical wavelet transform method described in the literature "Gilles, J. Empirical wavelet transform [J]. IEEE Transactions on Signal Processing, 2013, 61(16): 3999–4010."

[0094] S3.3: Establish an autoregressive moving average prediction model, use the historical current deviation modal components of each phase as the past observed values of the autoregressive moving average prediction model, and predict the future values, that is, use the autoregressive moving average prediction model to predict the theoretical prediction values of the historical current deviations of each phase The mathematical model of the existing autoregressive moving average model is as follows:

[0095]

[0096] Among them, p is the order of the autoregressive model; is the undetermined coefficient of the model; ε t is the error; q is the order of the moving average model; θ j (j = 1, 2,..., q) are the undetermined coefficients of the model.

[0097] In this embodiment, the autoregressive moving average prediction model can adopt the autoregressive moving average prediction model described in the literature "Box, G.E.P., Jenkins, G.M., et al. Time Series Analysis: Forecasting and Control, 5th edition [M]. Hoboken: Wiley, 2015."

[0098] S3.4: Perform inversion calculation on the theoretical prediction values of the historical current deviations of each phase The principle of the inversion calculation formula is the same as that of S3.1, so as to obtain the theoretical prediction values of the historical current data

[0099] S3.5: Based on the theoretical prediction values of the historical current data, perform secondary correction on the first correction data I of the current data x aff1 The specific steps are as follows:

[0100] S3.5.1: Take the quotient of the theoretical prediction values of the historical current data and the test data, and calculate the mean feature of the quotient value. The formula is expressed as

[0101] S3.5.2: Multiply the mean feature of the quotient value by the first correction data to obtain the second correction data I of the current datax aff2 .

[0102] S4: Dynamically and on-line detect the error state of the current transformer; specifically including the following sub-steps:

[0103] S4.1: The matrix form of the secondary correction data is Denoted as the secondary correction data matrix, perform standardization processing on the secondary correction data matrix to obtain the standardized matrix X = [X A X B X C (a+b)×3 ;

[0104] S4.2: Decompose the standardized matrix by using the principal component analysis (PCA) method according to the following formula:

[0105]

[0106] where, t i is the i-th score vector, p i T is the i-th loading vector, is the principal element model of the standardized matrix, E = T e P e T is the residual model of the standardized matrix, T is the principal element score matrix of the standardized matrix, P T is the transpose of the principal element loading matrix of the standardized matrix, T e is the residual score matrix of the standardized matrix, P e T is the transpose of the residual loading matrix of the standardized matrix.

[0107] S4.3: Perform singular value decomposition on the covariance matrix R of the standardized matrix to obtain n eigenvalues, sort the eigenvalues from large to small, and calculate the cumulative percentage of the first p1 eigenvalues where, λ j is the j-th eigenvalue, n is the total number of eigenvalues, CPV(p1) can usually set a limit value of 85%, so that when just the first p1 eigenvalues are substituted into the calculation, CPV(p1) is greater than 85%. After the limit value of CPV(p1) is determined, the corresponding p1 can be calculated, and then the principal element loading matrix P is constructed based on the first p1 eigenvalues, and the remaining (n - p1) eigenvalues are used to construct the residual loading matrix P e . The formula expression of singular value decomposition is R = X T X / (n - 1) = [PP e Λ[PP e T ​​, where Λ = diag(λ1, λ2, λ3) is the eigenvalue of the covariance matrix R, λ1 ≥ λ2 ≥ λ3, e and [P P

[0108] is the eigenmatrix composed of the eigenvectors corresponding to the eigenvalues λ1, λ2, and λ3. e S4.4: Standardize the residual score matrix T of the matrix e which is the projection of the standardized matrix X on the residual loading matrix P e , that is, T e = XP e . Having obtained the residual loading matrix P e through S4.3, substituting it into T e = XP e can obtain the residual score matrix T

[0109] S4.5: According to the formula Q = (XP e P e T )(XP e P e T ) T = XP e P e T X T calculate the Q statistic of the standardized matrix, and calculate the statistic control limit and determine the operating state of the three-phase current transformer at this time; where C α represents the critical value of the normal distribution at the given significance level α, and the parameter θ i represents the i-th power of the non-principal eigenvalue; n represents the total number of eigenvalues after singular value decomposition; k represents the number of retained principal components; j represents the j-th eigenvalue, which is a non-principal eigenvalue; represents the i-th power of the j-th eigenvalue.

[0110] If Q < Q c , it is determined that the three-phase current transformer is operating normally at this time;

[0111] If Q > Q c , it is determined that the error state of the three-phase current transformer is in an abnormal state.

[0112] S5. After detecting the b test data in the current round, calculate the number m of times the statistic Q exceeds the statistic control limit Q c and determine the update mode of the online detection model:

[0113] If m < 30%b, discard the earliest b test data of the current data, that is, the b test data of the first round, and update the next round of test data of the current data, then continue to execute S2. m < 30%b indicates that the state of the three-phase current transformer is basically normal. At this time, in order to prevent the online detection model from falling into a local optimum, the earliest test data is removed and new test data is added to improve the model training accuracy.

[0114] If m ≥ 30%b, discard the b test data of the current round, and update the next b test data of the current data, then continue to execute S2. m ≥ 30%b indicates that the error state of the three-phase current transformer is in an abnormal state. At this time, in order to prevent detection anomalies caused by incorrect test data, the test data of the current round is removed and the test data is updated to improve the detection accuracy of the model.

[0115] Through the above technical solutions, the online detection method for the error state of a current transformer based on data secondary correction provided by the present invention analyzes the error state of the current transformer based on the method of the current transformer's own measurement data, without using a standard current transformer. Compared with the traditional method for detecting the error state of a current transformer, it has the advantages of low detection cost and strong real-time performance; compared with the traditional method for detecting the error state of a current transformer, it can effectively reduce the off-line test workload and test safety hazards of power grid operation and maintenance personnel.

[0116] Embodiment 2

[0117] Based on Embodiment 1, Embodiment 2 of the present invention further provides an online detection system for the error of a current transformer based on data secondary correction. Its detection object is the three-phase current transformer in a specific interval, and its system composition is as Figure 2 shown, including a data correction module and an online detection module connected to each other. The data correction module includes a first correction module and a second correction module; the data correction module is used to correct the three-phase current data, establish a new comparison standard suitable for online detection of the error state of the current transformer based on the operating characteristics of the power system and data transformation, and send the corrected data to the online detection module; the online detection module is used to realize the online detection of the error state of the current transformer and the update of the current acquisition data by the detection model, and send the updated current data to the data correction module. The system also includes a data acquisition module and a training module, which are not shown in Figure 2 here. The data acquisition module is connected to the first correction module, the first correction module and the second correction module are connected, the second correction module is connected to the online detection module, and the training module trains the online detection model composed of the data acquisition module, the data correction module and the online detection module. The online detection system specifically includes:

[0118] A data acquisition module, which is used to collect the historical data and real-time test data of the current transformer to form current data;

[0119] The first correction module is used to calculate the three-phase unbalance degree sequence based on the current data, calculate the current standard deviation using the three-phase unbalance degree sequence and compare it with the data correction standard. If the current standard deviation is greater than the data correction standard, the current data is corrected once to obtain the first corrected data.

[0120] The second correction module is used to calculate the historical current deviation of each phase of the historical data, perform empirical wavelet transform on it and establish an autoregressive moving average prediction model to predict the theoretical predicted value of the historical current deviation of each phase; perform inverse calculation on the theoretical predicted value of the historical current deviation of each phase to obtain the theoretical predicted value of the historical current data; perform secondary correction on the first corrected data of the current data based on the theoretical predicted value of the historical current data to obtain the second corrected data.

[0121] The on-line detection module is used to calculate the Q statistic using the second corrected data, compare the Q statistic with the statistic control limit Q c for comparison to determine whether the error state of the current transformer is abnormal.

[0122] Specifically, the first correction module is further used for:

[0123] S2.1: The maximum value obtained by taking the absolute value of the difference between the A, B, and C phase data of the current data and the average value of the three-phase current of the current data is divided by the average value of the three-phase current of the current data to obtain the three-phase unbalance degree sequence CUF.

[0124] S2.2: Calculate the current standard deviation using the three-phase unbalance degree sequence CUF and the average value of the three-phase unbalance degree sequence CUF.

[0125] S2.3: Determine the data correction standard.

[0126] S2.4: Compare the current standard deviation with the data correction standard. If the current standard deviation is less than or equal to the data correction standard, there is no need to correct the current data. If the current standard deviation is greater than the data correction standard, the current data is corrected once to obtain the first corrected data.

[0127] More specifically, S2.4 includes:

[0128] S2.4.1: If the current standard deviation is less than the data correction standard, it is judged that the current data does not need to be corrected, and S2.4.3 is executed; if the current standard deviation is greater than the data correction standard, it is judged that the current data needs to be corrected, and S2.4.2 is executed.

[0129] S2.4.2: Perform an affine transformation on the current data, use the result of the affine transformation as the current data, and return to S2.1; the affine transformation is to multiply the current data by a linear transformation parameter and then add a translation transformation parameter.

[0130] S2.4.3. Output the current data at present as the first corrected data of the current data.

[0131] Specifically, the second correction module is also used for:

[0132] S3.1. Subtract the three-phase current average value of the historical data from the phase data of the historical data, and then divide by the three-phase current average value of the historical data to obtain the historical current deviation of each phase of the historical data;

[0133] S3.2. Perform empirical wavelet transform on the historical current deviation of each phase of the historical data to obtain the modal components of the historical current deviation of each phase;

[0134] S3.3. Establish an autoregressive moving average prediction model, use the modal components of the historical current deviation of each phase as the past observed values of the autoregressive moving average prediction model, and predict to obtain the theoretical prediction values of the historical current deviation of each phase;

[0135] S3.4. Perform inversion calculation on the theoretical prediction values of the historical current deviation of each phase. The inversion calculation formula is that the theoretical prediction value of the historical current data minus the three-phase current average value of the historical data and then divided by the three-phase current average value of the historical data is equal to the theoretical prediction value of the historical current deviation of each phase. Substitute the theoretical prediction values of the historical current deviation of each phase into the inversion calculation formula to solve and obtain the theoretical prediction values of the historical current data;

[0136] S3.5. Based on the theoretical prediction values of the historical current data, perform secondary correction on the first corrected data of the current data to obtain the secondary corrected data.

[0137] More specifically, S3.5 includes:

[0138] S3.5.1. Divide the theoretical prediction values of the historical current data by the test data, and calculate the three-phase average value of the quotient value as the mean feature of the quotient value;

[0139] S3.5.2. Multiply the mean feature of the quotient value by the first corrected data to obtain the secondary corrected data of the current data.

[0140] Specifically, the online detection module is also used for:

[0141] S4.1. Perform standardization processing on the secondary corrected data to obtain a standardization matrix;

[0142] S4.2. Decompose the standardization matrix by the principal component analysis method, decompose it into a principal component model and a residual model. The principal component model is the product of the principal component score matrix and the transpose of the principal component loading matrix, and the residual model is the product of the residual score matrix and the transpose of the residual loading matrix;

[0143] S4.3. Perform singular value decomposition on the covariance matrix of the standardized matrix to obtain multiple eigenvalues. Sort the eigenvalues from largest to smallest, calculate the cumulative percentage of the first p1 eigenvalues, such that the cumulative percentage calculated from the first p1 eigenvalues is just greater than the set limit value, and obtain the corresponding p1. Then, construct a principal component loading matrix based on the first p1 eigenvalues, and construct a residual loading matrix from the remaining eigenvalues;

[0144] S4.4. Calculate the projection of the standardized matrix on the residual loading matrix to obtain a residual score matrix;

[0145] S4.5. Multiply the residual score matrix, the transpose of the residual loading matrix, and the transpose of the standardized matrix to obtain the Q statistic of the standardized matrix; If Q < Q c , then it is judged that the three-phase current transformer is operating normally at this time. If Q > Q c , then it is judged that the error state of the three-phase current transformer is in an abnormal state.

[0146] Specifically, the on-line detection system for the current transformer error based on data secondary correction further includes a training module for model training. The training module is used for:

[0147] When the detection of the b test data in the current round is completed, count the number m of times that the Q statistic in the b test data of the current round exceeds the statistic control limit Q c . If m < 30%b, discard the b test data in the first round of the current data and update the next-round test data of the current data, and continue to execute S2; If m ≥ 30%b, discard the b test data in the current round and update the next-round b test data of the current data, and continue to execute the first correction module.

[0148] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: They can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; And these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A current transformer error online detection method based on secondary data correction, characterized in that: include: S1, collecting historical data and real-time test data of current transformer to form current data; S2, calculating the three-phase unbalance sequence based on the current data, using the three-phase unbalance sequence to calculate the current standard deviation and comparing it with the data correction standard, if the current standard deviation is greater than the data correction standard, correcting the current data once to obtain primary correction data; S3, calculating the historical current deviation of each phase of the historical data, performing empirical wavelet transform on it and establishing an autoregressive moving average prediction model to predict the theoretical prediction value of the historical current deviation of each phase; The theoretical prediction value of the historical current data is obtained by inverse calculation of the theoretical prediction value of the historical current deviation of each phase; Performing secondary correction on the primary correction data of the current data based on the theoretical prediction value of the historical current data to obtain secondary correction data; S4. Calculate the Q statistic using the secondary correction data and compare the Q statistic with the statistical control limit Q c Compare and determine whether the current transformer error state is abnormal.

2. The current transformer error online detection method based on data secondary correction according to claim 1 is characterized in that S2 include: S2.

1. Divide the absolute value of the three-phase data A, B, and C of the current data by the three-phase current average value of the current data to obtain the maximum value and the three-phase current average value of the current data to obtain the three-phase unbalance degree sequence CUF; S2.2, using the three-phase unbalance sequence CUF and the average value of the three-phase unbalance sequence CUF to calculate the standard deviation to obtain the current standard deviation; S2.3, determine the data correction standard; S2.

4. Compare the current standard deviation with the data correction standard. If the current standard deviation is less than or equal to the data correction standard, there is no need to correct the current data. If the current standard deviation is greater than the data correction standard, correct the current data once to obtain corrected data.

3. The current transformer error online detection method based on secondary data correction according to claim 2 is characterized in that: S2.4 includes: S2.4.

1. If the current standard deviation is less than the data correction standard, it is determined that the current data does not need to be corrected, and S2.4.3 is executed; if the current standard deviation is greater than the data correction standard, it is determined that the current data needs to be corrected, and S2.4.2 is executed; S2.4.2, perform affine transformation on the current data, use the affine transformation result as the current data, and return to S2.1; the affine transformation is to multiply the current data by the linear transformation parameter and then add the translation transformation parameter; S2.4.

3. Output the current data as the primary correction data of the current data.

4. The current transformer error online detection method based on data secondary correction according to claim 1 is characterized in that: S3 includes: S3.

1. Subtract the average value of the three-phase current of the historical data from the phase data of the historical data and then divide it by the average value of the three-phase current of the historical data to obtain the historical current deviation of each phase of the historical data; S3.2, performing empirical wavelet transform on the historical current deviation of each phase of the historical data to obtain the modal component of the historical current deviation of each phase; S3.

3. Establish an autoregressive moving average prediction model, use the modal component of the historical current deviation of each phase as the past observation value of the autoregressive moving average prediction model, and predict the theoretical prediction value of the historical current deviation of each phase; S3.

4. Perform inverse calculation on the theoretical prediction value of the historical current deviation of each phase. The inverse calculation formula is: the theoretical prediction value of the historical current data minus the average value of the three-phase current of the historical data and then divided by the average value of the three-phase current of the historical data is equal to the theoretical prediction value of the historical current deviation of each phase. Substitute the theoretical prediction value of the historical current deviation of each phase into the inverse calculation formula to obtain the theoretical prediction value of the historical current data; S3.

5. Perform a secondary correction on the primary correction data of the current data based on the theoretical prediction value of the historical current data to obtain secondary correction data.

5. The current transformer error online detection method based on secondary data correction according to claim 4 is characterized in that: S3.5 includes: S3.5.

1. The theoretical prediction value of the historical current data is divided by the test data, and the three-phase mean of the quotient is calculated as the mean characteristic of the quotient; S3.5.

2. Product the mean characteristic of the quotient value with the primary correction data to obtain the secondary correction data of the current data.

6. The current transformer error online detection method based on data secondary correction according to claim 1 is characterized in that: S4 includes: S4.1, standardize the secondary correction data to obtain a standardized matrix; S4.2, decomposing the standardized matrix by principal component analysis into a principal component model and a residual model, wherein the principal component model is the product of the transpose of the principal component score matrix and the principal component load matrix, and the residual model is the product of the transpose of the residual score matrix and the residual load matrix; S4.3, perform singular value decomposition on the covariance matrix of the standardized matrix to obtain multiple eigenvalues, sort the eigenvalues ​​from large to small, calculate the cumulative percentage of the first p1 eigenvalues, so that the cumulative percentage calculated by the first p1 eigenvalues ​​is just greater than the set limit, and obtain the corresponding p1, then construct the principal component load matrix based on the first p1 eigenvalues, and construct the residual load matrix based on the remaining eigenvalues; S4.4, calculate the projection of the standardized matrix on the residual loading matrix to obtain the residual score matrix; S4.

5. The transpose of the residual score matrix, the residual loading matrix, and the transpose of the standardized matrix are multiplied to obtain the Q statistic of the standardized matrix; if Q < Q c , then it is judged that the three-phase current transformer is operating normally at this time. If Q>Q c , it is judged that the error state of the three-phase current transformer is in an abnormal state at this time.

7. The current transformer error online detection method based on secondary data correction according to claim 6 is characterized in that: Also included is S5 for model training, which includes: After the current round of b test data is detected, the Q statistic in the current round of b test data exceeds the statistical control limit Q c The number of times is m. If m<30%b, the b test data of the first round of current data are discarded, and the next round of test data of current data are updated, and S2 is continued; if m≥30%b, the b test data of the current round are discarded, and the next round of b test data of current data are updated, and S2 is continued.

8. The current transformer error online detection system based on secondary data correction is characterized by: include: A data acquisition module is used to collect historical data and real-time test data of the current transformer to form current data; A first correction module is used to calculate a three-phase unbalance sequence based on current data, calculate a current standard deviation using the three-phase unbalance sequence and compare it with a data correction standard, and if the current standard deviation is greater than the data correction standard, correct the current data once to obtain primary correction data; The second correction module is used to calculate the historical current deviation of each phase of the historical data, perform empirical wavelet transform on it and establish an autoregressive sliding average prediction model to predict the theoretical prediction value of the historical current deviation of each phase; The theoretical prediction value of the historical current data is obtained by inverse calculation of the theoretical prediction value of the historical current deviation of each phase; Performing secondary correction on the primary correction data of the current data based on the theoretical prediction value of the historical current data to obtain secondary correction data; The online detection module is used to calculate the Q statistic using the secondary correction data and compare the Q statistic with the statistical control limit Q c Compare and determine whether the current transformer error state is abnormal.

9. The current transformer error online detection system based on secondary data correction according to claim 8 is characterized in that: The first correction module is also used to: S2.

1. Divide the absolute value of the three-phase data A, B, and C of the current data by the three-phase current average value of the current data to obtain the maximum value and the three-phase current average value of the current data to obtain the three-phase unbalance degree sequence CUF; S2.2, using the three-phase unbalance sequence CUF and the average value of the three-phase unbalance sequence CUF to calculate the standard deviation to obtain the current standard deviation; S2.3, determine the data correction standard; S2.

4. Compare the current standard deviation with the data correction standard. If the current standard deviation is less than or equal to the data correction standard, there is no need to correct the current data. If the current standard deviation is greater than the data correction standard, correct the current data once to obtain corrected data.

10. The current transformer error online detection system based on secondary data correction according to claim 9, characterized in that: S2.4 includes: S2.4.

1. If the current standard deviation is less than the data correction standard, it is determined that the current data does not need to be corrected, and S2.4.3 is executed; if the current standard deviation is greater than the data correction standard, it is determined that the current data needs to be corrected, and S2.4.2 is executed; S2.4.2, perform affine transformation on the current data, use the affine transformation result as the current data, and return to S2.1; the affine transformation is to multiply the current data by the linear transformation parameter and then add the translation transformation parameter; S2.4.

3. Output the current data as the primary correction data of the current data.

Images