A method and device for evaluating the error of a high-voltage current transformer for power distribution used for metering

By constructing a vector regression model and online error evaluation method, the problem of power outage in the transformer metering performance inspection in the prior art is solved, and the online monitoring and metering performance management of high-voltage current transformers in the distribution network is realized.

CN118797221BActive Publication Date: 2024-11-29WUHAN HANYANG POWER SUPPLY POWER ENG INSTALLATION TEAM
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411285444.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-13
Publication Date
2024-11-29
Estimated Expiration
2044-09-13

AI Technical Summary

Technical Problem

In the prior art, the transformer metering performance inspection requires power outage and the maintenance interval is long, which makes it difficult to detect abnormal transformers in a timely manner, affecting the power grid trade settlement.

Method used

The vector regression model is constructed based on the least squares method of the complex domain and the Kirchoff current law, collect current data and perform error evaluation, determine whether the transformer is normal, and online monitoring is realized.

Benefits of technology

The online metering performance monitoring of the distribution network high-voltage current transformers is realized, without power outage inspection, and abnormal transformers are discovered in a timely manner to ensure the accuracy of power grid trade settlement.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118797221B_ABST
    Figure CN118797221B_ABST
Patent Text Reader

Abstract

The present invention relates to the technical field of online monitoring of electric power metering, and specifically to a method and device for evaluating the error of a high-voltage current transformer for power distribution for metering, the method comprising the following steps: constructing a vector regression model based on the least squares method in the complex domain and Kirchhoff's current law; collecting current data within a set time period, and bringing it into the vector regression model to obtain the mean of the regression coefficient and the model prediction value, and judging whether the mean of the regression coefficient is valid, if not, repeating the step until the mean of the regression coefficient is valid; judging whether the transformer is normal according to the mean of the regression coefficient. It can solve the problem in the prior art that the metering performance inspection of the transformer requires power outage and the maintenance interval is long, which makes it difficult to find abnormal transformers in time.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of online monitoring of electric power metering, and in particular to an error evaluation method and device for a power distribution high-voltage current transformer for metering. 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 fixed periods, 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.

[0003] As an important reference for power grid trade settlement, metering transformers have very high requirements for accuracy. Considering that the metering performance of transformers may change with the increase of service life, it may not meet the accuracy requirements.

[0004] However, during the operation of the power grid, the maintenance intervals are long and the cost of power outage maintenance is high, which makes it difficult to detect abnormal transformers in time, thus affecting the trade settlement of the power grid. Summary of the invention

[0005] In view of the defects existing in the prior art, the purpose of the present invention is to provide a method and device for evaluating the error of a high-voltage current transformer for power distribution for metering, which can solve the problem in the prior art that the metering performance inspection of the transformer requires power outage and the maintenance interval is long, resulting in difficulty in timely detection of abnormal transformers.

[0006] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0007] In one aspect, the present invention provides a method for evaluating an error of a high-voltage current transformer for power distribution for metering, comprising the following steps:

[0008] A vector regression model is constructed based on the least squares method in the complex domain and Kirchhoff's current law;

[0009] Collect current data within a set time period and bring it into the vector regression model to obtain the mean of the regression coefficient and the model prediction value, and determine whether the mean of the regression coefficient is valid. If not, repeat this step until the mean of the regression coefficient is valid.

[0010] According to the mean value of the regression coefficient, determine whether the transformer is normal.

[0011] In some optional solutions, the Kirchhoff's current law based on the complex domain least squares method is used to construct a regression model, including:

[0012] Based on Kirchhoff's current law, the KCL equations for the current of each branch of the line to be tested are constructed;

[0013] Based on the least squares method in the complex field, a vector regression model for the KCL equations is constructed.

[0014] In some optional schemes, the KCL equations are ,in, is the current value of the secondary side current of the i-th branch converted to the primary side, is the current value of the secondary side current of the jth branch converted to the primary side, is the system error of the current transformer in the i-th branch, is the system error of the j-th branch current transformer, is the system error of the current transformer of the kth branch, n is the total number of branches, and i, j, k are the branch numbers.

[0015] In some optional schemes, the regression model is ,in, is the weight term of the complex regression coefficient corresponding to the independent variable based on the secondary-side-to-primary-side current acquisition value of the j-th branch in the line equation to be tested, which takes the current of the i-th branch as the dependent variable. is the complex regression coefficient bias term in the line equation to be tested, which takes the current of the i-th branch as the dependent variable. is the secondary current of the i-th branch, n is the total number of branches, and i, j, k are the branch numbers.

[0016] In some optional solutions, the current data collected within a set time period is brought into a vector regression model to obtain a regression coefficient mean and a model prediction value, including:

[0017] Collecting current data of all branches of the circuit to be tested within a first set time period;

[0018] The current data is evenly divided into a first set number of groups, and the current data in each group is brought into the regression model;

[0019] The regression equation corresponding to the i-th branch current data is extracted and solved to obtain the mean of the regression coefficient and the model prediction value.

[0020] In some optional solutions, the determination of whether the mean of the regression coefficient is valid includes:

[0021] Determine the fitting ability parameter of the regression model based on the mean of the regression coefficient and the model prediction value;

[0022] If the fitting ability parameter is greater than or equal to the set value, the regression coefficient mean is considered valid. If the fitting ability parameter is less than the set value, the regression coefficient mean is considered invalid.

[0023] In some optional solutions, judging whether the transformer is normal according to the mean value of the regression coefficient includes:

[0024] According to the mean of regression coefficient, the variation of ratio difference and angle difference is determined;

[0025] According to the ratio difference limit value of the transformer specification, determine the ratio difference variation range, and judge whether the transformer is normal according to whether the ratio difference variation is within the ratio difference variation range;

[0026] According to the angle difference limit value of the transformer specification, the range of the angle difference variation is determined, and whether the transformer is normal is determined based on whether the angle difference variation is within the range.

[0027] In some optional solutions, if the transformer is judged to be normal based on the mean value of the regression coefficient, secondary data collection is also performed and secondary judgment is performed, including:

[0028] Collecting current data of all branches of the circuit to be tested within a second set time period;

[0029] The current data is evenly divided into a second set number of groups, and the current data in each group is brought into the regression model and solved to obtain the regression coefficient of each group;

[0030] Cluster the difference between the regression coefficient of each group and the mean of the regression coefficient;

[0031] Determine whether the transformer is abnormal based on the clustering results.

[0032] In some optional solutions, judging whether the mutual inductor is abnormal according to the clustering result includes:

[0033] According to the number of ratio difference clusters and the number of angle difference clusters in the difference clustering results between the regression coefficients of each group and the mean of the regression coefficients, a group of data with the largest degree of change or sudden change in the state of the mutual inductor is determined;

[0034] According to the group regression coefficient and the mean value of the regression coefficient corresponding to the maximum change degree or sudden change of the state of the mutual inductor, the ratio difference and angle difference corresponding to each mutual inductor are determined;

[0035] According to the ratio difference and angle difference corresponding to each mutual inductor, it is judged whether the mutual inductor is abnormal.

[0036] On the other hand, the present invention also provides a device for evaluating an error of a power distribution high-voltage current transformer for metering, comprising the following steps:

[0037] A model building module, which is used to build a vector regression model based on the least square method in the complex domain and Kirchhoff's current law;

[0038] A fitting module is used to collect current data within a set time period and bring it into the vector regression model to obtain the mean of the regression coefficient and the model prediction value, and to determine whether the mean of the regression coefficient is valid. If not, the step is repeated until the mean of the regression coefficient is valid.

[0039] The judgment module is used to judge whether the mutual inductor is normal according to the mean value of the regression coefficient.

[0040] Compared with the prior art, the advantages of the present invention are: constructing a vector regression model by using the least squares method based on the complex domain and Kirchhoff's current law; collecting current data within a set time period and bringing it into the vector regression model to obtain the mean value of the regression coefficient and the model prediction value, and judging whether the transformer is normal based on the mean value of the regression coefficient. In this way, the transformer can be monitored online without the need for power outage inspection, realizing effective operation management and supervision of the online metering performance of the distribution network high-voltage current transformer. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0042] Figure 1 The present invention is a flowchart of a method for evaluating an error of a power distribution high-voltage current transformer for metering in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0044] The embodiments of the present invention are further described in detail below with reference to the accompanying drawings.

[0045] like Figure 1 As shown, on the one hand, the present invention provides a method for evaluating an error of a high-voltage current transformer for power distribution for metering, comprising the following steps:

[0046] S1: Construct a vector regression model based on the least squares method in the complex domain and Kirchhoff's current law.

[0047] In some optional embodiments, step S1 includes:

[0048] S11: Based on Kirchhoff's current law, construct the KCL equation group for the current of each branch of the line to be tested.

[0049] Based on Kirchhoff's current law, the sum of the input current and output current of any line is 0, that is:

[0050] (1)

[0051] in Indicates The primary current of the branch, Indicates the number of branches (including feeders) in the line. Formula (1) is expanded according to the corresponding relationship between current amplitude and phase as follows:

[0052] (2)

[0053] in Indicates The primary current amplitude of the branch, Indicates The primary current phase of the branch, w In the actual data collection process, the data source is taken from the secondary side of the current transformer. Considering that the measurement performance of the current transformer will change with the increase of service life, the differential ratio is usually used in the power grid system. and angle difference Indicates the system error of the transformer.

[0054] (3)

[0055] (4)

[0056] in Indicates the rated transformation ratio, Indicates the primary current amplitude, represents the secondary current amplitude, Indicates the primary current phase, Indicates the secondary current phase. , angular difference Substituting into formula (2), the KCL equation is transformed into:

[0057] (5)

[0058] Express equation (5) in plural form:

[0059] (6)

[0060] The secondary current of the i-th branch is , the system error of the current transformer of the i-th branch is ,in, is the rated transformation ratio of the ith branch, is the secondary current amplitude of the ith branch, is the secondary current phase of the ith branch, For time, is the ratio difference of the i-th branch, is the angle difference of the i-th branch, The j in is the imaginary unit.

[0061] Shift each line to the right to construct the KCL equation group, that is, the equation group of the line to be tested that meets the KCL characteristics considering the mutual inductor error:

[0062] (7)

[0063] in, is the current value of the secondary side current of the i-th branch converted to the primary side, is the current value of the secondary side current of the jth branch converted to the primary side, , , are the system errors of the current transformers of the i-th, j-th and k-th branches respectively, n is the total number of branches, i, j, k are the branch numbers.

[0064] S12: Based on the least squares method in the complex domain, a vector regression model for the KCL equations is constructed. Specifically:

[0065] For complex regression models , to ensure the loss function Minimum, using the conjugate property of complex numbers, we get:

[0066] (8)

[0067] in, is the current acquisition value converted from the secondary side to the primary side of the corresponding branch i in the line equation group to be tested, which takes the current of the i-th branch as the dependent variable. is the current acquisition value of the secondary side of other branches converted to the primary side of the line equation to be tested, which takes the current of the i-th branch as the dependent variable. The current prediction value of the secondary side of the tested branch i converted to the primary side of the tested line equation corresponding to the tested branch i in the tested line equation group with the current of the i-th branch as the dependent variable, W is the complex regression coefficient weight, is the loss function, T is the transpose, For conjugation.

[0068] The real part of the complex regression coefficient W and the imaginary part To guide, to make

[0069] (9)

[0070] The regression coefficient of the complex regression model is

[0071] (10)

[0072] Among them, B is the bias term of the complex regression coefficient.

[0073] Construct the regression model of the equation group in formula (7).

[0074] (11)

[0075] in, is the weight term of the complex regression coefficient corresponding to the independent variable based on the secondary-side-to-primary-side current acquisition value of the j-th branch in the line equation to be tested, which takes the current of the i-th branch as the dependent variable. is the complex regression coefficient bias term in the line equation to be tested, which takes the current of the i-th branch as the dependent variable. is the secondary current of the i-th branch, n is the total number of branches, and i, j, k are the branch numbers.

[0076] S2: Collect current data within a set time period and bring it into the vector regression model to obtain the mean of the regression coefficient and the model prediction value, and determine whether the mean of the regression coefficient is valid. If not, repeat this step until the mean of the regression coefficient is valid.

[0077] Since a reference system of normal state is required to analyze the operating transformer without stopping the operation, it is necessary to judge the initial state of the transformer.

[0078] In some optional embodiments, current data within a set time period is collected and brought into a vector regression model to obtain a regression coefficient mean and a model prediction value, including:

[0079] S21A: Collect current data of all branches of the circuit to be tested within a first set time period.

[0080] In this example, the first set time period is one week. Through the high-precision current acquisition device, the current data of all branches of the line to be tested are collected once a minute to reflect the current changes of each branch line user in normal use every week under normal conditions. T represents the number of collections within the collection time period.

[0081] S21B: Divide the current data into a first set number of groups, and bring the current data in each group into the regression model.

[0082] If it is believed that the current transformer has an error, its error includes systematic error and random error. The systematic error is a fixed value, which is used as the basis for judging the state of the current transformer. The random error result is an irregular random small value, which is usually ignored. In order to eliminate the influence of random errors, random sampling without replacement is used to collect current data. Randomly divide them into T / m groups, with m samples in each group. Substitute the current data of each group into the corresponding equations.

[0083] S21C: Extract the regression equation corresponding to the i-th branch current data and solve it to obtain the mean of the regression coefficient and the model prediction value.

[0084] In this example, the regression equation corresponding to the i-th branch current data is extracted as follows: . Its data set can be expressed as:

[0085] (12)

[0086] in, is the current data of the i-th branch of the h-th group at the t-th moment, is the full-time current data of the i-th branch of the h-th group, is the full-time current data of the branches of the hth group except the i-th branch, is the vector regression coefficient matrix of the regression model under the h-th group of current collection data of the i-th branch, = is the weight term matrix of the vector regression coefficient of the regression model under the h-th group of current acquisition data set of the i-th branch, is the weight term of the complex regression coefficient corresponding to the independent variable based on the current collection value of the first branch in the hth group of regression equations with the current of the i-th branch as the dependent variable, is the bias term of the complex regression coefficient in the hth group of regression equations with the current of the i-th branch as the dependent variable, T is the transpose, e represents the unit complex number, and m is the number of data in each group of samples.

[0087] The mean value of the regression coefficient is obtained according to formula (12), which is used to characterize the relative change of the system error of the corresponding transformer:

[0088] (13)

[0089] in, is the mean regression coefficient of the i-th branch, is the regression coefficient of the hth group of the ith branch.

[0090] , It represents the mean value of the weighted terms of the complex regression coefficients corresponding to the independent variables based on the current collection value of the i+1th branch in the regression equation with the current of the i-th branch as the dependent variable.

[0091] The model prediction value can be obtained based on the regression coefficient of each branch. The model prediction value is the result of multiplying X and B in formula (12).

[0092] In some optional embodiments, determining whether the mean of the regression coefficient is valid includes:

[0093] S22A: Determine the fitting ability parameter of the regression model according to the regression coefficient mean and the model prediction value. Specifically, according to the fitting ability parameter Evaluate the goodness of fit of the regression model.

[0094] (14)

[0095] Among them, SSR represents the sum of squares of the differences between the model prediction value and the actual observation value. SST represents the sum of squares of the differences between the actual observation value and the mean of the dependent variable. The mean of the dependent variable is determined based on the collected current acquisition data, and the model prediction value is determined based on the regression coefficient combined with the regression model.

[0096] S22B: If the fitting ability parameter is greater than or equal to the set value, the regression coefficient mean is considered valid; if the fitting ability parameter is less than the set value, the regression coefficient mean is considered invalid.

[0097] In this case, when the fitting capability parameter When , the corresponding equation group of the regression model fitting is better and can more accurately reflect the actual effect. You can proceed with subsequent operations.

[0098] Since the state of the transformers on each branch is considered to be stable within a period of time, when When the corresponding equation group of the regression model fitting does not meet the conditions, it can be considered that the data fluctuates or the state of the mutual inductor changes, and it is necessary to collect data again in step S21A.

[0099] S3: Determine whether the transformer is normal based on the mean value of the regression coefficient.

[0100] In some optional embodiments, step S3 includes:

[0101] S31: Determine the ratio difference change and the angle difference change according to the regression coefficient mean.

[0102] because The threshold requirement is high. It can be considered that the regression model (11) can correctly reflect the equation group (7) of the line to be tested that meets the KCL characteristics considering the mutual inductor error. Then, for the i-th equation in the equation group, the regression coefficient can be approximately considered to be The relative relationship with the error parameter in the equation to be tested is equal. Characterized by the relative error relationship parameters between mutual inductors.

[0103] (15)

[0104] Therefore, the difference between the i-th branch and the j-th branch can be approximately expressed according to the mean of the regression coefficient: And the angular difference change ,in, represents the systematic error of the CT of the i-th branch, represents the systematic error of the CT of the jth branch, n represents the total number of n branches on a line, i, j represents the branch number, represents the ratio difference of the i-th branch, is a natural constant, represents the angle difference of the i-th branch, The approximate characteristic quantity representing the relative change of the system error between the mutual inductors of the i-th branch and the j-th branch, Where j is an imaginary unit. CT is a current transformer.

[0105] The regression model (11) is an approximate fit of (7). The data correlation of branch j to branch i in the line can be expressed as the relative change between the system errors of branch j and branch i. In the regression model, Approximate expression.

[0106] S32: Determine a ratio difference variation range according to a ratio difference limit value of the transformer specification, and judge whether the transformer is normal according to whether the ratio difference variation is within the ratio difference variation range.

[0107] Considering the 0.2-level metering requirements based on the specification, the absolute value of the ratio difference of each transformer is The transformer is considered to be in normal state. The ratio difference and angle difference changes in equation (15) are considered separately.

[0108] For the ratio difference change , which can be expressed as follows:

[0109] (16)

[0110] According to the ratio difference limit of the transformer specification, the ratio difference variation range can be calculated by combining formula (16). Specifically, the conditions for judging whether the current transformers of branch i and branch j meet the 0.2 level measurement specification of the transformer are as follows: and hour, ;when and hour, . According to this, the current transformer ratio difference change of branch i and branch j is The upper bound of , the lower bound is That is to say, When the corresponding two transformers One of them is in an abnormal state, that is, the mutual inductor on the i-th branch or the mutual inductor on the j-th line is abnormal.

[0111] Based on current data collection Corresponding regression model, determine the regression coefficient in the i-th equation in the regression model. The weight term of the vector regression coefficient in the regression model is split into polar diameter and polar angle according to polar coordinates, and the split polar diameter is used to approximately represent the ratio difference change between the two current transformers. For each coefficient, more than half (n / 2) of the regression coefficient weight terms do not belong to ,express Abnormal. If less than half (n / 2) of the regression coefficient weight terms have a polar radius that does not belong to ,express , All abnormal.

[0112] S33: Determine the range of the angle difference variation according to the angle difference limit value of the transformer specification, and judge whether the transformer is normal according to whether the angle difference variation is within the range of the angle difference variation.

[0113] In this example, considering the 0.2 level metering requirement based on the specification, the absolute value of the angular difference of each transformer is The transformer is considered to be in normal condition.

[0114] For the angular difference , which can be expressed as the following formula

[0115] (17)

[0116] The conditions for judging whether the current transformers of branch i and branch j meet the 0.2 level measurement specification of the transformer are as follows: and hour, The change in the current transformer angle difference between branch i and branch j is The upper bound of , whose lower bound is If , the corresponding two transformers One of them is in an abnormal state.

[0117] Based on current data collection Corresponding regression model, determine the regression coefficient in the i-th equation in the regression model. The weight term of the vector regression coefficient in the regression model is split into polar diameter and polar angle according to polar coordinates, and the split polar angle is used to approximately represent the change in the angular difference between the two current transformers. For each coefficient, more than half (n / 2) of the regression coefficient weight terms do not belong to the polar angle ,express Abnormal. If less than half (n / 2) of the regression coefficient weight terms have polar angles that do not belong to ,express , All abnormal.

[0118] In some optional embodiments, if the mutual inductor is judged to be normal according to the mean value of the regression coefficient, S4: further performing secondary data collection and secondary judgment, including:

[0119] S41: Collecting current data of all branches of the circuit to be tested within a second set time period.

[0120] In this example, the second set time period is one week.

[0121] S42: Divide the current data into a second set number of groups, and bring the current data in each group into the regression model and solve it to obtain the regression coefficient of each group.

[0122] In this example, the current data is divided into the second set number of groups according to the time sequence. The second set number of groups is 7 groups, that is, one group per day, and the current data of the new week is extracted to construct seven current data sets by day. , , , k is the group number, represents the current data set of the kth group of the i-th branch, is the full-time current data of the i-th branch of the k-th group, is the full-time current data of the kth group of branches except the i-th branch. Input the new collected data to obtain the new group regression coefficient in the regression equation (Equation 12): , The approximate characteristic quantity representing the relative change of the system error between the mutual inductors of the i-th branch of the k-th group and the j+1-th branch, is the bias term of the complex regression coefficient in the kth group of regression equations with the current of the i-th branch as the dependent variable.

[0123] S43: clustering the difference between the regression coefficient of each group and the mean of the regression coefficient;

[0124] After the screening in step S3, the state of the transformer is considered to be normal, considering that the accuracy and stability of the metering transformer are highly required. Therefore, it is considered that the probability of the transformer changing within a period of time is low. If a change occurs, it is considered that only one group of transformers changes at a time within the same period of time. Based on this assumption, DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is used to cluster the regression coefficient difference between the initial state (first data collection) and the state to be measured (second data collection) of the transformer.

[0125] The difference between the new daily parameters and the training parameters is used to construct a parameter difference set:

[0126] (18)

[0127] Expanding formula (18), we get:

[0128] (19)

[0129] Through DBSCAN , Clustering is performed separately, where is the ratio difference change between the i-th branch and the j-th branch corresponding to the k-th group of data, is the change in the angle difference between the i-th branch and the j-th branch corresponding to the k-th group of data, n is the number of branches, is the change in the ratio difference between the i-th branch and the j-th branch corresponding to the k-th group of data and the change in the ratio difference between the i-th branch and the j-th branch corresponding to the initial state, is the change in the angular difference between the i-th branch and the j-th branch corresponding to the k-th group of data and the change in the angular difference between the i-th branch and the j-th branch corresponding to the initial state. Set the parameters minPts=1, r=1e-4, Set the parameters minPts=1, r=10'. r represents the neighborhood distance threshold of a sample, and minPts represents the threshold of the number of samples in the neighborhood of a sample with a distance of r.

[0130] Record the number of clusters of DBSCAN and the number of clusters of the difference difference of each group , the number of clusters of angle difference ,in Represents the clustering result of the i-th group.

[0131] S43: Determine whether the mutual inductor is abnormal according to the clustering result.

[0132] In some optional embodiments, step S43 includes:

[0133] S431: Determine a group of data with the largest degree of state change or sudden change of the mutual inductor according to the number of ratio difference clusters and the number of angle difference clusters in the difference clustering results between the group regression coefficients and the mean values ​​of the regression coefficients.

[0134] ① Considering the impression of abnormal fluctuations that may occur in the collection, record the number of clusters of the difference difference values ​​of each group , the number of clusters of angle difference As a result, if More than half of the total number is the same. In this example, if If the number of clusters is the same for more than 4 days, the transformer state is considered stable within the week, and no mutation occurs, which is considered to be a stable state. The last group s with a stable number of clusters is selected as the basis for the evaluation and analysis of the transformer state.

[0135] ② If the number of clusters of the difference values ​​of each group is recorded , the number of clusters of angle difference If there is a jump increase in the result, and the number of clusters in the subsequent days is the same as the jump result, it means that there is a sudden change in the state of the transformer, then the last group of data is used.

[0136] ③ If the number of cluster cores fluctuates over the seven days, it means that the state of the transformer fluctuates and the current measurement data is unstable. It is necessary to re-collect data to determine the state.

[0137] S432: Determine the ratio difference and angle difference corresponding to each mutual inductor according to the group regression coefficient and the mean value of the regression coefficient corresponding to a group of data with the largest degree of mutual inductor state change or sudden change.

[0138] (1) Transformer status assessment under stable conditions.

[0139] The last group s with stable cluster numbers is selected as the evaluation and analysis basis of the transformer status. The status prediction is performed by judging the change of the regression coefficient of the regression equation group corresponding to the newly collected sth group of data and the mean value of the regression coefficient obtained in step S2.

[0140]

[0141] (20)

[0143] (twenty one)

[0144] but express The largest corresponding transformer The state change is the greatest. is the branch number, and n is the total number of branches. It is assumed that the operating state of the transformer is normal and will not change greatly. It is assumed that the error change is caused by the largest change. Produced, therefore, according to express The largest corresponding transformer The state change degree is the largest, and the branch transformer i that is most likely to have a problem is found. Based on the group regression coefficient corresponding to the s-th group of collected data of the i-th branch transformer, according to the following formula (22):

[0145] (twenty two)

[0146] Calculate the ratio difference and angle difference results corresponding to each CT (transformer) to obtain the ratio difference and angle difference results of each CT.

[0147] (2) Transformer status evaluation under sudden change

[0148] The last set of data is used. In this example, the seventh set of data is used for analysis to evaluate the state of the transformer. The state prediction is performed by judging the change of the regression coefficient in the regression equation group of the seventh day newly collected and the regression coefficient in the training regression equation group.

[0149] Substitute the data in the last group into the data in equations (20) to (22), and then find the ratio difference and angle difference results of each CT.

[0150] (3) Transformer status evaluation cannot be performed

[0151] It is necessary to collect data again and judge the status again.

[0152] S433: Determine whether the mutual inductor is abnormal based on the ratio difference and angle difference corresponding to each mutual inductor.

[0153] In this example, whether the transformer is normal is determined by comparing the ratio difference and angle difference of each CT with the standard range corresponding to the transformer.

[0154] In a second aspect, the present invention provides an error evaluation device for a power distribution high-voltage current transformer for metering, comprising the following steps: a model building module, a fitting module and a judgment module.

[0155] Among them, the model building module is used to construct a vector regression model based on the least squares method in the complex domain and Kirchhoff's current law; the fitting module is used to collect current data within a set time period and bring it into the vector regression model to obtain the mean of the regression coefficient and the model prediction value, and to determine whether the mean of the regression coefficient is valid. If not, repeat this step until the mean of the regression coefficient is valid; the judgment module is used to determine whether the transformer is normal based on the mean of the regression coefficient.

[0156] Among them, the functional implementation of each module in the above-mentioned distribution high-voltage current transformer error assessment device for metering corresponds to the various steps in the above-mentioned distribution high-voltage current transformer error assessment method embodiment for metering, and its functions and implementation processes will not be repeated here one by one.

[0157] In summary, the error evaluation method for the distribution high-voltage current transformer for metering constructs a vector regression model through the least squares method based on the complex domain and Kirchhoff's current law; collects current data within a set time period and brings it into the vector regression model to obtain the mean value of the regression coefficient and the model prediction value, and judges whether the transformer is normal based on the mean value of the regression coefficient. In this way, the transformer can be monitored online without power outage inspection, realizing effective operation management and supervision of the online metering performance of the distribution network high-voltage current transformer.

[0158] In addition, secondary data collection and secondary judgment are also carried out. 3. DBSCAN is used to analyze the state changes of the transformer and conduct quantitative evaluation of the transformer.

[0159] In the description of the present application, it should be noted that the terms "upper", "lower", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present application. Unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be a connection between the two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances.

[0160] It should be noted that, in this application, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0161] The above description is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest range consistent with the principles and novel features applied for herein.

Claims

1. A method for evaluating the error of a high-voltage current transformer for power distribution for metering, characterized in that: The following steps are involved: A vector regression model is constructed based on the least squares method in the complex domain and Kirchhoff's current law, including: Based on Kirchhoff's current law, the KCL equations for the current of each branch of the line to be tested are constructed; Based on the least square method in the complex domain, a vector regression model for the KCL equations is constructed; The KCL equations are ,in, is the current value of the secondary side current of the i-th branch converted to the primary side, is the current value of the secondary side current of the jth branch converted to the primary side, is the system error of the current transformer in the i-th branch, is the system error of the j-th branch current transformer, is the system error of the current transformer of the kth branch, n is the total number of branches, i, j, k are the branch numbers, The regression model is ,in, is the weight term of the complex regression coefficient corresponding to the independent variable based on the secondary-side-to-primary-side current acquisition value of the j-th branch in the line equation to be tested, which takes the current of the i-th branch as the dependent variable. is the complex regression coefficient bias term in the line equation to be tested, which takes the current of the i-th branch as the dependent variable. is the secondary current of the i-th branch, n is the total number of branches, i, j, k are the branch numbers; Collect current data within a set time period and bring it into the vector regression model to obtain the mean of the regression coefficient and the model prediction value, and determine whether the mean of the regression coefficient is valid. If not, repeat this step until the mean of the regression coefficient is valid. According to the mean value of the regression coefficient, determine whether the transformer is normal.

2. The error evaluation method for a power distribution high voltage current transformer for metering as claimed in claim 1, characterized in that: The current data collected within the set time period is brought into the vector regression model to obtain the regression coefficient mean and the model prediction value, including: Collecting current data of all branches of the circuit to be tested within a first set time period; The current data is evenly divided into a first set number of groups, and the current data in each group is brought into the regression model; The regression equation corresponding to the i-th branch current data is extracted and solved to obtain the mean of the regression coefficient and the model prediction value.

3. The error evaluation method for a power distribution high voltage current transformer for metering as claimed in claim 2, characterized in that: The determination of whether the mean of the regression coefficient is effective includes: Determine the fitting ability parameter of the regression model based on the mean of the regression coefficient and the model prediction value; If the fitting ability parameter is greater than or equal to the set value, the regression coefficient mean is considered valid. If the fitting ability parameter is less than the set value, the regression coefficient mean is considered invalid.

4. The error evaluation method for a power distribution high voltage current transformer for metering as claimed in claim 2, characterized in that: The method of judging whether the transformer is normal according to the mean value of the regression coefficient includes: According to the mean of regression coefficient, the variation of ratio difference and angle difference is determined; According to the ratio difference limit value of the transformer specification, determine the ratio difference variation range, and judge whether the transformer is normal according to whether the ratio difference variation is within the ratio difference variation range; According to the angle difference limit value of the transformer specification, the range of the angle difference variation is determined, and whether the transformer is normal is determined based on whether the angle difference variation is within the range.

5. The error evaluation method for a power distribution high voltage current transformer for metering as claimed in claim 1, characterized in that: If the transformer is judged to be normal based on the mean value of the regression coefficient, secondary data collection is also performed and secondary judgment is performed, including: Collecting current data of all branches of the circuit to be tested within a second set time period; The current data is evenly divided into a second set number of groups, and the current data in each group is brought into the regression model and solved to obtain the regression coefficient of each group; Cluster the difference between the regression coefficient of each group and the mean of the regression coefficient; Determine whether the transformer is abnormal based on the clustering results.

6. The error evaluation method for a power distribution high voltage current transformer for metering as claimed in claim 5, characterized in that: The step of judging whether the mutual inductor is abnormal according to the clustering result includes: According to the number of ratio difference clusters and the number of angle difference clusters in the difference clustering results between the regression coefficients of each group and the mean of the regression coefficients, a group of data with the largest degree of change or sudden change in the state of the mutual inductor is determined; According to the group regression coefficient and the mean value of the regression coefficient corresponding to the maximum change degree or sudden change of the state of the mutual inductor, the ratio difference and angle difference corresponding to each mutual inductor are determined; According to the ratio difference and angle difference corresponding to each mutual inductor, it is judged whether the mutual inductor is abnormal.

7. A device for evaluating the error of a high-voltage current transformer for power distribution for metering, characterized in that: The following steps are involved: Model building module, which is used to build a vector regression model based on the least squares method in the complex domain and Kirchhoff's current law, including: Based on Kirchhoff's current law, the KCL equations for the current of each branch of the line to be tested are constructed; Based on the least square method in the complex domain, a vector regression model for the KCL equations is constructed; The KCL equations are ,in, is the current value of the secondary side current of the i-th branch converted to the primary side, is the current value of the secondary side current of the jth branch converted to the primary side, is the system error of the current transformer in the i-th branch, is the system error of the j-th branch current transformer, is the system error of the current transformer of the kth branch, n is the total number of branches, i, j, k are the branch numbers, The regression model is ,in, is the weight term of the complex regression coefficient corresponding to the independent variable based on the secondary-side-to-primary-side current acquisition value of the j-th branch in the line equation to be tested, which takes the current of the i-th branch as the dependent variable. is the complex regression coefficient bias term in the line equation to be tested, which takes the current of the i-th branch as the dependent variable. is the secondary current of the i-th branch, n is the total number of branches, i, j, k are the branch numbers; A fitting module is used to collect current data within a set time period and bring it into the vector regression model to obtain the mean of the regression coefficient and the model prediction value, and to determine whether the mean of the regression coefficient is valid. If not, the step is repeated until the mean of the regression coefficient is valid. The judgment module is used to judge whether the mutual inductor is normal according to the mean value of the regression coefficient.

Citation Information

Patent Citations

  • Method and system for analyzing performance of voltage transformer, electronic equipment and medium

    CN114492199A

  • Improved KNN-based current transformer measurement state evaluation method and system for distribution network metering, and medium

    CN118534400A