Current transformer error evaluation method based on current imbalance contribution degree analysis

Abstract

The invention relates to a current transformer error evaluation method based on current imbalance contribution analysis, and the method comprises the following specific steps: collecting CT group data in a normal error state, and constructing CT group modeling current vector data; constructing a current transformer error state evaluation model based on an overdetermined linear equation set; solving a reference weight coefficient matrix in the current transformer error state evaluation model; constructing an error fluctuation characteristic quantity; calculating a Q statistic of the error fluctuation characteristic quantity, and solving a statistic control limit Qcl of the error fluctuation characteristic quantity Ef based on empirical estimation of confidence; judging whether the operation error abnormal condition of the mutual inductor exists in the line or not; and if it is judged that the current transformer in the line at the to-be-detected moment is abnormal, positioning of the abnormal transformer is realized based on the contribution degree of each phase of current transformer to the real-time statistical magnitude. According to the invention, accurate discrimination of the metering performance of the current transformer in operation is realized, and technical support is provided for troubleshooting, state evaluation and the like of the current transformer.

Description

Technical Field

[0001] The present application relates to the field of current transformer measurement error evaluation, and in particular to a current transformer error evaluation method based on current imbalance contribution analysis. Background Art

[0002] High-voltage current transformers provide current signals to electric energy meters, and their measurement errors directly affect the accuracy of electric energy measurement. According to the verification regulations such as JJG 1021-2007 "Power Transformers", the verification of metering equipment in my country adopts the method of random power outages at a fixed period, that is, according to a fixed period of 4-10 years, the primary circuits corresponding to all metering equipment to be inspected are shut down one by one, and then the transformers are fully verified using verification equipment. However, high-voltage current transformers are easily affected by multiple factors during long-term energized operation, resulting in performance degradation or even inaccuracy. Therefore, achieving long-term dynamic tracking of the metering performance of high-voltage current transformers is of great significance to the fairness and justice of electric energy trade.

[0003] At present, there are a large number of high-voltage current transformers in operation for metering in distribution networks. Taking Wuhan Power Grid as an example, there are more than 120,000 (6-35) kV current transformers in operation. Faced with the rapid growth of calibration demand of hundreds of thousands of sets per year, the traditional fixed-period power outage calibration method has problems such as large workload, high operation risk, and difficulty in implementing power outages, which makes it impossible to timely discover the out-of-tolerance problems in the operation of distribution network transformers, which is easy to cause trade disputes.

[0004] Therefore, it is urgent to carry out research on online operation monitoring technology of high-voltage current transformers for distribution network metering, and to carry out effective operation management and supervision of the online metering performance of high-voltage current transformers in distribution networks. Summary of the invention

[0005] The purpose of the embodiments of the present application is to provide a current transformer error evaluation method based on current imbalance contribution analysis, which uses the Q statistical method to implement group CT error state evaluation and uses the contribution to locate abnormal line CTs.

[0006] To achieve the above objectives, an embodiment of the present application provides a current transformer error evaluation method based on current imbalance contribution analysis, comprising the following specific steps:

[0007] Collect CT group data with qualified error status and construct CT group modeling current vector data;

[0008] Using the current vector data of CT group modeling, a current transformer error state assessment model based on an overdetermined linear equations model is constructed;

[0009] The regularized iterative algorithm is used to solve the benchmark weight coefficient matrix X in the current transformer error state assessment model. standard ;

[0010] Using the benchmark weight coefficient matrix, the error fluctuation characteristic quantity E is constructed f ;

[0011] Calculate the error fluctuation characteristic E f The Q statistic is used to solve the error fluctuation characteristic quantity E based on the empirical estimate of confidence level. f The statistical control limit Q cl ;

[0012] Obtain the real-time measurement data of the CT, calculate the Q statistic of the real-time data, and compare it with the statistical control limit to determine whether there is abnormal operation error of the transformer in the line;

[0013] If it is determined that there is an abnormal current transformer in the line at the time of testing, the abnormal current transformer is located based on the contribution of each phase current transformer to the real-time statistics.

[0014] The acquisition of CT group data with qualified error states and the construction of CT group modeling current vector data are specifically as follows:

[0015] There will be a difference between the primary current value and the secondary current value of the current transformer CT, which is divided into ratio error f i With phase difference δ i , as shown in formula (1) and (2):

[0016]

[0017] δ=φ2-φ1 (2)

[0018] In the formula, K i is the rated transformation ratio of CT, I1 and φ1 are the amplitude and phase of the actual primary current of CT, I2 and φ2 are the amplitude and phase of the actual secondary current, according to the definition of the ratio difference and phase difference of current transformer, the relationship between the primary current measurement value and the secondary current measurement value is derived as follows:

[0019]

[0020]

[0021] in, represents the measured value of the primary current of the CT on the i-th line, represents the amplitude of the measured current on the primary side of the CT on the i-th line, φ 1i represents the phase of the measured current value of the primary side of the CT on the i-th line, represents the amplitude of the CT secondary current measurement value on the i-th line, φ 2i represents the phase of the CT secondary current measurement value on the i-th line, δ irepresents the angle difference of CT on the i-th line, f i represents the ratio difference of CT on the i-th line, r represents the phase sequence, and i represents the i-th line.

[0022] The current transformer error state evaluation model based on the overdetermined linear equations model is constructed by using the CT group modeling current vector data as follows:

[0023] According to Kirchhoff's current law, the vector sum of the primary current measurement values ​​of each line at a node is 0, so we can get:

[0024]

[0025]

[0026] By approximating formula (5), we can get:

[0027]

[0028] Perform offline calibration on any one of the n lines and measure the known ratio error f r , angle difference δ r , and then perform transposition processing on formula (7):

[0029]

[0030] Formula (8) is equivalent to:

[0031] AX=B (9)

[0032] This equation is an overdetermined linear equation system, where A represents the amplitude matrix of the CT secondary side current measurement value on the line, X represents the weight coefficient matrix of the CT secondary side current measurement value amplitude, and B represents the constant term of the known ratio difference angle difference, specifically:

[0033]

[0034]

[0035]

[0036] The regularized iterative algorithm is used to solve the benchmark weight coefficient matrix X in the current transformer error state assessment model. standard Specifically:

[0037] The solution process of the regularized iterative algorithm is as follows:

[0038] Multiply (9) by A on the left T , we can get:

[0039] A T AX=AT B (13)

[0040] In the coefficient matrix A T Adding a regularization parameter α to the elements on the main diagonal of A yields:

[0041] (A T A+αI)X=A T B+αX (14)

[0042] Where I is A T The identity matrix of the same order as A,

[0043] Continuing to transform formula (14), we can get:

[0044] X (k+1) =(A T A+αI) -1 (A T B+αX k ) (15)

[0045] Where k represents the number of iterations,

[0046] Δ=X (k+1) -X k (16)

[0047] When Δ is a minimum constant, solve the benchmark weight coefficient matrix X standard .

[0048] The reference weight coefficient matrix is ​​used to construct the error fluctuation feature value E f Specifically,

[0049] Error characteristic fluctuation E f :

[0050]

[0051]

[0052]

[0053] Indicates the measured current, S pre Represents the estimated and current.

[0054] The calculation error fluctuation characteristic quantity E f The Q statistic is used to solve the error fluctuation characteristic quantity E based on the empirical estimate of confidence level. f The statistical control limit Q cl Specifically,

[0055] S51: Error fluctuation characteristic quantity E fPerform standardized data processing to obtain a standardized data matrix As well as the mean vector b and the variance matrix Σ;

[0056]

[0057] Among them, 1 n =[1,1,…,1] T ∈Φ n×1 , b=(E f ) T 1 n / n is the mean vector of the error fluctuation characteristic quantity, is the variance matrix of the error fluctuation characteristic quantity;

[0058] S52: Standardized matrix Perform singular value decomposition and calculate its eigenvalues ​​λ1,λ2,…,λ m and the corresponding eigenvector [PP e ];

[0059]

[0060] In the formula is the score matrix, is the load matrix;

[0061] For any i' and j', when i'≠j', we have Established, that is, each score vector satisfies mutual orthogonality, and similarly, each load vector also satisfies mutual orthogonality, that is:

[0062]

[0063] According to the characteristics of the sampled voltage data decomposition, both sides of equation (22) are multiplied by p i‘ We can get:

[0064]

[0065] From a geometric point of view, we can know from formula (23) that the score vector t i‘ is the data matrix In the corresponding load vector p i‘ The length of the mapping on represents the data matrix The proportion of the corresponding load vector is arranged in descending order according to its length, as shown in the following formula:

[0066] ||t1||>||t2||>…>||t m || (24)

[0067] Then the load vector p1 is the data matrix The direction with the largest change, p2 is the data matrix The second largest direction of change, p m The data matrix The direction with the smallest change,

[0068] If there is some linear correlation between the variables in the data matrix, then The change will be mainly due to the largest loads The projection in other directions will be small. The following principal component decomposition can be performed:

[0069]

[0070] In the formula, is the data matrix The principal subspace model of is the data matrix The residual subspace model of , T is the principal component score matrix, P is the principal component loading matrix, T e is the residual score matrix, P e is the residual loading matrix,

[0071] Load matrices P and P e The data matrix can be The covariance matrix R of is obtained by singular value decomposition, as shown below:

[0072]

[0073] In the formula, Λ=diag(λ1,λ2,…,λ m ), λ1≧λ2≧…≧λ m is the eigenvalue of the covariance matrix R, is the load vector composed of the corresponding eigenvectors. The larger the eigenvalue, the stronger the correlation of the represented variables.

[0074] S53: Select the number of principal components p according to the cumulative contribution percentage of variance CPV method, and determine the load matrix P of the principal element subspace and the load matrix P of the residual subspace e ;

[0075] The CPV method determines the number of principal components by calculating the cumulative variance percentage of the first p principal components. The calculation formula is:

[0076]

[0077] S54: Select the test confidence β and calculate the statistical threshold Q of the Q statistic cl

[0078] Q clis the statistical control threshold with a significance level of β, calculated as follows:

[0079]

[0080] In the formula, C β is the critical value of the normal distribution at the detection level β.

[0081] The method of obtaining the real-time measurement data of the CT, calculating the Q statistic of the real-time data, and comparing it with the statistical control limit to determine whether there is an abnormal operation error of the transformer in the line is specifically as follows:

[0082] CT real-time measurement data, first calculate the real-time error fluctuation characteristic quantity E f , and then calculate the Q statistic of the process information according to formula (29): if it is less than the statistical threshold Q c , indicating that the current transformers on each line are in normal operation; if it is greater than the statistical threshold Q c , indicating that the current transformer on the line is in abnormal operation state.

[0083]

[0084] Among them, E f Represents the real-time error fluctuation characteristic, P e Represents the loading matrix of the residual subspace.

[0085] If it is determined that there is an abnormal current transformer in the circuit at the time of testing, based on the contribution of each phase current transformer to the real-time statistics, the abnormal current transformer is located as follows:

[0086] When the real-time statistic Q exceeds the statistic threshold Q c Finally, it is necessary to determine which specific line CT metering error state has changed abnormally, that is, by calculating the role of each line in the construction of the over-limit statistics, the importance of the variable to the over-limit statistics, that is, the contribution, is obtained. It is considered that the line with the highest contribution has changed abnormally.

[0087] When the Q statistic exceeds the statistical control limit, the contribution of the current transformer measurement data on the i-th line to the Q statistic is Con i :

[0088]

[0089] represents the maximum value of the product of the weight and the current measurement value in the n-1th line, Indicates the minimum value of the product of the weight and the current measurement value in the n-1th line.

[0090] Compared with the prior art, the beneficial effects of this application are:

[0091] The present invention realizes the accuracy of judging the metering performance of the current transformer in operation, provides technical support for current transformer fault troubleshooting, status evaluation, etc., promotes the in-depth application of metering data, and advances the development of metering device status detection technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0093] Figure 1 This is a flow chart of the method of this application. DETAILED DESCRIPTION

[0094] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings.

[0095] The terms "comprises," "comprising," or any other variation thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0096] The terms "first", "second", etc. are only used to distinguish one entity or operation from another entity or operation, and should not be understood as indicating or implying relative importance, nor should they be understood as requiring or implying any such actual relationship or order between these entities or operations.

[0097] like Figure 1 As shown, a current transformer error evaluation method based on current imbalance contribution analysis is characterized by comprising the following specific steps:

[0098] S1. Collect CT group data with qualified error status and construct CT group modeling current vector data;

[0099] S2. Using the CT population modeling current vector data, a current transformer error state assessment model based on an overdetermined linear equations model is constructed;

[0100] S3. Use regularized iterative algorithm to solve the reference weight coefficient matrix X in the current transformer error state evaluation model standard ;

[0101] S4. Use the benchmark weight coefficient matrix to construct the error fluctuation feature E f ;

[0102] S5. Calculate the error fluctuation characteristic value E f The Q statistic is used to solve the error fluctuation characteristic quantity E based on the empirical estimate of confidence level. f The statistical control limit Q cl ;

[0103] S6. Obtain the real-time measurement data of the CT, calculate the Q statistic of the real-time data, and compare it with the statistical control limit to determine whether there is an abnormal operation error of the transformer in the line;

[0104] S7. If it is determined that there is an abnormal current transformer in the line at the time to be tested, the abnormal current transformer is located based on the contribution of each phase current transformer to the real-time statistics.

[0105] Collect CT group data with qualified error states, and construct CT group modeling current vector data, specifically,

[0106] There is a difference between the primary current value and the secondary current value of the current transformer (CT), which is divided into ratio error f i With phase difference δ i , as shown in formula (1) and (2):

[0107]

[0108] δ=φ2-φ1['] (2)

[0109] In the formula, K i is the rated transformation ratio of CT, I1 and φ1 are the amplitude and phase of the actual primary current of CT, and I2 and φ2 are the amplitude and phase of the actual secondary current. The ratio difference and phase difference limits of CT are shown in Table 1:

[0110] Table 1 CT ratio difference and phase difference limits

[0111]

[0112]

[0113] According to the definition of current transformer ratio difference and phase difference, the relationship between the primary current measurement value and the secondary current measurement value is derived as follows:

[0114]

[0115]

[0116] in, represents the measured value of the primary current of the CT on the i-th line, represents the amplitude of the measured current on the primary side of the CT on the i-th line, φ 1i represents the phase of the measured current value of the primary side of the CT on the i-th line, represents the amplitude of the CT secondary current measurement value on the i-th line, φ 2i represents the phase of the CT secondary current measurement value on the i-th line, δ i represents the angle difference of CT on the i-th line, f i represents the ratio difference of CT on the i-th line, r represents the phase sequence (r=A, B, C), and i represents the i-th line.

[0117] Using the CT group modeling current vector data, a current transformer error state evaluation model based on an overdetermined linear equation model is constructed. Specifically,

[0118] According to Kirchhoff's current law, the vector sum of the primary current measurement values ​​of each line at a node is 0, so we can get:

[0119]

[0120]

[0121] By approximating formula (6), we can get:

[0122]

[0123] Perform offline calibration on any one of the n lines and measure the known ratio error f r , angle difference δ r , and then perform transposition processing on formula (7):

[0124]

[0125] Formula (8) can be equivalent to:

[0126] AX=B (9)

[0127] This equation is an overdetermined linear equation system, where A represents the amplitude matrix of the CT secondary side current measurement value on the line, X represents the weight coefficient matrix of the CT secondary side current measurement value amplitude, and B represents the constant term of the known ratio difference angle difference, specifically:

[0128]

[0129]

[0130]

[0131] The regularized iterative algorithm (least squares method, partial least squares method, conjugate gradient method, etc.) is used to solve the reference weight coefficient matrix X in the current transformer error state evaluation model. standard Specifically,

[0132] The solution process of the regularized iterative algorithm is as follows:

[0133] Multiply (9) by A on the left T , we can get:

[0134] A T AX=A T B (13)

[0135] In the coefficient matrix A T Adding a regularization parameter α to the elements on the main diagonal of A yields:

[0136] (A T A+αI)X=A T B+αX (14)

[0137] Where I is A T The identity matrix of A of the same rank.

[0138] Continuing to transform formula (14), we can get:

[0139] X (k+1) =(A T A+αI) -1 (A T B+αX k ) (15)

[0140] Where k represents the number of iterations.

[0141] Δ=X (k+1) -X k (16)

[0142] When Δ is a minimum constant, we can solve the benchmark weight coefficient matrix X standard .

[0143] Using the benchmark weight coefficient matrix, the error fluctuation characteristic quantity E is constructed f Specifically,

[0144] Error characteristic fluctuation E f :

[0145]

[0146]

[0147]

[0148] Indicates the measured current, S pre Represents the estimated and current.

[0149] Calculate the error fluctuation characteristic E f The Q statistic is used to solve the error fluctuation characteristic quantity E based on the empirical estimate of confidence level. f The statistical control limit Q cl :

[0150] The specific process is:

[0151] S51: Error fluctuation characteristic quantity E f Perform standardized data processing to obtain a standardized data matrix As well as the mean vector b and the variance matrix Σ;

[0152]

[0153] Among them, 1 n =[1,1,…,1] T ∈Φ n×1 , b=(E f ) T 1 n / n is the mean vector of the error fluctuation characteristic quantity, is the variance matrix of the error fluctuation characteristic.

[0154] S52: Standardized matrix Perform singular value decomposition and calculate its eigenvalues ​​λ1,λ2,…,λ m and the corresponding eigenvector [PP e ];

[0155]

[0156] In the formula is the score matrix, is the load matrix.

[0157] For any i' and j', when i'≠j', we have Established, that is, each score vector satisfies mutual orthogonality, and similarly, each load vector also satisfies mutual orthogonality, that is:

[0158]

[0159] According to the characteristics of the sampled voltage data decomposition, both sides of equation (22) are multiplied by p i‘ We can get:

[0160]

[0161] From a geometric point of view, we can know from formula (23) that the score vector t i‘ is the data matrix In the corresponding load vector p i‘ The length of the mapping on represents the data matrix The proportion of the corresponding load vector. The score vectors are arranged in descending order according to their length, as follows:

[0162] ||t1||>||t2||>…>||t m || (24)

[0163] Then the load vector p1 is the data matrix The direction with the largest change, p2 is the data matrix The second largest direction of change, p m The data matrix The direction with the smallest change,

[0164] If there is some linear correlation between the variables in the data matrix, then The change will be mainly due to the largest loads The projection in other directions will be small. The following principal component decomposition can be performed:

[0165]

[0166] In the formula, is the data matrix The principal subspace model of is the data matrix The residual subspace model of . T is the principal component score matrix, P is the principal component loading matrix, T e is the residual score matrix, P e is the residual loading matrix.

[0167] Load matrices P and P e The data matrix can be The covariance matrix R of is obtained by singular value decomposition, as shown below:

[0168]

[0169] In the formula, Λ=diag(λ1,λ2,…,λ m ), λ1≧λ2≧…≧λ m is the eigenvalue of the covariance matrix R, is the load vector composed of the corresponding eigenvectors. The larger the eigenvalue, the stronger the correlation of the represented variables.

[0170] S53: Select the number of principal components p according to the Cumulative Percent Variance (CPV) method, and determine the loading matrix P of the principal element subspace and the loading matrix P of the residual subspace. e ;

[0171] The CPV method determines the number of principal components by calculating the cumulative variance percentage of the first p principal components. The calculation formula is:

[0172]

[0173] S54: Select the test confidence β and calculate the statistical threshold Q of the Q statistic cl

[0174] Q cl is the statistical control threshold with a significance level of β, which can be calculated as follows:

[0175]

[0176] In the formula, C β is the critical value of the normal distribution at the detection level β.

[0177] Get the real-time measurement data of CT, calculate the Q statistic of the real-time data, and compare it with the statistical control limit to determine whether there is abnormal operation error of the transformer in the line. Specifically,

[0178] CT real-time measurement data, first calculate the real-time error fluctuation characteristic quantity E f , and then calculate the Q statistic of the process information according to formula (29): if it is less than the statistical threshold Q c , indicating that the current transformers on each line are in normal operation; if it is greater than the statistical threshold Q c , indicating that the current transformer on the line is in abnormal operating state.

[0179]

[0180] If it is determined that there is an abnormal current transformer in the line at the time of testing, the abnormal current transformer is located based on the contribution of each phase current transformer to the real-time statistics.

[0181] When the real-time statistic Q exceeds the statistic threshold Q c After that, it is necessary to determine which specific line CT metering error state has changed abnormally. That is, by calculating the role of each line in the construction of the over-limit statistics, the importance of the variable to the over-limit statistics, that is, the contribution, can be obtained. It is considered that the line with the highest contribution has changed abnormally.

[0182] When the Q statistic exceeds the statistical control limit, the contribution of the current transformer measurement data on the i-th line to the Q statistic is Con i :

[0183]

[0184] represents the maximum value of the product of the weight and the current measurement value in the n-1th line, Indicates the minimum value of the product of the weight and the current measurement value in the n-1th line.

[0185] The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A current transformer error evaluation method based on current imbalance contribution analysis, characterized in that: The specific steps include: Collect CT group data with normal error status and construct CT group modeling current vector data; Using the current vector data of CT population modeling, a current transformer error state assessment model based on overdetermined linear equations is constructed; The regularized iterative algorithm is used to solve the benchmark weight coefficient matrix X in the current transformer error state assessment model. standard ; Using the benchmark weight coefficient matrix X standard , construct the error fluctuation characteristic E f ; Calculate the error fluctuation characteristic E f The Q statistic is used to solve the error fluctuation characteristic quantity E based on the empirical estimate of confidence level. f The statistical control limit Q cl ; Get the real-time measurement data of CT, calculate the Q statistic of the real-time data, and compare it with the statistical control limit Q cl Compare and judge whether there is abnormal operation error of transformer in the line; If it is determined that there is an abnormal current transformer in the line at the time of testing, the abnormal current transformer is located based on the contribution of each phase current transformer to the real-time statistics.

2. A current transformer error evaluation method based on current imbalance contribution analysis according to claim 1, characterized in that: The acquisition of CT group data with qualified error states and the construction of CT group modeling current vector data are specifically as follows: There will be a difference between the primary current value and the secondary current value of the current transformer CT, which is divided into ratio error f i With phase difference δ i , as shown in formula (1) and (2): δ=φ2-φ1 (2) In the formula, K i is the rated transformation ratio of CT, I1 and φ1 are the amplitude and phase of the actual primary current of CT, I2 and φ2 are the amplitude and phase of the actual secondary current, according to the definition of the ratio difference and phase difference of current transformer, the relationship between the primary current measurement value and the secondary current measurement value is derived as follows: in, represents the measured value of the primary current of the CT on the i-th line, represents the amplitude of the measured current on the primary side of the CT on the i-th line, φ 1i represents the phase of the measured current value of the primary side of the CT on the i-th line, represents the amplitude of the CT secondary current measurement value on the i-th line, φ 2i represents the phase of the CT secondary current measurement value on the i-th line, δ i represents the angle difference of CT on the i-th line, f i represents the ratio difference of CT on the i-th line, r represents the phase sequence, and i represents the i-th line.

3. A current transformer error evaluation method based on current imbalance contribution analysis according to claim 2, characterized in that: The current vector data of the CT group modeling is used to construct a current transformer error state evaluation model based on an overdetermined linear equation group model: According to Kirchhoff's current law, the vector sum of the primary current measurement values ​​of each line at a node is 0, so we can get: By approximating formula (5), we can get: Perform offline calibration on any one of the n lines and measure the known ratio error f r , angle difference δ r , and then perform transposition processing on formula (7): Formula (8) is equivalent to: AX=B (9) This equation is an overdetermined linear equation system, where A represents the amplitude matrix of the CT secondary side current measurement value on the line, X represents the weight coefficient matrix of the CT secondary side current measurement value amplitude, and B represents the constant term of the known ratio difference angle difference, specifically:

4. A current transformer error evaluation method based on current imbalance contribution analysis according to claim 3, characterized in that: The regularized iterative algorithm is used to solve the benchmark weight coefficient matrix X in the current transformer error state assessment model. standard Specifically: The solution process of the regularized iterative algorithm is as follows: Multiply (9) by A on the left T , we can get: A T AX=A T B (13) In the coefficient matrix A T Adding a regularization parameter α to the elements on the main diagonal of A yields: (A T A+αI)X=A T B+αX (14) Where I is A T The identity matrix of the same order as A, Continuing to transform formula (14), we can get: X (k+1) =(A T A+αI) -1 (A T B+αX k ) (15) Where k represents the number of iterations, Δ=X (k+1) -X k (16) When Δ is a minimum constant, solve the benchmark weight coefficient matrix X standard .

5. A current transformer error evaluation method based on current imbalance contribution analysis according to claim 4, characterized in that: The reference weight coefficient matrix is ​​used to construct the error fluctuation feature value E f Specifically, Error characteristic fluctuation E f : Indicates the measured current, S pre Represents the estimated and current.

6. A current transformer error evaluation method based on current imbalance contribution analysis according to claim 5, characterized in that: The calculation error fluctuation characteristic quantity E f The Q statistic is used to solve the error fluctuation characteristic quantity E based on the empirical estimate of confidence level. f The statistical control limit Q cl Specifically, S51: Error fluctuation characteristic quantity E f Perform standardized data processing to obtain a standardized data matrix As well as the mean vector b and the variance matrix Σ; Among them, 1 n =[1,1,…,1] T ∈Φ n×1 , b=(E f ) T 1 n / n is the mean vector of the error fluctuation characteristic quantity, is the variance matrix of the error fluctuation characteristic quantity; S52: Standardized matrix Perform singular value decomposition and calculate its eigenvalues ​​λ1,λ2,…,λ m and the corresponding eigenvector [PP e ]; In the formula is the score matrix, is the load matrix; For any i' and j', when i'≠j', we have Established, that is, each score vector satisfies mutual orthogonality, and similarly, each load vector also satisfies mutual orthogonality, that is: According to the characteristics of the sampled voltage data decomposition, both sides of equation (22) are multiplied by p i‘ We can get: From a geometric point of view, we can know from formula (23) that the score vector t i‘ is the data matrix In the corresponding load vector p i‘ The length of the mapping on represents the data matrix The proportion of the corresponding load vector is arranged in descending order according to its length, as shown in the following formula: ||t1||>||t2||>…>||t m || (24) Then the load vector p1 is the data matrix The direction with the largest change, p2 is the data matrix The second largest direction of change, p m The data matrix The direction with the smallest change, If there is some linear correlation between the variables in the data matrix, then The change will be mainly due to the largest loads The projection in other directions will be small. The following principal component decomposition can be performed: In the formula, is the data matrix The principal subspace model of is the data matrix The residual subspace model of , T is the principal component score matrix, P is the principal component loading matrix, T e is the residual score matrix, P e is the residual loading matrix, Load matrices P and P e The data matrix can be The covariance matrix R of is obtained by singular value decomposition, as shown below: In the formula, Λ=diag(λ1,λ2,…,λ m ), λ1≧λ2≧…≧λ m is the eigenvalue of the covariance matrix R, is the load vector composed of the corresponding eigenvectors. The larger the eigenvalue, the stronger the correlation of the represented variables. S53: Select the number of principal components p according to the cumulative contribution percentage of variance CPV method, and determine the load matrix P of the principal element subspace and the load matrix P of the residual subspace e ; The CPV method determines the number of principal components by calculating the cumulative variance percentage of the first p principal components. The calculation formula is: S54: Select the test confidence β and calculate the statistical threshold Q of the Q statistic cl Q cl is the statistical control threshold with a significance level of β, calculated as follows: In the formula, C β is the critical value of the normal distribution at the detection level β.

7. A current transformer error evaluation method based on current imbalance contribution analysis according to claim 6, characterized in that: The method of obtaining the real-time measurement data of the CT, calculating the Q statistic of the real-time data, and comparing it with the statistical control limit to determine whether there is an abnormal operation error of the transformer in the line is specifically as follows: CT real-time measurement data, first calculate the real-time error fluctuation characteristic quantity E f , and then calculate the Q statistic of the process information according to formula (29): if it is less than the statistical threshold Q c , indicating that the current transformers on each line are in normal operation; if it is greater than the statistical threshold Q c , indicating that the current transformer on the line is in abnormal operation state. Among them, E f Represents the real-time error fluctuation characteristic, P e Represents the loading matrix of the residual subspace.

8. A current transformer error evaluation method based on current imbalance contribution analysis according to claim 7, characterized in that: If it is determined that there is an abnormal current transformer in the circuit at the time of testing, based on the contribution of each phase current transformer to the real-time statistics, the abnormal current transformer is located as follows: When the real-time statistic Q exceeds the statistic threshold Q c Finally, it is necessary to determine which specific line CT metering error state has changed abnormally, that is, by calculating the role of each line in the construction of the over-limit statistics, the importance of the variable to the over-limit statistics, that is, the contribution, is obtained. It is considered that the line with the highest contribution has changed abnormally. When the Q statistic exceeds the statistical control limit, the contribution of the current transformer measurement data on the i-th line to the Q statistic is Con i : represents the maximum value of the product of the weight and the current measurement value in the n-1th line, Indicates the minimum value of the product of the weight and the current measurement value in the n-1th line.

Images