Online detection method, device and storage medium for out-of-tolerance electric meter considering suspicion coefficient

Abstract

The online detection method, device and storage medium of out-of-tolerance electric meters considering the suspicion coefficient provided in the present application can first determine the first electricity data of the total meter in the target area in a preset historical period and the second electricity data of the sub-meters whose electricity is greater than the preset electricity threshold in the preset historical period when detecting the smart meters in the target area. Then, based on the first electricity data and the second electricity data corresponding to each sub-meter, the line loss electricity, sub-meter electricity, sub-meter relative error and confidence interval of the sub-meter relative error of each sub-meter at each moment in the preset historical period are determined, and the average suspicion coefficient of each sub-meter in the preset historical period is calculated through the confidence interval, so as to analyze the comprehensive influence of the relative errors of the sub-meters at different moments, and obtain the suspected electric meters through a series of measures from a probability perspective, so that the detection result is more objective, the on-site false detection rate is lower, and the detection rate is higher.

Description

Technical Field

[0001] The present application relates to the technical field of smart meter detection, and in particular to an online detection method, device, storage medium and computer equipment for out-of-tolerance meters taking suspicion coefficient into consideration. Background Art

[0002] At present, all industries are striving to promote digital construction, and power grid companies have put into operation a large number of smart meters to collect massive amounts of electric energy data with high accuracy and good data quality. However, with the increase in the service life and the influence of factors such as environmental effects, equipment abnormalities, and human damage, the measurement accuracy of smart meters will decrease, thereby reducing the data quality.

[0003] At present, smart meters mainly perform error detection through periodic on-site manual inspection. This method is inefficient, has a small detection range, consumes a lot of manpower and material resources, and has high maintenance costs. It is difficult to meet the requirements of smart grid digital transformation and daily operation and maintenance needs. Summary of the invention

[0004] The purpose of this application is to solve at least one of the above-mentioned technical defects, especially the technical defects of the prior art that error detection is performed through periodic on-site manual detection. This method is inefficient, has a small detection range, consumes a lot of manpower and material resources, and has high maintenance costs, making it difficult to meet the requirements of the digital transformation of smart grids and the needs of daily operation and maintenance.

[0005] The present application provides an online detection method for an out-of-tolerance electric meter taking into account a suspicion coefficient, the method comprising:

[0006] Determine the first power data of the total meter of the target area in the preset historical period and the second power data of the sub-meters whose power is greater than the preset power threshold in the preset historical period;

[0007] According to the first electricity quantity data and the second electricity quantity data corresponding to each substation table, determine the line loss electricity quantity, sub-meter electricity quantity, sub-meter relative error and the confidence interval of the sub-meter relative error of each substation table at each moment in the preset historical period;

[0008] Based on the confidence interval of the relative error of each differentiation table at each moment in the preset historical period, the average suspicion coefficient of each differentiation table in the preset historical period is calculated;

[0009] According to the line loss electricity and sub-meter electricity of each differentiated meter at each moment in the preset historical period, the correlation between the line loss electricity and sub-meter electricity of each differentiated meter in the preset historical period is calculated, and the average suspicion coefficient correction factor of each differentiated meter in the preset historical period is determined by using the correlation influence mechanism;

[0010] The average suspicion coefficient correction factor of each station division table is used to correct the average suspicion coefficient of the corresponding station division table, and the station division tables are sorted in descending order according to the corrected average suspicion coefficient, and the station division table with the largest ranking is taken as the out-of-tolerance meter.

[0011] Optionally, determining the first power data of the total meter of the target area in a preset historical period and the second power data of the sub-meters whose power is greater than a preset power threshold in the preset historical period includes:

[0012] Determine the first original data of the total table of the target area in the preset historical period and the second original data of each area table in the preset historical period;

[0013] Performing a preprocessing operation on the first original data and the second original data to obtain first electric quantity data and second electric quantity data, wherein the preprocessing operation includes removing abnormal data, filling missing data, correcting clock deviation data, and correcting user variable data using user variable files;

[0014] According to the second electricity data of each substation table, the substation tables whose sub-meter electricity is not greater than the preset electricity threshold in the substation tables are eliminated to obtain the second electricity data of each sub-meter whose sub-meter electricity is greater than the preset electricity threshold in the preset historical period.

[0015] Optionally, the preprocessing the first original data and the second original data to obtain the first electric quantity data and the second electric quantity data includes:

[0016] Acquire the area data of the target area; wherein the area data includes the household change file of the electric meter, the first original data includes the clock deviation of the total meter and the total meter power of the area total meter at each moment in the preset historical period, and the second original data includes the clock deviation of the sub-meter and the sub-meter power of each sub-meter at each moment in the preset historical period;

[0017] Using a statistical discrimination method to eliminate abnormal data in the station area data, the first original data, and the second original data;

[0018] Filling the missing data in the area data after the abnormal data is removed, the first original data, and the second original data;

[0019] Using the clock deviation of the total meter and the clock deviation of the sub-meters of each substation, the total meter power of the substation total meter at each moment in the preset historical period and the sub-meter power of each substation substation at each moment in the preset historical period after filling in the missing data are corrected to obtain second power data;

[0020] According to the meter household change files in the area data after the missing data is filled, the total meter electricity of the corrected area total meter at each moment in the preset historical period is further corrected to obtain the first electricity data.

[0021] Optionally, according to the second power data of each substation table, the substation tables whose sub-meter power is not greater than a preset power threshold are removed, and the second power data of each substation table whose sub-meter power is greater than the preset power threshold in the preset historical period is obtained, including:

[0022] Determine the sub-metered electric quantity of each sub-meter at each moment in the preset historical period according to the second electric quantity data of each sub-meter in the preset historical period;

[0023] The median 1 / K of the sub-meter electricity of each sub-meter in each sub-meter in the preset historical period is used as the low electricity threshold at that moment, and at each moment, the sub-meter in each sub-meter whose electricity is less than the low electricity threshold at that moment is assigned a judgment value of 1, and the other sub-meters are assigned a judgment value of 0;

[0024] The average value of the sum of the judgment values ​​of each sub-meter at all times in the preset historical period is calculated, and the sub-meters with average values ​​greater than the preset probability threshold in the sub-meters are eliminated as sub-meters with power not greater than the preset power threshold, so as to obtain the second power data of the sub-meters with power greater than the preset power threshold in the preset historical period.

[0025] Optionally, determining the relative error of each sub-meter at each moment in the preset historical period and the confidence interval of the relative error of each sub-meter according to the first power data and the second power data corresponding to each sub-meter, includes:

[0026] Using a sliding window of preset length, the first electricity data of the total meter of the substation area in the preset historical period and the second electricity data of the substation area table in the preset historical period are divided into time periods to obtain the total meter electricity and the sub-meter electricity in each time period;

[0027] Based on the law of conservation of energy, a group of equations for solving relative errors of sub-meters corresponding to the total meter electricity and sub-meter electricity in multiple time periods is established, and after solving the group of equations for solving relative errors of sub-meters, a point estimate of the desired quantity is obtained, wherein the desired quantity includes the relative error of the sub-meter, the line loss rate, and the fixed loss;

[0028] The relative error of each table for each area is determined according to the point estimate of the quantity to be determined, and the confidence interval corresponding to the relative error of each table for each area is calculated using the T test in the significance test.

[0029] Optionally, the calculating the average suspicion coefficient of each differentiation table in the preset historical period based on the confidence interval of the relative error of each differentiation table at each moment in the preset historical period includes:

[0030] The suspicion coefficient corresponding to the confidence interval of the relative error of each partition meter at each moment in the preset historical period is calculated by using a credible range confirmation mechanism;

[0031] According to the suspicion coefficient of each classification table at each moment in the preset historical period, the average suspicion coefficient of each classification table in the preset historical period is calculated.

[0032] Optionally, the calculation of the correlation between the line loss electricity and the sub-meter electricity of each differentiated meter in the preset historical period according to the line loss electricity and the sub-meter electricity of each differentiated meter at each moment in the preset historical period, and the determination of the average suspicion coefficient correction factor of each differentiated meter in the preset historical period by using the correlation influence mechanism include:

[0033] According to the line loss electricity and sub-meter electricity of each differentiated meter at each moment in the preset historical period, the correlation between the line loss electricity and the sub-meter electricity of each differentiated meter in the preset historical period is calculated, and the correlation includes negative correlation and positive correlation;

[0034] For each station differentiation table, when the line loss electricity and the sub-meter electricity of the station differentiation table in the preset historical period are negatively correlated, the average suspicion coefficient correction factor of the station differentiation table in the preset historical period is set to a value greater than 1;

[0035] When there is a positive correlation between the line loss electricity and the sub-meter electricity of the station differentiation meter in the preset historical period, the average suspicion coefficient correction factor of the station differentiation meter in the preset historical period is set to a value less than 1.

[0036] The present application also provides an out-of-tolerance electric meter online detection device taking into account a suspicion coefficient, comprising:

[0037] A data acquisition module, used to determine the first power data of the total meter of the target area in the preset historical period and the second power data of the sub-meters whose power is greater than the preset power threshold in the preset historical period;

[0038] An error determination module, configured to determine the line loss electricity, sub-meter electricity, sub-meter relative error and confidence interval of the sub-meter relative error of each sub-meter at each moment in the preset historical period according to the first electricity data and the second electricity data corresponding to each sub-meter;

[0039] A suspicion coefficient calculation module, used to calculate the average suspicion coefficient of each differentiation table in the preset historical period based on the confidence interval of the relative error of each differentiation table at each moment in the preset historical period;

[0040] A correction factor determination module is used to calculate the correlation between the line loss electricity and the sub-meter electricity of each differentiated meter in the preset historical period according to the line loss electricity and the sub-meter electricity of each differentiated meter at each moment in the preset historical period, and determine the average suspicion coefficient correction factor of each differentiated meter in the preset historical period by using the correlation influence mechanism;

[0041] The out-of-tolerance meter detection module is used to correct the average suspicion coefficient of the corresponding station division table using the average suspicion coefficient correction factor of each station division table, and sort the station division tables in descending order according to the corrected average suspicion coefficient, and take the station division table with the largest ranking as the out-of-tolerance meter.

[0042] The present application also provides a storage medium, in which computer-readable instructions are stored. When the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the online detection method of out-of-tolerance electric meters considering the suspicion coefficient as described in any of the above embodiments.

[0043] The present application also provides a computer device, comprising: one or more processors, and a memory;

[0044] The memory stores computer-readable instructions, and when the computer-readable instructions are executed by the one or more processors, the steps of the online detection method for out-of-tolerance electric meters taking into account the suspicion coefficient as described in any one of the above embodiments are performed.

[0045] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0046] The present application provides an online detection method, device and storage medium for out-of-tolerance meters taking suspicion coefficients into consideration. When detecting smart meters in a target area, the first electricity data of the total meter in the target area in a preset historical period and the second electricity data of the area meters whose electricity is greater than a preset electricity threshold in a preset historical period can be first determined, so as to avoid the influence of the small electricity area meters on the final result; then, the present application can determine the line loss electricity, sub-meter electricity, sub-meter relative error and confidence interval of the sub-meter relative error of each area meter at each moment in the preset historical period according to the first electricity data and the second electricity data corresponding to each area meter, and calculate the average suspicion coefficient of each area meter in the preset historical period through the confidence interval, so as to analyze the comprehensive influence of the relative errors of the sub-meters at different moments, and obtain the suspected electricity meter from a probability perspective through a series of measures, so that the detection result is more objective, the on-site false detection rate is lower, and the detection rate is higher; secondly, the present application takes into account that when the line loss electricity and the sub-meter electricity are strongly correlated, the shortage and sub-meter electricity of the line loss electricity The shortfall of meter electricity will be transferred to each other, and the relative error of the sub-meter obtained is inaccurate. Based on this, the present application calculates the correlation between the line loss electricity and the sub-meter electricity of each differentiation meter in the preset historical period according to the line loss electricity and the sub-meter electricity of each differentiation meter at each moment in the preset historical period, and uses the correlation influence mechanism to determine the average suspicion coefficient correction factor of each differentiation meter in the preset historical period, and uses the average suspicion coefficient correction factor to correct the average suspicion coefficient of the corresponding differentiation meter, so as to eliminate the above-mentioned influencing factors from the perspective of probability, thereby improving the accuracy of identifying out-of-tolerance meters; further, after obtaining the corrected average suspicion coefficient of each differentiation table, the present application can sort each differentiation table in descending order according to the corrected average suspicion coefficient, and use the differentiation table with the largest ranking as the out-of-tolerance meter, so that the out-of-tolerance meter detected in this way has higher accuracy, and the process does not require manual detection, thereby effectively improving detection efficiency and reducing maintenance costs, and meeting the requirements of digital transformation of smart grids and daily operation and maintenance needs to a great extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art 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 paying creative labor.

[0048] Figure 1 A flow chart of an online detection method for an out-of-tolerance electric meter taking into account a suspicion coefficient provided in an embodiment of the present application;

[0049] Figure 2A schematic diagram of the structure of an online detection device for out-of-tolerance electric meter taking suspicion coefficient into consideration provided in an embodiment of the present application;

[0050] Figure 3 A schematic diagram of the internal structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only 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 creative work are within the scope of protection of this application.

[0052] At present, all industries are striving to promote digital construction, and power grid companies have put into operation a large number of smart meters to collect massive amounts of electric energy data with high accuracy and good data quality. However, with the increase in the service life and the influence of factors such as environmental effects, equipment abnormalities, and human damage, the measurement accuracy of smart meters will decrease, thereby reducing the data quality.

[0053] At present, smart meters mainly perform error detection through periodic on-site manual detection. This method is inefficient, has a small detection range, consumes a lot of manpower and material resources, and has high maintenance costs. It is difficult to meet the requirements of smart grid digital transformation and daily operation and maintenance needs. Based on this, this application proposes the following technical solutions, see below for details:

[0054] In one embodiment, Figure 1 As shown, Figure 1 A flow chart of an online detection method for an out-of-tolerance meter taking into account a suspicion coefficient provided in an embodiment of the present application; the present application provides an online detection method for an out-of-tolerance meter taking into account a suspicion coefficient, and the method may include:

[0055] S110: Determine the first power data of the total meter in the target area in a preset historical period and the second power data of the sub-meters whose power is greater than a preset power threshold in the preset historical period.

[0056] In this step, when performing online detection on the smart meter, the target substation can be determined first, and then the first electricity data of the substation total meter in the target substation in the preset historical period and the second electricity data of the substation meters whose electricity quantities are greater than the preset electricity threshold in the preset historical period are obtained, so as to calculate the relative errors of the sub-meters of each substation meter through the first electricity data and the second electricity data, and determine which smart meters are out-of-tolerance meters through the relative errors of the sub-meters.

[0057] Among them, the preset historical time period of the present application can be set according to the number of substation tables in the target substation and the accuracy of the detection results. When the number of substation tables in the target substation is large, in order to obtain more accurate detection results, the preset historical time period can be set to a longer time period. However, considering that the longer the time period, the greater the amount of data to be calculated, therefore, it can be set according to the actual situation, such as setting it to collect the first electricity data and the second electricity data within the past month, or two months, three months, etc., without any restriction here.

[0058] It can be understood that the substation in this application refers to the power supply range or area of ​​(a) transformer, and the target substation is the area where the out-of-tolerance meter needs to be detected online in this application. The out-of-tolerance meter here refers to the metering error range of the electric energy meter that exceeds the accuracy of the electric energy meter. The out-of-tolerance electric energy meter has positive out-of-tolerance and negative out-of-tolerance. The out-of-tolerance electric energy meter is no longer accurate and can no longer be used, so online detection is required.

[0059] Furthermore, in actual application, each substation will be equipped with a substation master meter and several substation submeters. The substation master meter and substation submeters are independently measured, and different substation submeters are used to measure the electricity of different users. If line loss and fixed loss are not considered, the total meter electricity of the substation master meter is equal to the sum of the submeter electricity of each substation submeter. However, since line loss and fixed loss are objective, the total meter electricity of the substation master meter is generally greater than the sum of the submeter electricity of each substation submeter.

[0060] Therefore, when the present application needs to detect out-of-tolerance electricity meters, in order to improve the detection accuracy, the first electricity data of the substation total meter in the preset historical period and the second electricity data of each substation meter in the preset historical period can be combined for analysis. In this way, not only the sub-meter electricity of each substation meter at each moment in the preset historical period can be obtained, but also the line loss electricity of each substation meter can be obtained. Furthermore, the present application can also use the first electricity data of the substation total meter and the second electricity data of each substation meter to calculate the sub-meter relative error of each substation meter, and use this to detect out-of-tolerance electricity meters.

[0061] Furthermore, when the present application detects each substation meter in the target substation, due to the different commissioning time of each substation meter, some substation meters may have just been put into operation, and some substation meters may have been in operation for a period of time; or due to the influence of complex factors in the target substation itself, some users have low electricity consumption, and some users have high electricity consumption, which will also lead to a certain difference in the electricity consumption of each sub-meter in the substation meter. When there is a small electricity sub-meter in the substation meter, it may affect the judgment of the detection result of the over-tolerance meter. For example, when the substation meter error is estimated by using the multivariate linear regression method or its improved algorithm to obtain the relative error of the sub-meter, the condition number of the substation meter electricity matrix of the multivariate linear regression equation group will increase due to the small electricity sub-meter, which may cause the equation to be ill-conditioned and reduce the stability of the solution result. Therefore, in order to avoid the misjudgment of the over-tolerance meter caused by the small electricity sub-meter, the present application can screen the first electricity data of each substation meter when obtaining it, and retain the second electricity data of the substation meter whose electricity is greater than the preset electricity threshold in the preset historical period.

[0062] S120: Determine the line loss electricity, sub-meter electricity, sub-meter relative error and confidence interval of the sub-meter relative error of each sub-meter at each moment in a preset historical period according to the first electricity data and the second electricity data corresponding to each substation submeter.

[0063] In this step, after obtaining the first electricity data of the total meter in the target substation in the preset historical period and the second electricity data of the substation meters whose electricity is greater than the preset electricity threshold in the preset historical period through S110, the present application can determine the line loss electricity, sub-meter electricity, sub-meter relative error and the confidence interval corresponding to the sub-meter relative error of each substation meter at each moment in the preset historical period based on the first electricity data and the second electricity data corresponding to each substation meter.

[0064] It is understandable that, since most of the prior art equates the identification of out-of-tolerance meters with the calculation of meter errors, this approach is actually too ideal and unnecessary. The reasons are: first, the accuracy of the measured data used for meter error estimation is limited, resulting in inaccurate error estimation results; second, assuming that the error estimation results show that a certain meter is out of tolerance, no matter how much the out-of-tolerance is, it requires on-site manual secondary verification; third, according to the simulation experiment results, the meter error calculation results are time-varying, and there may be large differences at different times. It is difficult for the prior art to combine multiple error results for out-of-tolerance judgment. Therefore, a more effective approach is shown in this application. After obtaining the first power data of the total meter in the target area in the preset historical period and the second power data of the sub-meter meter in the preset historical period whose power is greater than the preset power threshold, the relative error of each sub-meter meter at each moment in the preset historical period and the confidence interval of the relative error of each sub-meter meter are calculated according to the first power data and the second power data corresponding to each sub-meter meter, and the probability of the meter out of tolerance is judged by the confidence interval of the relative error of the sub-meter, thereby obtaining the suspected meter.

[0065] Furthermore, when considering the influence of line loss on the estimation of meter error, the existing methods often reduce the influence of line loss randomness on the meter error solution equation by estimating line loss, but line loss is time-varying and difficult to be accurately estimated. And according to the simulation experiment results, for the meter error solution equation, when the line loss and sub-meter electricity are strongly correlated, the shortage of line loss and the shortage of sub-meter electricity will be transferred to each other, and the sub-meter error obtained is inaccurate. Therefore, the present application can calculate the line loss and sub-meter electricity of each sub-meter at each moment in a preset historical period through the first electricity data and the second electricity data corresponding to each sub-meter, and determine the correlation between the line loss and sub-meter electricity at each moment in the preset historical period, so as to further screen the suspected meters obtained using the relative errors of the sub-meters according to the correlation, thereby improving the recognition accuracy of out-of-tolerance meters.

[0066] S130: Calculate the average suspicion coefficient of each differentiation table in the preset historical period based on the confidence interval of the relative error of each differentiation table at each moment in the preset historical period.

[0067] In this step, after determining the line loss electricity, sub-meter electricity, sub-meter relative error and confidence interval of the sub-meter relative error of each sub-meter at each moment in the preset historical period through S120, in order to further screen the suspected electricity meters obtained using the sub-meter relative error through the correlation between the line loss electricity and the sub-meter electricity, the present application can calculate the average suspicion coefficient of each sub-meter in the preset historical period according to the confidence interval of the sub-meter relative error of each sub-meter at each moment in the preset historical period. In this way, the average suspicion coefficient can be corrected by using the correlation between the line loss electricity and the sub-meter electricity, so as to obtain a more accurate identification result.

[0068] It can be understood that since the confidence interval obtained in the present application is the confidence interval corresponding to the relative error of each station area table at each moment in the preset historical period, when calculating the average suspicion coefficient of each station area table in the preset historical period, the suspicion coefficient of each station area table at each moment in the preset historical period can be calculated first, and then the final average suspicion coefficient can be obtained based on the suspicion coefficient at each moment.

[0069] S140: Calculate the correlation between the line loss electricity and the sub-meter electricity of each differentiated meter in the preset historical period based on the line loss electricity and the sub-meter electricity of each differentiated meter at each moment in the preset historical period, and determine the average suspicion coefficient correction factor of each differentiated meter in the preset historical period using the correlation influence mechanism.

[0070] In this step, after determining the line loss electricity, sub-meter electricity, sub-meter relative error and confidence interval of the sub-meter relative error of each substation meter at each moment in the preset historical period through S120, in order to avoid the situation where the calculation result of the sub-meter relative error is inaccurate due to the influence of the correlation between the line loss electricity and the sub-meter electricity when solving the sub-meter relative error of each substation meter, the present application can calculate the line loss electricity and the sub-meter electricity of each substation meter at each moment in the preset historical period through the first electricity data and the second electricity data corresponding to each substation meter, and determine the correlation between the line loss electricity and the sub-meter electricity at each moment in the preset historical period. Then, the present application can use the correlation influence mechanism to determine the average suspicion coefficient correction factor of each substation meter in the preset historical period, so that the average suspicion coefficient obtained in S130 can be corrected by the average suspicion coefficient correction factor, so as to improve the recognition accuracy of out-of-tolerance meters.

[0071] Among them, the correlation influence mechanism of the present application refers to the influence on the average suspicion coefficient of a station-divided meter determined according to the correlation between the line loss electricity and the sub-meter electricity of a station-divided meter within a preset time period. For example, when the line loss electricity of a station-divided meter is negatively correlated with the electricity of a sub-meter, this situation is consistent with the performance of electricity theft, so the average suspicion coefficient of the station-divided meter should be increased; and when the line loss electricity and the sub-meter electricity of a station-divided meter are negatively correlated, when the sub-meter relative error of the station-divided meter is estimated using the multivariate linear regression method or its improved algorithm, the sub-meter electricity and the line loss electricity have the same status in mathematics, so the error will be shared between the relative error of the sub-meter and the line loss rate, resulting in deviations in the calculation results of both, and at this time the average suspicion coefficient of the station-divided meter should be reduced.

[0072] S150: using the average suspicion coefficient correction factor of each station division table to correct the average suspicion coefficient of the corresponding station division table, and sorting each station division table in descending order according to the corrected average suspicion coefficient, and taking the station division table with the largest ranking as the out-of-tolerance meter.

[0073] In this step, after obtaining the average suspicion coefficient and the average suspicion coefficient correction factor corresponding to each station classification table, the present application can use the average suspicion coefficient correction factor to correct the average suspicion coefficient, and sort each station classification table in descending order according to the corrected average suspicion coefficient, and finally take the station classification table with the largest ranking as the out-of-tolerance meter.

[0074] Specifically, when correcting the average suspicion coefficient of the station area table, the average suspicion coefficient correction factor corresponding to the station area table can be multiplied by the corresponding average suspicion coefficient to obtain the corrected average suspicion coefficient. When the average suspicion coefficients of all station area tables are corrected, the present application can sort the average suspicion coefficients of each station area table in descending order. The station area table with the largest ranking indicates that its suspicion coefficient is the highest. Therefore, the station area table can be used as an out-of-tolerance meter. Furthermore, since there may be multiple station area tables with the largest ranking, when there are multiple station area tables with the largest corrected average suspicion coefficients, all station area tables with the largest ranking can be used as out-of-tolerance meters.

[0075] In the above embodiment, when the smart meter in the target area is detected, the first electricity data of the area total meter in the target area in the preset historical period and the second electricity data of the area meter whose electricity of each sub-meter is greater than the preset electricity threshold in the preset historical period can be determined first, so as to avoid the influence of the small electricity area meter on the final result; then, the present application can determine the line loss electricity, sub-meter electricity, sub-meter relative error and confidence interval of the sub-meter relative error of each area meter at each moment in the preset historical period according to the first electricity data and the second electricity data corresponding to each area meter, and calculate the average suspicion coefficient of each area meter in the preset historical period through the confidence interval, so as to analyze the comprehensive influence of the relative error of the sub-meter at different moments, and obtain the suspected electricity meter through a series of measures from the probability perspective, so that the detection result is more objective, the on-site false detection rate is lower, and the detection rate is higher; secondly, the present application takes into account that when the line loss electricity and the sub-meter electricity are strongly correlated, the shortage of line loss electricity and the shortage of sub-meter electricity will be transferred to each other, and the confidence interval of the sub-meter relative error will be transferred to each other. The relative error of the sub-meter obtained in this way is inaccurate. Based on this, the present application calculates the correlation between the line loss electricity and the sub-meter electricity of each differentiation meter in the preset historical period according to the line loss electricity and the sub-meter electricity of each differentiation meter at each moment in the preset historical period, and uses the correlation influence mechanism to determine the average suspicion coefficient correction factor of each differentiation meter in the preset historical period, and uses the average suspicion coefficient correction factor to correct the average suspicion coefficient of the corresponding differentiation meter, so as to eliminate the above-mentioned influencing factors from the perspective of probability, thereby improving the accuracy of identifying out-of-tolerance meters; further, after obtaining the corrected average suspicion coefficient of each differentiation table, the present application can sort each differentiation table in descending order according to the corrected average suspicion coefficient, and take the differentiation table with the largest ranking as the out-of-tolerance meter, so that the out-of-tolerance meter detected in this way has higher accuracy, and the process does not require manual detection, thereby effectively improving detection efficiency and reducing maintenance costs, and meeting the requirements of digital transformation of smart grids and daily operation and maintenance needs to a great extent.

[0076] In one embodiment, determining in S110 the first power data of the total meter in the target meter area during the preset historical period and the second power data of the sub-meters whose power is greater than the preset power threshold during the preset historical period may include:

[0077] S111: Determine the first original data of the total table of the target area in a preset historical period and the second original data of each area table in the preset historical period.

[0078] S112: Preprocessing the first original data and the second original data to obtain first power data and second power data, wherein the preprocessing operation includes removing abnormal data, filling missing data, correcting clock deviation data, and correcting user variable data using user variable files.

[0079] S113: Based on the second electricity data of each substation table, the substation tables whose sub-meter electricity is not greater than a preset electricity threshold are eliminated, and the second electricity data of each sub-meter whose sub-meter electricity is greater than the preset electricity threshold in the preset historical period are obtained.

[0080] In this embodiment, when determining the first power data of the total meter of the area in the target area in the preset historical period, the first original data of the total meter of the area in the preset historical period can be obtained first, and then the first original data can be preprocessed to obtain the first power data of the total meter of the area in the preset historical period. When determining the second power data of the area table in the preset historical period in which the power of each sub-meter in the target area is greater than the preset power threshold, the second original data of all the area tables in the target area in the preset historical period can be obtained first, and then the second original data can be preprocessed to eliminate the area tables in which the power of the sub-meter is not greater than the preset power threshold according to the second power data of each area table, so as to obtain the second power data of the area table in the preset historical period in which the power of each sub-meter is greater than the preset power threshold.

[0081] Furthermore, the prior art does not consider the data quality of the measured data when identifying out-of-tolerance meters, lacks comprehensive data repair methods, including the lack of consideration of factors such as data anomalies, missing data, clock deviations, and household change files, resulting in a low accuracy rate of the final recognition result. Therefore, before using the first original data and the second original data for data analysis, the present application may perform a preprocessing operation on the first original data and the second original data, and the preprocessing operation may include operations such as removing abnormal data, filling missing data, correcting clock deviation data, and using household change files to correct household change data, thereby effectively improving the recognition accuracy of out-of-tolerance meters.

[0082] In one embodiment, performing a preprocessing operation on the first original data and the second original data in S112 to obtain the first power data and the second power data may include:

[0083] S1121: Obtain the substation data of the target substation; wherein the substation data includes the household change file of the electric meter, the first original data includes the clock deviation of the total meter and the total meter power of the substation total meter at each moment in the preset historical period, and the second original data includes the clock deviation of the sub-meter and the sub-meter power of each substation submeter at each moment in the preset historical period.

[0084] S1122: using a statistical discrimination method to eliminate abnormal data in the station area data, the first original data, and the second original data.

[0085] S1123: Fill in the missing data in the area data after the abnormal data is removed, the first original data, and the second original data.

[0086] S1124: Utilizing the clock deviation of the total meter and the clock deviation of the sub-meters of each substation, correct the total meter power of the substation area total meter at each moment in the preset historical period and the sub-meter power of each substation area meter at each moment in the preset historical period after filling in the missing data to obtain the second power data.

[0087] S1125: According to the meter household change files in the area data after the missing data is filled, the total meter electricity of the corrected area total meter at each moment in the preset historical period is further corrected to obtain the first electricity data.

[0088] In this embodiment, when the first original data and the second original data are preprocessed, the substation data of the target substation can be obtained first, and the substation data can include a meter household change file, and a household change means that a certain user belongs to substation A before a certain time, but belongs to substation B after the time. Therefore, the ownership relationship between users and target substations in different periods can be clarified through the meter household change file, so as to assist in correcting the total meter power of the substation total meter in the target substation. Of course, the substation data of the present application can also include the substation belonging area name, substation belonging line name, substation name, substation number, meter measurement point number, etc., which are not limited here.

[0089] After obtaining the station area data, the statistical discrimination method can be used to eliminate the abnormal data in the station area data, the first original data and the second original data. There are many statistical discrimination methods. The basic method is to give a confidence level (such as 95%, 99%), and then find the corresponding confidence interval, and then determine the data outside the confidence interval as an abnormal value and eliminate it. The present application can eliminate the abnormal data in the station area data, the first original data and the second original data respectively by setting the confidence interval.

[0090] After the abnormal data is removed, the present application can also fill in the missing data in the area data after the abnormal data is removed, the first original data, and the second original data. There are many methods for filling in missing data, which can specifically include: filling in fixed values, filling in means, filling in modes, filling in medians, interpolation filling, filling in adjacent values ​​before or after the missing position, and establishing a regression model with non-missing values ​​as features to obtain missing values. The present application can choose any missing data filling method to fill in the missing data in the area data after the abnormal data is removed, the first original data, and the second original data.

[0091] Next, since the first original data in the present application may include not only the clock deviation of the total meter, but also the total meter power of the total meter of the substation at each moment in the preset historical period, the second original data may include not only the clock deviation of the sub-meter, but also the sub-meter power of each substation at each moment in the preset historical period. Therefore, the present application may also use the clock deviation of the total meter and the clock deviation of the sub-meter of each substation to correct the total meter power of the substation at each moment in the preset historical period and the sub-meter power of each substation at each moment in the preset historical period after filling in the missing data, and obtain the second power data.

[0092] For example, the present application can compare the clock deviation of the total meter and the clock deviation of a certain sub-meter with the standard time. When the comparison result is that the total meter of the substation is 2 minutes slow and the sub-meter of the substation is 5 minutes fast, it is necessary to re-align the total meter power of the substation at each moment in the preset historical period and the sub-meter power of the substation at each moment in the preset historical period. Otherwise, the total meter data and the sub-meter data at the same time cannot correspond one to one.

[0093] Furthermore, since there are multiple users in a substation, and there is only one substation total meter in a substation, each user corresponds to a substation table. When a user has a household change, that is, a user belongs to substation A before a certain moment, but belongs to substation B after that moment, then the user's sub-meter electricity should be counted into the total meter electricity of substation A before that moment, and should be counted into the total meter electricity of substation B after that moment. If the household change relationship is not considered, a large error will be introduced. Therefore, the present application can continue to correct the total meter electricity of the corrected substation total meter at each moment in the preset historical period according to the meter household change file in the substation data after filling in the missing data, so as to obtain the final first electricity data.

[0094] In one embodiment, in S113, based on the second power data of each substation table, the substation tables in which the power of each submeter is not greater than the preset power threshold are eliminated, and the second power data of each submeter table in which the power of each submeter is greater than the preset power threshold in the preset historical period is obtained, which may include:

[0095] S1131: Determine the sub-meter power of each sub-meter at each moment in the preset historical period according to the second power data of each sub-meter in the preset historical period.

[0096] S1132: Take 1 / K of the median of the sub-meter electricity consumption of each sub-meter at each moment in the preset historical period as the low electricity consumption threshold at that moment, and at each moment, assign a judgment value of 1 to the sub-meter electricity consumption in each sub-meter of the sub-meter less than the low electricity consumption threshold at that moment, and assign a judgment value of 0 to the other sub-meters.

[0097] S1133: Calculate the average value of the sum of the judgment values ​​of each sub-meter at all times in the preset historical period, and remove the sub-meters whose average values ​​in the sub-meters are greater than the preset probability threshold as sub-meters whose power is not greater than the preset power threshold, to obtain the second power data of the sub-meters whose power is greater than the preset power threshold in the preset historical period.

[0098] In this embodiment, when eliminating the sub-meter electric quantity of each sub-meter in each sub-meter table that is not greater than the preset electric quantity threshold value, the present application can first determine the sub-meter electric quantity of each sub-meter at each moment in the preset historical period according to the second electric quantity data of each sub-meter in the preset historical period, and then the present application can use 1 / K of the median of the sub-meter electric quantity of all sub-meters at each moment in the preset historical period as the small electric quantity threshold value at that moment, and at each moment, assign a judgment value of 1 to the sub-meter in each sub-meter table whose electric quantity is less than the small electric quantity threshold value at that moment, and assign a judgment value of 0 to the other sub-meters, and finally calculate the average value of the sum of the judgment values ​​of each sub-meter at all moments in the preset historical period, and eliminate the sub-meter tables whose average value in each sub-meter table is greater than the preset probability threshold as the sub-meter tables not greater than the preset electric quantity threshold value, thereby obtaining the second electric quantity data of each sub-meter in the preset historical period for each sub-meter whose electric quantity is greater than the preset electric quantity threshold value.

[0099] It can be understood that after assigning a judgment value of 1 or 0 to the substation table at each moment, the present application can make the average value of the sum of the judgment values ​​of each substation table at all moments in the preset historical period obtained by subsequent calculation be in the range of [0, 1], which is physically equivalent to the probability of the occurrence of a small electricity sub-meter. The greater the probability of the occurrence of a small electricity sub-meter, the greater the deviation of the calculation result. Therefore, the value range of the preset probability threshold in the present application can be set to [0.05, 0.1] to control the probability of the occurrence of a small electricity sub-meter. Furthermore, when determining the small electricity threshold, the present application can also set the value range of K according to the above analysis, thereby effectively improving the recognition accuracy of out-of-tolerance meters.

[0100] In one embodiment, in S120, determining the relative error of each sub-meter at each moment in the preset historical period and the confidence interval of the relative error of each sub-meter according to the first power data and the second power data corresponding to each sub-meter, may include:

[0101] S121: Use a sliding window of preset length to divide the first electricity data of the area total meter in the preset historical period and the second electricity data of each area meter in the preset historical period into time periods to obtain the total meter electricity and sub-meter electricity in each time period.

[0102] S122: Based on the law of conservation of energy, a group of equations for solving sub-meter relative errors corresponding to the total meter power and sub-meter power in multiple time periods are established, and after solving the group of equations for solving the sub-meter relative errors, point estimates of the quantities to be determined are obtained, and the quantities to be determined include the sub-meter relative errors, line loss rate, and fixed loss.

[0103] S123: Determine the relative error of each table for each area according to the point estimate of the quantity to be determined, and calculate the confidence interval corresponding to the relative error of each table for each area by using the T test in the significance test.

[0104] In this embodiment, after obtaining the first electricity data of the total meter in the target substation in the preset historical period and the second electricity data of the substation meters whose electricity is greater than the preset electricity threshold in the preset historical period, the present application can calculate the sub-meter relative error and the confidence interval of the sub-meter relative error of each substation meter at each moment in the preset historical period based on the first electricity data and the second electricity data corresponding to each substation meter, and judge the probability of the meter being out of tolerance through the confidence interval of the sub-meter relative error, thereby obtaining the suspected meter.

[0105] Specifically, when calculating the relative error of each sub-meter at each moment in a preset historical period for each sub-meter, the present application may first set the sliding window size to D (assuming D ≥ (J+2)), and then use the sliding window to divide the first electricity data of the total meter in the preset historical period and the second electricity data of each sub-meter in the preset historical period into time periods, thereby obtaining the total meter electricity and sub-meter electricity in each time period. Next, the present application may establish a sub-meter relative error solution equation group corresponding to the total meter electricity and sub-meter electricity in multiple time periods based on the law of conservation of energy, and after solving the sub-meter relative error solution equation group, obtain the point estimate of the quantity to be determined, which includes the sub-meter relative error, line loss rate and fixed loss. Finally, the present application may determine the sub-meter relative error of each sub-meter according to the point estimate of the quantity to be determined, and use the T test in the significance test to calculate the confidence interval corresponding to the sub-meter relative error of each sub-meter.

[0106] Furthermore, based on the law of conservation of energy, the relative error solution equation of the sub-table is established as follows:

[0107]

[0108] Where J is the number of meters in the substation; y(i) is the total meter power of the substation at the measurement time i; x j (i) is the electricity consumption of the jth sub-meter at the measurement time i; e j (i) is the relative error of the j-th sub-meter at the measurement time i; e line (i) is the line loss rate at the measurement time i; e line(i)y(i) is the line loss at metering time i, which is proportional to the total meter power of the substation; e0(i) is the fixed loss at metering time i.

[0109] Since the first electricity data and the second electricity data in the present application are divided into D time periods, the present application can construct a group of equations for solving the relative errors of the sub-meters at D moments through the above-mentioned equations for solving the relative errors of the sub-meters. In the solving equations, the total meter electricity of the substation area and the meter electricity of the substation area are known quantities, and the relative errors of the sub-meters, the line loss rate and the fixed loss are quantities to be determined. Therefore, the matrix representation of the group of multivariate linear regression equations is:

[0110] Y=XB (2)

[0111] Where Y is the total power vector of the area meter, Y = [y(1), y(2), ..., y(D)] T ; X is the electricity matrix of the substation, X=[x(1),x(2),...,x(D)] T , where x(D)=[x1(D),x2(D),...,x J (D), y(D), 1]; B is the matrix of the quantity to be determined, B = [B(1), B(2), ..., B(D)], B(D) = [1-e1(D), 1-e2(D), ..., 1-e J (D),e line (D),e0(D)] T .

[0112] According to the matrix transformation, the point estimate of the unknown quantity B is:

[0113]

[0114] Furthermore, the present application can use the T test in the significance test to calculate the confidence interval corresponding to the relative error of each table in each table. Specifically, from formula (3), it can be seen that the point estimate of the quantity B to be determined is Affected by the measured data of total meter power and meter power in the substation area, due to the systematic error and random error in the measured data, the point estimate of the quantity B to be determined is There are also errors. The confidence interval centered on the point estimate can contain the point estimate with a certain degree of confidence. Therefore, this application can use the confidence interval to measure the point estimate. Specifically, from the T test in the significance test, we can know that under the confidence level of (1-α) The confidence interval for is:

[0115]

[0116] Among them, t can be queried in the T test critical value table through the confidence level (1-α) and the degree of freedom (D-1). α / 2 The value of s D The result of repeated calculations The standard deviation of .

[0117] Then, according to equations (2) and (4), the confidence interval of the relative error of the easy-to-score table (θL, θU) is:

[0118]

[0119] In one embodiment, in S130, based on the confidence interval of the relative error of each table at each moment in the preset historical period, calculating the average suspicion coefficient of each table in the preset historical period may include:

[0120] S131: Calculate the suspicion coefficient corresponding to the confidence interval of the relative error of each partition table at each moment in the preset historical period by using the credible range confirmation mechanism.

[0121] S132: Calculate the average suspicion coefficient of each classification table in the preset historical period according to the suspicion coefficient of each classification table at each moment in the preset historical period.

[0122] In this embodiment, since the confidence interval obtained in this application is the confidence interval corresponding to the relative error of each station area table at each moment in the preset historical period, when calculating the average suspicion coefficient of each station area table in the preset historical period, the suspicion coefficient of each station area table at each moment in the preset historical period can be calculated first, and then the final average suspicion coefficient can be obtained based on the suspicion coefficient at each moment.

[0123] When calculating the suspicion coefficient of each station table at each moment in the preset historical period, the credible range (θ L0 ,θ U0 ), and then calculate the corresponding suspicion coefficient. The specific calculation formula is:

[0124]

[0125] The smaller the credible range is set in this application, the higher the on-site precision of the suspect sub-table is. Of course, it can also be set according to on-site experience, and the credible range is generally set to (-2, 2). After obtaining the suspicion coefficient of each sub-table at each moment in the preset historical period, the sum of the suspicion coefficients of the corresponding sub-table can be divided by the total measurement time to obtain the average suspicion coefficient.

[0126] In one embodiment, in S140, according to the line loss electricity and sub-meter electricity of each differentiation table at each moment in the preset historical period, the correlation between the line loss electricity and the sub-meter electricity of each differentiation table in the preset historical period is calculated, and the average suspicion coefficient correction factor of each differentiation table in the preset historical period is determined by using the correlation influence mechanism, which may include:

[0127] S141: Calculate the correlation between the line loss electricity and the sub-meter electricity of each differentiated meter in the preset historical period according to the line loss electricity and the sub-meter electricity of each differentiated meter at each moment in the preset historical period, wherein the correlation includes negative correlation and positive correlation.

[0128] S142: For each station differentiation table, when the line loss electricity and the sub-meter electricity in the preset historical period are negatively correlated, the average suspicion coefficient correction factor of the station differentiation table in the preset historical period is set to a value greater than 1.

[0129] S143: When there is a positive correlation between the line loss electricity and the sub-meter electricity of the station division meter in the preset historical period, the average suspicion coefficient correction factor of the station division meter in the preset historical period is set to a value less than 1.

[0130] In the present application, in order to avoid the situation where the calculation result of the relative error of the sub-meter is inaccurate due to the influence of the correlation between the line loss electricity and the sub-meter electricity when solving the relative error of the sub-meters of each sub-meter, the present application can calculate the correlation between the line loss electricity and the sub-meter electricity of each sub-meter in the preset historical period according to the line loss electricity and the sub-meter electricity of each sub-meter at each moment in the preset historical period. Then, the present application can use the correlation influence mechanism to determine the average suspicion coefficient correction factor of each sub-meter in the preset historical period. In this way, the average suspicion coefficient can be corrected by the average suspicion coefficient correction factor, so as to improve the recognition accuracy of out-of-tolerance meters.

[0131] For example, when the line loss electricity of a station-specific meter is negatively correlated with the electricity of a sub-meter, this situation is consistent with the performance of electricity theft. Therefore, the average suspicion coefficient of the station-specific meter should be increased. At this time, the average suspicion coefficient correction factor of the station-specific meter in the preset historical period can be set to a value greater than 1; and when the line loss electricity of a station-specific meter is negatively correlated with the sub-meter electricity, when the sub-meter relative error of the station-specific meter is estimated using the multivariate linear regression method or its improved algorithm, the sub-meter electricity and the line loss electricity have the same mathematical status. Therefore, the error will be shared between the relative error of the sub-meter and the line loss rate, resulting in deviations in the calculation results of the two. At this time, the average suspicion coefficient of the station-specific meter should be reduced, such as setting the average suspicion coefficient correction factor of the station-specific meter in the preset historical period to a value less than 1.

[0132] Based on the above considerations, the calculation formula for the average suspicion coefficient correction factor λ in this application is as follows:

[0133]

[0134] Among them, c is the average suspicion coefficient of the sub-meter table of the station area, a1 and a2 are correction coefficients, 0 < a1 < 1, 1 < a2 < 1.5, and in the example, a1 = 0.5 and a2 = 1.2 are taken.

[0135] The on-line detection device for out-of-tolerance electric meters considering the suspicion coefficient provided in the embodiments of this application will be described below. The on-line detection device for out-of-tolerance electric meters considering the suspicion coefficient described below can be correspondingly referred to the on-line detection method for out-of-tolerance electric meters considering the suspicion coefficient described above.

[0136] In one embodiment, as Figure 2 shown, Figure 2 is a schematic structural diagram of an on-line detection device for out-of-tolerance electric meters considering the suspicion coefficient provided in the embodiments of this application; this application also provides an on-line detection device for out-of-tolerance electric meters considering the suspicion coefficient, which may include a data acquisition module 210, an error determination module 220, a suspicion coefficient calculation module 230, a correction factor determination module 240, and an out-of-tolerance electric meter detection module 250, specifically including the following:

[0137] The data acquisition module 210 is configured to determine the first power consumption data of the main meter of the target station area in a preset historical period and the second power consumption data of the sub-meter tables of the station area where the power consumption of each sub-meter is greater than the preset power threshold in the preset historical period.

[0138] The error determination module 220 is configured to determine the line loss power, sub-meter power, sub-meter relative error, and the confidence interval of the sub-meter relative error at each moment in the preset historical period according to the first power consumption data and the second power consumption data corresponding to each sub-meter table of the station area.

[0139] The suspicion coefficient calculation module 230 is configured to calculate the average suspicion coefficient of each sub-meter table in the preset historical period based on the confidence interval of the sub-meter relative error at each moment in the preset historical period of each sub-meter table of the station area.

[0140] The correction factor determination module 240 is configured to calculate the correlation between the line loss power and the sub-meter power of each sub-meter table in the preset historical period according to the line loss power and the sub-meter power of each sub-meter table at each moment in the preset historical period, and use the correlation influence mechanism to determine the average suspicion coefficient correction factor of each sub-meter table in the preset historical period.

[0141] The out-of-tolerance meter detection module 250 is used to correct the average suspicion coefficient of the corresponding station division table using the average suspicion coefficient correction factor of each station division table, and sort the station division tables in descending order according to the corrected average suspicion coefficients, and take the station division table with the largest ranking as the out-of-tolerance meter.

[0142] In the above embodiment, when the smart meter in the target area is detected, the first electricity data of the area total meter in the target area in the preset historical period and the second electricity data of the area meter whose electricity of each sub-meter is greater than the preset electricity threshold in the preset historical period can be determined first, so as to avoid the influence of the small electricity area meter on the final result; then, the present application can determine the line loss electricity, sub-meter electricity, sub-meter relative error and confidence interval of the sub-meter relative error of each area meter at each moment in the preset historical period according to the first electricity data and the second electricity data corresponding to each area meter, and calculate the average suspicion coefficient of each area meter in the preset historical period through the confidence interval, so as to analyze the comprehensive influence of the relative error of the sub-meter at different moments, and obtain the suspected electricity meter through a series of measures from the probability perspective, so that the detection result is more objective, the on-site false detection rate is lower, and the detection rate is higher; secondly, the present application takes into account that when the line loss electricity and the sub-meter electricity are strongly correlated, the shortage of line loss electricity and the shortage of sub-meter electricity will be transferred to each other, and the confidence interval of the sub-meter relative error will be transferred to each other. The relative error of the sub-meter obtained in this way is inaccurate. Based on this, the present application calculates the correlation between the line loss electricity and the sub-meter electricity of each differentiation meter in the preset historical period according to the line loss electricity and the sub-meter electricity of each differentiation meter at each moment in the preset historical period, and uses the correlation influence mechanism to determine the average suspicion coefficient correction factor of each differentiation meter in the preset historical period, and uses the average suspicion coefficient correction factor to correct the average suspicion coefficient of the corresponding differentiation meter, so as to eliminate the above-mentioned influencing factors from the perspective of probability, thereby improving the accuracy of identifying out-of-tolerance meters; further, after obtaining the corrected average suspicion coefficient of each differentiation table, the present application can sort each differentiation table in descending order according to the corrected average suspicion coefficient, and take the differentiation table with the largest ranking as the out-of-tolerance meter, so that the out-of-tolerance meter detected in this way has higher accuracy, and the process does not require manual detection, thereby effectively improving detection efficiency and reducing maintenance costs, and meeting the requirements of digital transformation of smart grids and daily operation and maintenance needs to a great extent.

[0143] In one embodiment, the present application also provides a storage medium storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the online detection method of out-of-tolerance electric meters considering the suspicion coefficient as described in any of the above embodiments.

[0144] In one embodiment, the present application also provides a computer device, including: one or more processors, and a memory.

[0145] The memory stores computer-readable instructions, and when the computer-readable instructions are executed by the one or more processors, the steps of the online detection method for out-of-tolerance electric meters taking into account the suspicion coefficient as described in any one of the above embodiments are performed.

[0146] Indicatively, Figure 3 As shown, Figure 3 This is a schematic diagram of the internal structure of a computer device provided in an embodiment of the present application. The computer device 300 may be provided as a server. Figure 3 The computer device 300 includes a processing component 302, which further includes one or more processors, and a memory resource represented by a memory 301, for storing instructions executable by the processing component 302, such as an application. The application stored in the memory 301 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 302 is configured to execute instructions to perform the online detection method of an out-of-tolerance electric meter considering a suspicion coefficient of any of the above embodiments.

[0147] The computer device 300 may further include a power supply component 303 configured to perform power management of the computer device 300, a wired or wireless network interface 304 configured to connect the computer device 300 to a network, and an input / output (I / O) interface 305. The computer device 300 may operate based on an operating system stored in the memory 301, such as Windows Server TM, Mac OS X TM, Unix TM, Linux TM, Free BSD TM, or the like.

[0148] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0149] Finally, it should be noted that, in this article, relational terms such as first and second, etc. 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.

[0150] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can refer to each other.

[0151] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may 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 scope consistent with the principles and novel features disclosed herein.

Claims

1. An online detection method for out-of-tolerance electric meters considering suspicion coefficient, characterized in that: The method comprises: Determine the first power data of the total meter of the target area in the preset historical period and the second power data of the sub-meters whose power is greater than the preset power threshold in the preset historical period; According to the first electricity quantity data and the second electricity quantity data corresponding to each substation table, determine the line loss electricity quantity, sub-meter electricity quantity, sub-meter relative error and the confidence interval of the sub-meter relative error of each substation table at each moment in the preset historical period; Based on the confidence interval of the relative error of each differentiation table at each moment in the preset historical period, the average suspicion coefficient of each differentiation table in the preset historical period is calculated; According to the line loss electricity and sub-meter electricity of each differentiated meter at each moment in the preset historical period, the correlation between the line loss electricity and sub-meter electricity of each differentiated meter in the preset historical period is calculated, and the average suspicion coefficient correction factor of each differentiated meter in the preset historical period is determined by using the correlation influence mechanism; The average suspicion coefficient correction factor of each station division table is used to correct the average suspicion coefficient of the corresponding station division table, and the station division tables are sorted in descending order according to the corrected average suspicion coefficient, and the station division table with the largest ranking is taken as the out-of-tolerance meter; The determining, based on the first power data and the second power data corresponding to each substation table, the sub-meter relative error of each substation table at each moment in the preset historical period and the confidence interval of the sub-meter relative error includes: Using a sliding window of preset length, the first electricity data of the total meter of the substation area in the preset historical period and the second electricity data of the substation area table in the preset historical period are divided into time periods to obtain the total meter electricity and the sub-meter electricity in each time period; Based on the law of conservation of energy, a group of equations for solving relative errors of sub-meters corresponding to the total meter electricity and sub-meter electricity in multiple time periods is established, and after solving the group of equations for solving relative errors of sub-meters, a point estimate of the desired quantity is obtained, wherein the desired quantity includes the relative error of the sub-meter, the line loss rate, and the fixed loss; Determine the relative error of each table for each area according to the point estimate of the quantity to be determined, and calculate the confidence interval corresponding to the relative error of each table for each area using the T test in the significance test; The calculation of the average suspicion coefficient of each differentiation table in the preset historical period based on the confidence interval of the relative error of each differentiation table at each moment in the preset historical period includes: The suspicion coefficient corresponding to the confidence interval of the relative error of each partition meter at each moment in the preset historical period is calculated by using a credible range confirmation mechanism; According to the suspicion coefficient of each classification table at each moment in the preset historical period, the average suspicion coefficient of each classification table in the preset historical period is calculated.

2. The method for online detection of out-of-tolerance electric meters considering the suspicion coefficient according to claim 1 is characterized in that: The step of determining the first power data of the total meter of the target area in the preset historical period and the second power data of the sub-meters whose power is greater than the preset power threshold in the preset historical period includes: Determine the first original data of the total table of the target area in the preset historical period and the second original data of each area table in the preset historical period; Performing a preprocessing operation on the first original data and the second original data to obtain first electric quantity data and second electric quantity data, wherein the preprocessing operation includes removing abnormal data, filling missing data, correcting clock deviation data, and correcting user variable data using user variable files; According to the second electricity data of each substation table, the substation tables whose sub-meter electricity is not greater than the preset electricity threshold in the substation tables are eliminated to obtain the second electricity data of each sub-meter whose sub-meter electricity is greater than the preset electricity threshold in the preset historical period.

3. The method for online detection of out-of-tolerance electric meters considering the suspicion coefficient according to claim 2 is characterized in that: The preprocessing operation is performed on the first original data and the second original data to obtain the first electric quantity data and the second electric quantity data, including: Acquire the area data of the target area; wherein the area data includes the household change file of the electric meter, the first original data includes the clock deviation of the total meter and the total meter power of the area total meter at each moment in the preset historical period, and the second original data includes the clock deviation of the sub-meter and the sub-meter power of each sub-meter at each moment in the preset historical period; Using a statistical discrimination method to eliminate abnormal data in the station area data, the first original data, and the second original data; Filling the missing data in the area data after the abnormal data is removed, the first original data, and the second original data; Using the clock deviation of the total meter and the clock deviation of the sub-meters of each substation, the total meter power of the substation total meter at each moment in the preset historical period and the sub-meter power of each substation substation at each moment in the preset historical period after filling in the missing data are corrected to obtain second power data; According to the meter household change files in the area data after the missing data is filled, the total meter electricity of the corrected area total meter at each moment in the preset historical period is further corrected to obtain the first electricity data.

4. The method for online detection of out-of-tolerance electric meters considering the suspicion coefficient according to claim 2 is characterized in that: According to the second power data of each substation table, the substation tables whose sub-meter power is not greater than the preset power threshold are eliminated, and the second power data of each sub-meter table whose sub-meter power is greater than the preset power threshold in the preset historical period are obtained, including: Determine the sub-metered electric quantity of each sub-meter at each moment in the preset historical period according to the second electric quantity data of each sub-meter in the preset historical period; The median 1 / K of the sub-meter electricity of each sub-meter in each sub-meter in the preset historical period is used as the low electricity threshold at that moment, and at each moment, the sub-meter in each sub-meter whose electricity is less than the low electricity threshold at that moment is assigned a judgment value of 1, and the other sub-meters are assigned a judgment value of 0; The average value of the sum of the judgment values ​​of each sub-meter at all times in the preset historical period is calculated, and the sub-meters with average values ​​greater than the preset probability threshold in the sub-meters are eliminated as sub-meters with power not greater than the preset power threshold, so as to obtain the second power data of the sub-meters with power greater than the preset power threshold in the preset historical period.

5. The method for online detection of out-of-tolerance electric meters considering suspicion coefficient according to claim 1, characterized in that: The method of calculating the correlation between the line loss electricity and the sub-meter electricity of each differentiated meter in the preset historical period according to the line loss electricity and the sub-meter electricity of each differentiated meter at each moment in the preset historical period, and determining the average suspicion coefficient correction factor of each differentiated meter in the preset historical period by using the correlation influence mechanism includes: According to the line loss electricity and sub-meter electricity of each differentiated meter at each moment in the preset historical period, the correlation between the line loss electricity and the sub-meter electricity of each differentiated meter in the preset historical period is calculated, and the correlation includes negative correlation and positive correlation; For each station differentiation table, when the line loss electricity and the sub-meter electricity of the station differentiation table in the preset historical period are negatively correlated, the average suspicion coefficient correction factor of the station differentiation table in the preset historical period is set to a value greater than 1; When there is a positive correlation between the line loss electricity and the sub-meter electricity of the station differentiation meter in the preset historical period, the average suspicion coefficient correction factor of the station differentiation meter in the preset historical period is set to a value less than 1.

6. An online detection device for out-of-tolerance electric meter considering suspicion coefficient, characterized in that: include: A data acquisition module, used to determine the first power data of the total meter of the target area in the preset historical period and the second power data of the sub-meters whose power is greater than the preset power threshold in the preset historical period; An error determination module, configured to determine the line loss electricity, sub-meter electricity, sub-meter relative error and confidence interval of the sub-meter relative error of each sub-meter at each moment in the preset historical period according to the first electricity data and the second electricity data corresponding to each sub-meter; A suspicion coefficient calculation module, used to calculate the average suspicion coefficient of each differentiation table in the preset historical period based on the confidence interval of the relative error of each differentiation table at each moment in the preset historical period; A correction factor determination module is used to calculate the correlation between the line loss electricity and the sub-meter electricity of each differentiated meter in the preset historical period according to the line loss electricity and the sub-meter electricity of each differentiated meter at each moment in the preset historical period, and determine the average suspicion coefficient correction factor of each differentiated meter in the preset historical period by using the correlation influence mechanism; The out-of-tolerance meter detection module is used to correct the average suspicion coefficient of the corresponding station division table by using the average suspicion coefficient correction factor of each station division table, and sort the station division tables in descending order according to the corrected average suspicion coefficient, and take the station division table with the largest ranking as the out-of-tolerance meter; The error determination module determines the relative error of each sub-meter at each moment in the preset historical period and the confidence interval of the relative error of each sub-meter according to the first power data and the second power data corresponding to each sub-meter, including: Using a sliding window of preset length, the first electricity data of the total meter of the substation area in the preset historical period and the second electricity data of the substation area table in the preset historical period are divided into time periods to obtain the total meter electricity and the sub-meter electricity in each time period; Based on the law of conservation of energy, a group of equations for solving relative errors of sub-meters corresponding to the total meter electricity and sub-meter electricity in multiple time periods is established, and after solving the group of equations for solving relative errors of sub-meters, a point estimate of the desired quantity is obtained, wherein the desired quantity includes the relative error of the sub-meter, the line loss rate, and the fixed loss; Determine the relative error of each table for each area according to the point estimate of the quantity to be determined, and calculate the confidence interval corresponding to the relative error of each table for each area using the T test in the significance test; The suspicion coefficient calculation module calculates the average suspicion coefficient of each differentiation table in the preset historical period based on the confidence interval of the relative error of each differentiation table at each moment in the preset historical period, including: The suspicion coefficient corresponding to the confidence interval of the relative error of each partition meter at each moment in the preset historical period is calculated by using a credible range confirmation mechanism; According to the suspicion coefficient of each classification table at each moment in the preset historical period, the average suspicion coefficient of each classification table in the preset historical period is calculated.

7. A storage medium, characterized in that: The storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the online detection method for out-of-tolerance electric meters considering the suspicion coefficient as described in any one of claims 1 to 5.

8. A computer device, characterized in that: include: one or more processors, and memory; The memory stores computer-readable instructions, and when the computer-readable instructions are executed by the one or more processors, the steps of the online detection method for out-of-tolerance electric meters taking into account the suspicion coefficient as described in any one of claims 1 to 5 are performed.

Images