Monitoring Method and System for Internal Defects of Distribution Transformers Based on Secondary Voltage Sag Recognition

Abstract

The present invention belongs to the technical field of internal defect detection of distribution transformers. To reduce the cost of online detection and achieve real-time monitoring, a method and system for monitoring internal defects of distribution transformers based on the identification of secondary voltage dips are provided, including the following steps: establishing a linear regression model of the secondary voltage of the distribution transformer to be measured, updating the linear regression prediction model of the secondary voltage of the distribution transformer to be measured daily, predicting the secondary voltage data of the distribution transformer to be measured on the same day, calculating the secondary voltage prediction difference according to the secondary voltage prediction value and the measured value, calculating the cumulative value of the secondary voltage drop of the distribution transformer to be measured compared with the predicted data by using the cumulative sum algorithm, and when the cumulative drop value is greater than the cumulative drop threshold, it is determined that an internal defect has occurred. The internal defect monitoring system of the distribution transformer includes a communication module, a collection module, a storage module, and a data processing module. The present invention automatically identifies internal defects according to the data automatically collected in the smart meter, improving economy, real-time performance, and power supply reliability.

Description

Technical Field

[0001] The present invention belongs to the technical field of internal defect detection of distribution transformers. Background Art

[0002] The distribution transformer, abbreviated as distribution transformer, is an important part of the distribution network, with a large number and a wide distribution range. Although the service life of the distribution transformer can generally reach more than 50 years, with the increase in the types of loads and the sharp growth of the total social load capacity, the distribution transformers in some power supply areas are in a heavy load state for a long time. Under high load conditions, the possibility of defects in the distribution transformer will increase significantly, and the failure rate will also increase accordingly. After the distribution transformer fails, it will be shut down for maintenance or replacement, resulting in long-term power outages, affecting the production and life of users, and reducing the power supply reliability of the distribution system. Therefore, it is necessary to perform non-stop maintenance operations on the distribution transformer in time when defects occur to prevent the many adverse effects caused by its failure.

[0003] When internal defects occur in the distribution transformer, its equivalent parameters and voltage and current transfer characteristics are less different from the normal state, making it difficult to detect. The existing internal defect detection methods for distribution transformers are mainly divided into offline detection methods and online detection methods. The offline detection method mainly analyzes defects by using various detection devices in the power-off state of the distribution transformer, that is, in the offline state. Since the distribution transformer needs to be shut down, it will affect the power supply to users and cause economic losses. The online detection method mainly measures the voltage and current on the primary side of the distribution transformer through an online detection device, and calculates the difference between the primary side voltage and the rated value to determine whether internal defects occur. Although the accuracy is relatively high, due to the high cost of the online detection device and the need to be equipped with operators, the cost is relatively high, and it is not economical for the transformers in the distribution system. Summary of the Invention

[0004] The purpose of the present invention is to solve the problems existing in the above-mentioned prior art, and provide a method for monitoring internal defects of a distribution transformer based on the identification of secondary voltage drop, which can reduce the online detection cost, improve the automation level, and realize real-time monitoring.

[0005] The present invention is realized through the following technical solutions: A method for monitoring internal defects of a distribution transformer based on the identification of secondary voltage drop, comprising the following steps:

[0006] Establish a linear regression model of the secondary voltage of the distribution transformer to be measured:

[0007] Based on the relationship between the secondary side voltage and current between the distribution transformer to be measured and the nearest distribution transformer, establish a linear regression model of the secondary voltage of the distribution transformer to be measured; the nearest distribution transformer refers to the distribution transformer with the shortest electrical distance from the distribution transformer to be measured;

[0008] Update the linear regression prediction model of the secondary voltage of the distribution transformer to be measured daily:

[0009] Collect the secondary voltage and secondary current data of each distribution transformer through the smart meters connected to the secondary side of each distribution transformer, and obtain a single-day sample data set by summarizing the secondary voltage and secondary current data daily; summarize the single-day sample data sets of the previous k days to obtain a k-day sample data set;

[0010] Summarize the k-day sample data set of the distribution transformer to be measured and the k-day sample data set of the nearest distribution transformer of the distribution transformer to be measured as the regression calculation data set; solve the linear regression constant and linear regression coefficient of the secondary voltage linear regression model of the distribution transformer to be measured on the current day according to the regression calculation data set, so as to obtain the secondary voltage linear regression prediction model of the distribution transformer to be measured on the current day;

[0011] Predict the secondary voltage data of the distribution transformer to be measured on the current day:

[0012] According to the secondary voltage linear regression prediction model of the previous day of the distribution transformer to be measured, and the secondary current sample data set in the single-day sample data set of the distribution transformer to be measured on the current day and the single-day sample data set of the nearest distribution transformer on the current day, calculate the secondary voltage prediction data set of the distribution transformer to be measured on the current day;

[0013] Identify the degree of secondary voltage drop:

[0014] According to the secondary voltage prediction data set of the distribution transformer to be measured on the current day and the secondary voltage sample data set in the single-day sample data set of the distribution transformer to be measured on the current day, calculate the secondary voltage prediction difference set;

[0015] According to the secondary voltage prediction difference set, use the cumulative sum algorithm to calculate the cumulative drop value of the secondary voltage of the distribution transformer to be measured compared with the predicted data;

[0016] Identify internal defects: When the cumulative drop value is greater than the cumulative drop threshold, it is determined that the distribution transformer to be measured has internal defects.

[0017] Further, establish the general formula of the secondary voltage linear regression model of the distribution transformer to be measured, including the following steps:

[0018] According to the primary side voltage-current relationship and the transformation relationship between the primary side voltage and the secondary side voltage between the distribution transformer to be measured and the nearest distribution transformer, establish the secondary side voltage-current relationship;

[0019] Ignore the transmission line impedance between the distribution transformer to be measured and the nearest distribution transformer to simplify the secondary side voltage-current relationship, and rewrite the simplified secondary side voltage-current relationship as the secondary side voltage expression of the distribution transformer to be measured;

[0020] Transform the secondary side voltage expression of the distribution transformer to be measured into a linear expression, so as to obtain the secondary voltage linear regression model.

[0021] Further, the expression of the secondary voltage linear regression model of the distribution transformer to be measured is:

[0022]

[0023] Among them, i represents the distribution transformer to be measured, j represents the nearest distribution transformer, and V i-2 represents the secondary voltage of the distribution transformer to be measured, and V j-2 represents the secondary voltage of the nearest distribution transformer, and I i represents the secondary current of the distribution transformer to be measured, and I j represents the secondary current of the nearest distribution transformer. α0 represents the linear regression constant, and α1, α2, and α3 all represent the linear regression coefficients, and ε represents the random error.

[0024] Furthermore, the linear regression constant and the linear regression coefficients are calculated by using the multiple linear regression calculation method, and the calculation equations are as follows:

[0025]

[0026] In the formula, respectively represent the average secondary voltage and the average secondary current of the distribution transformer to be measured in the regression calculation dataset; respectively represent the average secondary voltage and the average secondary current of the nearest distribution transformer in the regression calculation dataset; L pq all represent calculation notations, and p, q = 1, 2, 3;

[0027] The expression of

[0028]

[0029] L pq is as follows:

[0030]

[0031] In the formula, n represents the number of groups of sample data in the regression calculation dataset. Each group of sample data includes the secondary voltage of the distribution transformer to be measured, the secondary current of the distribution transformer to be measured, the secondary voltage of the nearest distribution transformer, and the secondary current of the nearest distribution transformer collected at the same moment; V i-2a and I ia respectively represent the secondary voltage data and the secondary current data of the distribution transformer to be measured in the a-th group of sample data; V j-2a and I ja respectively represent the secondary voltage data and the secondary current data of the nearest distribution transformer in the a-th group of sample data; a = 1, 2, 3..., n.

[0032] Furthermore, the expression of the linear regression prediction model of the secondary voltage of the distribution transformer to be measured on the same day is as follows:

[0033]

[0034] Among them, V j-2m and I jm respectively represent the secondary voltage data and secondary current data in the m-th group of sampling data in the single-day sample dataset of the nearest distribution transformer on the current day; I im represents the secondary current data in the m-th group of sampling data in the single-day sample dataset of the distribution transformer to be measured on the current day; represents the predicted value of the secondary voltage of the distribution transformer to be measured on the current day; m = 1, 2, 3, …, M, where M represents the number of groups of sampling data in the single-day sample dataset.

[0035] Further, the data calculation formulas it contains are as follows: The calculation formula for the predicted difference of the secondary voltage is as follows:

[0036]

[0037] Among them, V i-2m represents the secondary voltage data in the m-th group of sampling data in the single-day sample dataset of the distribution transformer to be measured on the current day, represents the predicted value of the secondary voltage of the distribution transformer to be measured on the current day, and ΔV i-2m represents the predicted difference of the secondary voltage of the distribution transformer to be measured on the current day.

[0038] Further, the calculation formula for the cumulative drop value is as follows:

[0039]

[0040] In the formula, S i (t) and S i (t - 1) respectively represent the cumulative drop values of the distribution transformer to be measured on the current day and the previous day.

[0041] Further, the cumulative drop threshold is determined as follows:

[0042] Step1: Select a test distribution transformer with serious internal defects and the same model as the distribution transformer to be measured, and conduct an open-circuit test on the test distribution transformer under rated conditions, that is, connect its primary side to an AC power supply with a voltage level equal to its rated primary voltage, and measure the voltage on its secondary side, which is the secondary side voltage V f-2 of the test distribution transformer in the state of serious internal defects;

[0043] Step2: Calculate the difference ΔV f-2 between the secondary side voltage V f-2N of the test distribution transformer in the state of serious internal defects and the rated secondary voltage V f-2 of the test distribution transformer. Its expression is: ΔV f-2 = V f-2N - V f-2;

[0044] Step 3: Calculate the cumulative drop value S within p days under the severe internal defect state fp , where 1 < p < 10, and its expression is: S fp = pMΔV f-2 , where M represents the number of groups of sample data in the single-day sample data set;

[0045] Step 4: Calculate the cumulative drop threshold S of the distribution transformer according to the severe defect degree of 0.1(10 - p) T , and its expression is: S T = 0.1(10 - p)S fp .

[0046] The present invention also provides a monitoring system for internal defects of a distribution transformer, which is characterized in that it is used to execute the method for monitoring internal defects of a distribution transformer based on the identification of secondary voltage drops as described in claims 1 to 9, and includes a communication module, a collection module, a storage module, and a data processing module;

[0047] The communication module is used for real-time communication with the distribution network information management system to obtain in real time the secondary voltage data and secondary current data recorded by the intelligent meters connected to the secondary sides of each distribution transformer in the distribution network and send them to the collection module, and forward the alarm information to the distribution network information management system;

[0048] The collection module is used for regularly summarizing the sampling data of each distribution transformer to form a single-day sample data set, a k-day sample data set, and a regression calculation data set for each distribution transformer and sending them to the storage module;

[0049] The storage module is used for storing the single-day sample data set, the k-day sample data set, and the alarm information of each distribution transformer;

[0050] The data processing module is used for daily updating the secondary voltage linear regression prediction model of the distribution transformer to be measured, predicting the secondary voltage data of the distribution transformer to be measured on the same day, identifying the degree of secondary voltage drop, identifying whether the distribution transformer to be measured has internal defects, generating alarm information when it is determined that the distribution transformer to be measured has internal defects, and sending the alarm information to the storage module and the communication module respectively.

[0051] Compared with the prior art, the beneficial effects of the present invention are:

[0052] 1. The present invention uses the data collected by the intelligent meters connected to the secondary sides of the distribution transformers in the distribution network to detect the internal defects of the distribution transformers, without the need to purchase additional on-site detection equipment suitable for transformers, reducing the cost of a large number of transformer defect detections and improving the operation economy of the power system.

[0053] 2. The internal defect detection of the present invention only requires the operating data of the voltage and current on the secondary side of the distribution transformer, and there is no need to shut down the transformer during the detection, avoiding the power outage of users caused by off-line detection, reducing the number of power outages and the power outage time caused by the operation and inspection of the transformer, and improving the power supply reliability of the power system.

[0054] 3. The internal defect monitoring system of the distribution transformer of the present invention automatically collects data from the intelligent electricity meter, automatically identifies the degree of secondary voltage drop according to the sampled data, and automatically identifies internal defects according to the cumulative threshold of the drop, without the need for operators, so real-time monitoring is achieved, internal defects can be found in time, and the trend of increasing failure rate can be adapted. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 is the flowchart of the distribution internal defect monitoring method based on secondary voltage drop identification in this specific embodiment;

[0056] Figure 2 is the electrical connection schematic diagram of adjacent distribution transformers in this specific embodiment;

[0057] Figure 3 is the structural composition diagram of the internal defect monitoring system of the distribution transformer in this specific embodiment. SPECIFIC EMBODIMENT

[0058] The present invention will be further described in detail below with reference to the accompanying drawings:

[0059] Reference Figure 1 As shown, a distribution transformer internal defect monitoring method based on secondary voltage drop identification includes the following steps: establishing a secondary voltage linear regression model of the distribution transformer to be measured, updating the secondary voltage linear regression prediction model of the distribution transformer to be measured daily, predicting the secondary voltage data of the distribution transformer to be measured on the same day, identifying the degree of secondary voltage drop and identifying internal defects. Each step will be specifically described below.

[0060] 1) Establishing a secondary voltage linear regression model of the distribution transformer to be measured

[0061] Based on the secondary side voltage and current relationship formula between the distribution transformer to be measured and the nearest distribution transformer, establish a secondary voltage linear regression model of the distribution transformer to be measured; the nearest distribution transformer refers to the distribution transformer with the shortest electrical distance from the distribution transformer to be measured.

[0062] Specifically, establishing a secondary voltage linear regression model of the distribution transformer to be measured includes the following steps:

[0063] According to the primary side voltage and current relationship formula between the distribution transformer to be measured and the nearest distribution transformer and the transformation relationship between the primary side voltage and the secondary side voltage, establish a secondary side voltage and current relationship formula.

[0064] Reference Figure 2As shown, the distribution transformer to be measured is numbered i, and the distribution transformer with the shortest adjacent electrical distance is numbered j. First, establish the primary side voltage and current relationship between the two transformers:

[0065] V i-1 =V j-1 +I ij Z ij

[0066] Among them, V i-1 、V j-1 represent the effective values of the primary side voltages of the two distribution transformers i and j respectively, and I ij 、Z ij represent the effective value of the transmission line current and the impedance value between the two distribution transformers i and j respectively. According to the voltage transformation relationship of the transformer, establish the secondary side voltage and current relationship between the two transformers:

[0067] N i (V i-2 +I i Z i )=N j (V j-2 +I j Z j )+I ij Z ij

[0068] Among them, V i-2 、V j-2 、I i 、I j 、Z i 、Z j 、N i 、N j represent the effective value of the voltage, the effective value of the secondary side current, the equivalent impedance, and the turns ratio measured by the smart meters connected to the secondary sides of the two distribution transformers i and j respectively.

[0069] Since the impedance of the transmission line between two distribution transformers with the shortest adjacent electrical distance is small, that is, Z ij is small, the impedance of the transmission line between the distribution transformer to be measured and the nearest distribution transformer is ignored to simplify the secondary side voltage and current relationship. The simplified secondary side voltage and current relationship is as follows:

[0070] N i (V i-2 +I i Z i )≈N j (V j-2 +I j Z j )

[0071] Rewrite the simplified secondary - side voltage - current relationship as the secondary - side voltage expression of the distribution transformer under test:

[0072]

[0073] Transform the secondary - side voltage expression of the distribution transformer under test into a linear expression, so as to obtain a secondary - voltage linear regression model:

[0074]

[0075] Among them, i represents the distribution transformer under test, j represents the nearest distribution transformer, V i-2 represents the secondary voltage of the distribution transformer under test, V j-2 represents the secondary voltage of the nearest distribution transformer, I i represents the secondary current of the distribution transformer under test, I j represents the secondary current of the nearest distribution transformer, α0 represents the linear regression constant, and α1, α2, α3 all represent linear regression coefficients; ε represents the random error, and ε reflects that the regression model is not absolutely accurate. The subsequent calculations do not need to calculate this parameter, that is, this parameter is omitted.

[0076] 2) Update the secondary - voltage linear regression prediction model of the distribution transformer under test daily

[0077] Collect the secondary - voltage and secondary - current data of each distribution transformer through the smart meters connected to the secondary side of each distribution transformer respectively. Obtain a single - day sample data set by summarizing the secondary - voltage and secondary - current data daily; summarize the single - day sample data sets of the previous k days, and obtain a k - day sample data set.

[0078] Summarize the k - day sample data set of the distribution transformer under test and the k - day sample data set of the nearest distribution transformer of the distribution transformer under test as the regression calculation data set; solve the linear regression constant and linear regression coefficient of the secondary - voltage linear regression model of the distribution transformer under test on the current day according to the regression calculation data set, so as to obtain the secondary - voltage linear regression prediction model of the distribution transformer under test on the current day.

[0079] Use the multiple - linear - regression calculation method to calculate the linear regression constant and linear regression coefficient. The calculation equations are as follows:

[0080]

[0081] In the formula, respectively represent the average value of the secondary voltage and the average value of the secondary current of the distribution transformer under test in the regression calculation data set; respectively represent the average value of the secondary voltage and the average value of the secondary current of the nearest distribution transformer in the regression calculation data set; L pq all represent calculation notations, p, q = 1, 2, 3;

[0082] L pq The expression of

[0083]

[0084] is as follows:

[0085]

[0086] In the formula, n represents the number of groups of sample data in the regression calculation dataset. Each group of sample data includes the secondary voltage of the distribution transformer to be measured, the secondary current of the distribution transformer to be measured, the secondary voltage of the nearest distribution transformer, and the secondary current of the nearest distribution transformer collected at the same moment; V i-2a 、I ia respectively represent the secondary voltage data and secondary current data of the distribution transformer to be measured in the a-th group of sample data; V j-2a 、I ja respectively represent the secondary voltage data and secondary current data of the nearest distribution transformer in the a-th group of sample data; a = 1, 2, 3..., n.

[0087] 3) Predict the secondary voltage data of the distribution transformer to be measured on the current day

[0088] According to the secondary voltage linear regression prediction model of the distribution transformer to be measured on the previous day, and the secondary current sample dataset in the single-day sample dataset of the distribution transformer to be measured on the current day and the single-day sample dataset of the nearest distribution transformer on the current day, calculate the secondary voltage prediction dataset of the distribution transformer to be measured on the current day.

[0089] The expression of the secondary voltage linear regression prediction model of the distribution transformer to be measured on the current day is as follows:

[0090]

[0091] Among them, V j-2m 、I jm respectively represent the secondary voltage data and secondary current data in the m-th group of sampling data in the single-day sample dataset of the nearest distribution transformer on the current day; I im represents the secondary current data in the m-th group of sampling data in the single-day sample dataset of the distribution transformer to be measured on the current day; represents the predicted value of the secondary voltage of the distribution transformer to be measured on the current day; m = 1, 2, 3,…, M, where M represents the number of groups of sampling data in the single-day sample dataset. Let k = 7 and the sampling interval be 30 minutes, then M = 48 and n = 336.

[0092] 4) Identify the degree of secondary voltage drop

[0093] Calculate the secondary voltage prediction difference set based on the secondary voltage prediction data set of the distribution transformer to be measured on the current day and the secondary voltage sample data set in the single-day sample data set of the distribution transformer to be measured on the current day. The calculation formula for the secondary voltage prediction difference in the secondary voltage prediction data set is as follows:

[0094]

[0095] Where, V i-2m represents the secondary voltage data in the mth group of sampling data in the single-day sample data set of the distribution transformer to be measured on the current day, represents the secondary voltage prediction value of the distribution transformer to be measured on the current day, and ΔV i-2m represents the secondary voltage prediction difference of the distribution transformer to be measured on the current day.

[0096] According to the secondary voltage prediction difference set, use the cumulative sum algorithm to calculate the cumulative drop value of the secondary voltage of the distribution transformer to be measured compared with the prediction data. The calculation formula for the cumulative drop value is as follows:

[0097]

[0098] In the formula, S i (t) and S i (t - 1) represent the cumulative drop values of the distribution transformer to be measured on the current day and the previous day respectively. The calculation formula for the cumulative drop value is as follows:

[0099]

[0100] In the formula, S i (t) and S i (t - 1) represent the cumulative drop values of the distribution transformer to be measured on the current day and the previous day respectively.

[0101] 5) Identify internal defects

[0102] When the cumulative drop value is greater than the cumulative drop threshold, it is determined that the distribution transformer to be measured has internal defects.

[0103] The cumulative drop threshold is determined as follows:

[0104] Step1: Select a test distribution transformer with serious internal defects and the same model as the distribution transformer to be measured. Conduct an open-circuit test on the test distribution transformer under rated operating conditions, that is, connect its primary side to an AC power supply with a voltage level equal to its rated primary voltage, and measure the voltage on its secondary side, which is the secondary side voltage V f-2 ;

[0105] Step2: Calculate the difference ΔV between the secondary side voltage V f-2 of the test distribution transformer under serious internal defects and the rated secondary voltage V f-2N of the test distribution transformerf-2 , and its expression is: ΔV f-2 = V f-2N - V f-2 ;

[0106] Step3: Calculate the cumulative drop value S within p days in the state of severe internal defects fp , where 1 < p < 10, and its expression is: S fp = pMΔV f-2 , where M represents the number of groups of sample data in the single-day sample data set;

[0107] Step4: Calculate the cumulative drop threshold S of the distribution transformer according to the severe defect degree of 0.1(10 - p) T , and its expression is: S T = 0.1(10 - p)S fp . For example, when p takes 6, 7, and 8 respectively, the cumulative drop thresholds S T are 0.4S fp , 0.3S fp , 0.2S fp respectively.

[0108] To implement the distribution transformer internal defect monitoring method based on secondary voltage drop identification of the present invention, the present invention also correspondingly provides a distribution transformer internal defect monitoring system. Refer to Figure 3 as shown, which includes a communication module, a collection module, a storage module, and a data processing module.

[0109] The communication module is used for real-time communication with the distribution network information management system to obtain in real time the secondary voltage data and secondary current data recorded by the smart meters connected to the secondary sides of each distribution transformer in the distribution network and send them to the collection module, and forward the alarm information to the distribution network information management system.

[0110] The collection module is used for regularly summarizing the sampling data of each distribution transformer to form the single-day sample data set, k-day sample data set, and regression calculation data set of each distribution transformer and sending them to the storage module.

[0111] The storage module is used for storing the single-day sample data set, k-day sample data set, and alarm information of each distribution transformer.

[0112] The data processing module is used for daily updating the secondary voltage linear regression prediction model of the distribution transformer to be measured, predicting the secondary voltage data of the distribution transformer to be measured on the same day, identifying the degree of secondary voltage drop, identifying whether the distribution transformer to be measured has internal defects, generating alarm information when it is determined that the distribution transformer to be measured has internal defects, and sending the alarm information to the storage module and the communication module respectively.

[0113] The above technical solution is only one implementation mode of the present invention. For those skilled in the art, based on the disclosed principle of the present invention, it is very easy to make various types of improvements or deformations, and it is not limited to the technical solution described in the above specific embodiments of the present invention. Therefore, the foregoing description is only preferred and does not have a restrictive meaning.

Claims

1. A monitoring method for internal defects of distribution transformers based on the identification of secondary voltage dips, characterized in that, Including the following steps: Establish a quadratic voltage linear regression model for the distribution transformer to be measured: Based on the quadratic side voltage and current relationship formula between the distribution transformer to be measured and the nearest distribution transformer, establish a quadratic voltage linear regression model for the distribution transformer to be measured; the nearest distribution transformer refers to the distribution transformer with the shortest electrical distance from the distribution transformer to be measured; Update the quadratic voltage linear regression prediction model of the distribution transformer to be measured daily: Collect the secondary voltage and secondary current data of each distribution transformer through the smart meters connected to the secondary side of each distribution transformer, and obtain a single-day sample data set by summarizing the secondary voltage and secondary current data daily; summarize the single-day sample data sets of the days before the current day k to obtain k a daily sample data set; Summarize the k daily sample data set of the distribution transformer to be measured and the k daily sample data set of the nearest distribution transformer to be measured, as the regression calculation data set; Solve the linear regression constant and linear regression coefficient of the quadratic voltage linear regression model of the distribution transformer to be measured on the current day according to the regression calculation data set, so as to obtain the quadratic voltage linear regression prediction model of the distribution transformer to be measured on the current day; Predict the quadratic voltage data of the distribution transformer to be measured on the current day: According to the quadratic voltage linear regression prediction model of the distribution transformer to be measured on the previous day, and the quadratic current sample data set in the single-day sample data set of the distribution transformer to be measured on the current day and the single-day sample data set of the nearest distribution transformer on the current day, calculate the quadratic voltage prediction data set of the distribution transformer to be measured on the current day; Identify the degree of quadratic voltage drop: According to the quadratic voltage prediction data set of the distribution transformer to be measured on the current day and the quadratic voltage sample data set in the single-day sample data set of the distribution transformer to be measured on the current day, calculate the quadratic voltage prediction difference set; According to the quadratic voltage prediction difference set, use the cumulative sum algorithm to calculate the cumulative drop value of the quadratic voltage of the distribution transformer to be measured compared with the predicted data; Identify internal defects: When the cumulative drop value is greater than the cumulative drop threshold, it is determined that the distribution transformer to be measured has internal defects.

2. The method for monitoring internal defects of a distribution transformer based on secondary voltage sag identification according to claim 1, wherein Establishing a quadratic voltage linear regression model for the distribution transformer to be measured includes the following steps: Establish a quadratic side voltage and current relationship formula according to the primary side voltage and current relationship formula and the transformation relationship between the primary side voltage and the quadratic side voltage between the distribution transformer to be measured and the nearest distribution transformer; Ignore the transmission line impedance between the distribution transformer to be measured and the nearest distribution transformer to simplify the quadratic side voltage and current relationship formula, and rewrite the simplified quadratic side voltage and current relationship formula as the quadratic side voltage expression of the distribution transformer to be measured; Transform the quadratic side voltage expression of the distribution transformer to be measured into a linear expression, so as to obtain a quadratic voltage linear regression model.

3. The method for monitoring internal defects of a distribution transformer based on the identification of secondary voltage dips according to claim 2, characterized in that, The expression of the quadratic voltage linear regression model of the distribution transformer to be measured is: Among them, i represents the distribution transformer to be measured, j represents the nearest distribution transformer, represents the secondary voltage of the distribution transformer to be measured, represents the secondary voltage of the nearest distribution transformer, represents the secondary current of the distribution transformer to be measured, represents the secondary current of the nearest distribution transformer, represents the linear regression constant, both represent the linear regression coefficients, represents the random error.

4. The method for monitoring internal defects of a distribution transformer based on the identification of secondary voltage dips according to claim 3, wherein Use the multiple linear regression calculation method to calculate the linear regression constant and linear regression coefficient, and the calculation equation is as follows: In the formula, respectively represent the average secondary voltage and the average secondary current of the distribution transformer under test in the regression calculation dataset; respectively represent the average secondary voltage and the average secondary current of the nearest distribution transformer in the regression calculation dataset; both represent calculation notations, p , q = 1, 2, 3; The expression is as follows: The expression is as follows: In the formula, n represents the number of groups of sample data in the regression calculation dataset, and each group of sample data includes the secondary voltage of the distribution transformer under test, the secondary current of the distribution transformer under test, the secondary voltage of the nearest distribution transformer, and the secondary current of the nearest distribution transformer collected at the same moment; respectively represent the secondary voltage data and secondary current data of the distribution transformer under test in the -th group of sample data; respectively represent the secondary voltage data and secondary current data of the nearest distribution transformer in the -th n group of sample data; = 1, 2, 3, …, n .

5. The method for monitoring internal defects of a distribution transformer based on the identification of secondary voltage dips according to claim 4, characterized in that, The expression of the quadratic voltage linear regression prediction model of the distribution transformer to be measured on the current day is as follows: Among them, respectively represent the secondary voltage data and secondary current data in the m th group of sampling data in the single-day sample dataset of the nearest distribution transformer on the current day; represents the secondary current data in the m th group of sampling data in the single-day sample dataset of the distribution transformer to be measured on the current day; represents the predicted value of the secondary voltage of the distribution transformer to be measured on the current day; m = 1, 2, 3, …, M, where M represents the number of groups of sampling data in the single-day sample dataset.

6. The method for monitoring internal defects of a distribution transformer based on the identification of secondary voltage dips according to claim 5, characterized in that, The calculation formula of the quadratic voltage prediction difference is as follows: Among them, represents the secondary voltage data in the m th group of sampling data in the single-day sample dataset of the distribution transformer to be measured on the current day, represents the predicted value of the secondary voltage of the distribution transformer to be measured on the current day, represents the predicted difference of the secondary voltage of the distribution transformer to be measured on the current day.

7. The method for monitoring internal defects of a distribution transformer based on the identification of secondary voltage dips according to claim 1, characterized in that The calculation formula of the cumulative drop value is as follows: In the formula, respectively represent the cumulative drop values of the distribution transformer to be measured on the current day and the previous day.

8. The method for monitoring internal defects of a distribution transformer based on secondary voltage sag identification according to claim 5, wherein Let k = 7, and the sampling interval is 30 minutes, then M = 48, n = 336.

9. The method for monitoring internal defects of a distribution transformer based on the identification of secondary voltage dips according to claim 1, characterized in that The cumulative drop threshold is determined as follows: Step 1: Select a test distribution transformer with serious internal defects and the same model as the distribution transformer to be tested. Conduct an no-load test on the test distribution transformer under rated conditions, that is, connect its primary side to an AC power supply with a voltage level equal to its rated primary voltage, and measure the voltage on its secondary side, which is the secondary side voltage of the test distribution transformer in the state of serious internal defects. V f-2 ; Step 2: Calculate the secondary side voltage of the test distribution transformer under severe internal defects and the rated secondary voltage of the test distribution transformer difference , and its expression is: ; Step 3: Calculate the cumulative value of drops during p the number of days in the severe internal defect state , where 1 < p < 10, and its expression is: , M represents the number of groups of sampled data in the single-day sample dataset; Step4: Calculate the cumulative threshold of the distribution transformer dropout according to the severe defect level of 0.1 (10 - p ), and its expression is: . .

10. A monitoring system for internal defects of a distribution transformer, characterized in that, Used to execute the distribution transformer internal defect monitoring method based on quadratic voltage drop identification as described in any one of claims 1 to 9, and includes a communication module, a collection module, a storage module, and a data processing module; The communication module is used for real-time communication with the distribution network information management system to obtain the quadratic voltage data and quadratic current data recorded by the smart meters connected to the secondary side of each distribution transformer in the distribution network in real time and send them to the collection module, and forward the alarm information to the distribution network information management system; The acquisition module is used to regularly summarize the sampling data of each distribution transformer, so as to form a single-day sample data set for each distribution transformer, k a daily sample data set, a regression calculation data set, and send them to the storage module; The storage module is used to store the single-day sample data sets of each distribution transformer, k daily sample data sets, and alarm information; The data processing module is used to update the quadratic voltage linear regression prediction model of the distribution transformer to be measured daily, predict the quadratic voltage data of the distribution transformer to be measured on the same day, identify the degree of quadratic voltage drop, identify whether there are internal defects in the distribution transformer to be measured, generate an alarm message when it is determined that there are internal defects in the distribution transformer to be measured, and send the alarm message to the storage module and the communication module respectively.

Images