A method, device, and storage medium for identifying inaccurate measurement of low-voltage meters

Abstract

The present invention discloses a method and device for identifying inaccurate measurement of low-voltage meters, and a storage medium. The method comprises the following steps: determining the first mutation date of the line loss rate of a target area and the abnormal area in the target area according to the line loss rate data; determining the second mutation date of the power consumption data of each user in the abnormal area according to the power consumption data; determining the third mutation date of the line loss value of the target area according to the line loss value data; if the first mutation date and the third mutation date are both the same as the second mutation date of the power consumption data of a certain user, the user is determined to be an abnormal user; calculating the similarity between the line loss mutation value corresponding to the abnormal user and the power consumption mutation value of the abnormal user, if the similarity corresponding to a certain abnormal user is greater than the similarity threshold, the abnormal user is determined to be the final abnormal user. The present invention is targeted at a wide range and can cover the power consumption characteristics of most users who continuously steal electricity.

Description

Technical Field

[0001] The present invention relates to the technical field of electric energy metering and acquisition, and particularly relates to a method and device for identifying inaccurate metering of low-voltage meters, and a storage medium. Background Art

[0002] Traditional methods for identifying abnormal metering of low-voltage meters mainly rely on analyzing the data of unbalanced zero-fire currents. However, many electricity theft methods can make the zero-fire current data entered into the database equal or the zero-fire current data empty (or 0), resulting in the failure of the analysis method and thus low identification efficiency. Summary of the Invention

[0003] In view of the above technical problems, the purpose of the present invention is to provide a method and device for identifying inaccurate metering of low-voltage meters, and a storage medium, so as to solve the problem of the failure of the analysis method caused by the method for identifying abnormal metering of low-voltage meters.

[0004] The present invention adopts the following technical solutions:

[0005] A method for identifying inaccurate metering of low-voltage meters includes the following steps:

[0006] Obtain the line loss rate data of the target substation area within a preset time period, and determine the first mutation date of the line loss rate of the target substation area and the abnormal substation areas in the target substation area according to the line loss rate data;

[0007] Obtain the power consumption data of each user in the abnormal substation area within a preset time period, and determine the second mutation date of the power consumption data of each user in the abnormal substation area according to the power consumption data;

[0008] Obtain the line loss value data of the abnormal substation area within a preset time period, and determine the third mutation date of the line loss value of the target substation area according to the line loss value data;

[0009] If both the first mutation date and the third mutation date are the same as the second mutation date of the power consumption data of a certain user, then determine that user as an abnormal user;

[0010] Calculate the similarity between the line loss mutation value corresponding to the abnormal user and the power consumption mutation value of the abnormal user to obtain the similarity of each abnormal user;

[0011] If the similarity corresponding to a certain abnormal user is greater than the similarity threshold, then determine that abnormal user as the final abnormal user.

[0012] Preferably, the determining the first mutation date of the line loss rate of the target substation area according to the line loss rate data includes:

[0013] After obtaining the line loss rate data of the target substation area within a preset time period, preprocess the line loss rate data to obtain a preprocessed line loss rate time series;

[0014] Calculate the first mutation degree and the second mutation degree corresponding to each line loss rate in the time series of the line loss rate; wherein, the first mutation degree corresponding to each line loss rate is calculated according to the first left sequence and the first right sequence corresponding to each line loss rate, and the second mutation degree corresponding to each line loss rate is calculated according to the second left sequence and the second right sequence corresponding to each line loss rate; divide the time series of the line loss rate based on a certain line loss rate to obtain the first left sequence and the first right sequence corresponding to the line loss rate; obtain the second left sequence based on the first left sequence, and obtain the second right sequence based on the first right sequence;

[0015] Calculate the first mutation index corresponding to each line loss rate in the time series of the line loss rate, and the first mutation index is the product of the first mutation degree and the second mutation degree; determine the first mutation date of the line loss rate of the target substation area according to the first mutation indexes corresponding to all line loss rates.

[0016] Preferably, the determination method of the abnormal substation area in the target substation area is as follows:

[0017] If there is a first mutation index corresponding to a line loss rate in the time series of the line loss rate of the target substation area that is greater than a preset threshold, it is determined that the target substation area is an abnormal substation area.

[0018] Preferably, the determining the first mutation date of the line loss rate of the target substation area according to the first mutation indexes corresponding to all line loss rates includes:

[0019] Obtain the maximum value among all the first mutation indexes greater than the preset threshold, and use the time date corresponding to the maximum value as the first mutation date of the line loss rate of the target substation area.

[0020] Preferably, the determining the second mutation date of the power consumption data of each user in the abnormal substation area according to the power consumption data includes:

[0021] After obtaining the power consumption data of each user in the preset time period of the abnormal substation area, preprocess the power consumption data to obtain a preprocessed time series of the power consumption.

[0022] Calculate the third mutation degree and the fourth mutation degree corresponding to each power consumption data in the time series of the power consumption; wherein, the third mutation degree corresponding to each power consumption data is calculated according to the third left sequence and the third right sequence corresponding to each power consumption data, and the fourth mutation degree corresponding to each power consumption data is calculated according to the fourth left sequence and the fourth right sequence corresponding to each power consumption data; divide the time series of the power consumption based on a certain power consumption data to obtain the third left sequence and the third right sequence corresponding to the power consumption data, obtain the fourth left sequence based on the third left sequence, and obtain the fourth right sequence based on the third right sequence;

[0023] Calculate the second mutation index corresponding to each power consumption data in the time series of power consumption, where the second mutation index is the product of the third mutation degree and the fourth mutation degree; determine the second mutation date of the abnormal substation area power consumption data based on the second mutation indexes corresponding to all power consumption data.

[0024] Preferably, the determining the third mutation date of the target substation area line loss value according to the line loss value data includes:

[0025] After obtaining the line loss value data of the abnormal substation area in a preset time period, preprocess the line loss value data to obtain a preprocessed time series of line loss values;

[0026] Calculate the fifth mutation degree and the sixth mutation degree corresponding to each line loss rate in the time series of line loss values; wherein, the fifth mutation degree corresponding to each line loss value is calculated according to the fifth left sequence and the fifth right sequence corresponding to each line loss value, and the second mutation degree corresponding to each line loss value is calculated according to the sixth left sequence and the sixth right sequence corresponding to each line loss value; divide the time series of line loss values based on a certain line loss value to obtain the fifth left sequence and the fifth right sequence corresponding to the line loss value; obtain the sixth left sequence based on the fifth left sequence, and obtain the sixth right sequence based on the fifth right sequence;

[0027] Calculate the third mutation index corresponding to each line loss value in the time series of line loss values, where the third mutation index is the product of the fifth mutation degree and the sixth mutation degree; determine the third mutation date of the target substation area line loss value based on the third mutation indexes corresponding to all line loss values.

[0028] Preferably, the similarity between the line loss mutation value corresponding to the abnormal user and the power consumption mutation value of the abnormal user satisfies the following formula:

[0029] loss_sim = min(delta_mean_lqp, delta_mean_kwh) / max(delta_mean_lqp, delta_mean_kwh); where, delta_mean_lpq is the line loss mutation value corresponding to the abnormal user, delta_mean_kqh is the power consumption mutation value of the abnormal user, and loss_sim is the similarity between the line loss mutation value corresponding to the abnormal user and the power consumption mutation value of the abnormal user.

[0030] Preferably, the line loss mutation value corresponding to the abnormal user is obtained based on the seventh left sequence and the seventh right sequence, and the time series of line loss values is divided into the seventh left sequence and the seventh right sequence according to the third mutation date; the mutation value of the power consumption of the abnormal user is obtained based on the eighth left sequence and the eighth right sequence, and the time series of power consumption is divided into the eighth left sequence and the eighth right sequence according to the second mutation date.

[0031] A device for identifying inaccurate metering of low-voltage meters, comprising:

[0032] A first mutation date determination unit, configured to obtain the line loss rate data of a target power distribution area within a preset time period, and determine the first mutation date of the line loss rate of the target power distribution area and the abnormal power distribution areas in the target power distribution area according to the line loss rate data;

[0033] A second mutation date determination unit, configured to obtain the power consumption data of each user in the abnormal power distribution area within a preset time period, and determine the second mutation date of the power consumption data of each user in the abnormal power distribution area according to the power consumption data;

[0034] A third mutation date determination unit, configured to obtain the line loss value data of the abnormal power distribution area within a preset time period, and determine the third mutation date of the line loss value of the target power distribution area according to the line loss value data;

[0035] A first determination unit, configured to determine that a user is an abnormal user when both the first mutation date and the third mutation date are the same as the second mutation date of the power consumption data of a certain user;

[0036] A similarity determination unit, configured to calculate the similarity between the line loss mutation value corresponding to the abnormal user and the power consumption mutation value of the abnormal user, and obtain the similarity of each abnormal user;

[0037] A second determination unit, configured to determine that an abnormal user is a final abnormal user when the similarity corresponding to a certain abnormal user is greater than the similarity threshold.

[0038] A computer storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method for identifying inaccurate metering of low-voltage meters is implemented.

[0039] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0040] The present invention determines an abnormal user when both the first mutation date and the third mutation date are the same as the second mutation date of the power consumption data of a certain user; then calculates the similarity between the line loss mutation value corresponding to the abnormal user and the power consumption mutation value of the abnormal user to obtain the similarity of each abnormal user; when the similarity corresponding to a certain abnormal user is greater than the similarity threshold, the final abnormal user is determined. By analyzing from two dimensions of calculating the mutation degree of the user's power consumption and the line loss rate of the power distribution area and evaluating the impact of the change in the user's power consumption on the line loss of the power distribution area, it is completely independent of the zero-fire current data, and the full coverage of the scenario of inaccurate metering of low-voltage meters is realized.

[0041] Specifically, the mutation conditions of the substation area line loss rate, the substation area line loss value, and the user power consumption are calculated through mutation indexes (i.e., the first mutation index, the second mutation index, and the third mutation index) (i.e., the mutation dates are calculated); this includes first circularly calculating the mutation indexes corresponding to each date of various time series data, then further comparing the mutation indexes between different dates, and then confirming the mutation dates; then, based on the mutation date conditions of the substation area line loss rate, the substation area line loss value, and the user power consumption (for example, identifying abnormal users according to the first mutation date, the second mutation date, and the third mutation date), and then, combining the similarity of the change values between the substation area line loss value and the user power consumption to output the final result of abnormal users. During the identification process, only the power consumption data of the electricity meters and the line loss data between substations are relied on to output the final result of abnormal users, which is applicable to a wide range and can cover the power consumption characteristics of most users with continuous electricity theft; it avoids the problem that when identifying by analyzing the data of the zero-fire current imbalance, the zero-fire current data is equal or the zero-fire current data is empty, resulting in the failure of the analysis method. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 FIG. is a schematic flowchart of a method for identifying inaccurate metering of low-voltage meters provided by an embodiment of the present invention;

[0043] Figure 2 FIG. is a schematic diagram of a device for identifying inaccurate metering of low-voltage meters provided by an embodiment of the present invention;

[0044] Figure 3 FIG. is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] Next, in combination with the drawings and specific embodiments, the present invention will be further described. It should be noted that, on the premise of no conflict, the following described embodiments or technical features can be arbitrarily combined to form new embodiments:

[0046] In this embodiment, the data used mainly includes the electric energy data of the electricity meters and the daily line loss information data of the substation area, where the daily line loss information data includes the daily line loss rate data and the daily line loss value data.

[0047] Specifically, the line loss rate is the percentage of the electric energy lost in the power network (line loss load) to the electric energy supplied to the power network (power supply load); the line loss rate is used to evaluate the economy of the operation of the power system.

[0048] Specifically, line loss rate = (line loss electricity / power supply)*100% = (power supply - power sales) / power supply*100% = (1-power sales / power supply)*100%; or, line loss rate = (purchased electricity - power sales) / purchased electricity×100% (the purchased electricity consists of the unified accounting power plant grid-connected electricity, the purchased electricity from independent power plants, and the net electricity received from other power grids).

[0049] Specifically, the line loss value can be calculated by using the loss factor and the power loss at the maximum load to calculate the line loss value in the time period T.

[0050] Embodiment 1:

[0051] Please refer to Figure 1 As shown, Figure 1 A method for identifying inaccurate measurement of a low-pressure meter according to the present invention is shown, comprising the following steps:

[0052] Step S1: obtaining the line loss rate data of the target area in a preset time period, and determining the first mutation date of the line loss rate of the target area and the abnormal area in the target area according to the line loss rate data;

[0053] Specifically, determining the first mutation date of the target area line loss rate according to the line loss rate data includes:

[0054] Step S11: after acquiring the line loss rate data of the preset time period of the target area, preprocessing the line loss rate data to obtain a preprocessed line loss rate time series;

[0055] In this embodiment, preprocessing can be performed by the following method:

[0056] Specifically, obtain the line loss rate data of the substation in the last N days [N≥30], and eliminate the abnormal data with a line loss rate less than or equal to 0% or greater than or equal to 30%. If the number of remaining data days after screening is still greater than n; (n is set to N*66% by default, and the threshold can be adjusted according to actual needs); then regenerate the line loss rate time series sequence from the filtered data in the original time order as the preprocessed line loss rate time series sequence.

[0057] It should be noted that if the number of days remaining after screening is not greater than n, the area will be regarded as an area with abnormal data and will not be identified in subsequent steps.

[0058] Step S12: Calculate the first mutation degree K1 and the second mutation degree corresponding to each line loss rate in the time series of line loss rates; wherein, the first mutation degree corresponding to each line loss rate is calculated based on the first left sequence and the first right sequence corresponding to each line loss rate, and the second mutation degree corresponding to each line loss rate is calculated based on the second left sequence and the second right sequence corresponding to each line loss rate; divide the time series of line loss rates based on a certain line loss rate to obtain the first left sequence and the first right sequence corresponding to this line loss rate; obtain the second left sequence based on the first left sequence, and obtain the second right sequence based on the first right sequence;

[0059] In specific implementation, dividing the time series of line loss rates based on a certain line loss rate to obtain the first left sequence and the first right sequence corresponding to this line loss rate may include the following steps:

[0060] Obtain the time points corresponding to each line loss rate in the time series of line loss rates, arrange the line loss rates in the time series of line loss rates in chronological order, take the sequence composed of the line loss rates before a certain line loss rate as the first left sequence, and take the sequence composed of a certain line loss rate and the line loss rates after this line loss rate as the first right sequence.

[0061] In specific implementation, obtaining the second left sequence based on the first left sequence and obtaining the second right sequence based on the first right sequence may include the following steps:

[0062] In the first left sequence, take the sequence composed of m line loss rates before a certain line loss rate as the second left sequence; in the first right sequence, take the sequence composed of a certain line loss rate and m - 1 line loss rates after this line loss rate as the second right sequence; wherein, m is a natural number, and m is adjusted according to the number of line loss rates in the time series of line loss rates.

[0063] Specifically, the first mutation degree K1 and the second mutation degree respectively satisfy the following formulas:

[0064] K1 = abs(mean_left - mean_right) / max(std_left, std_right);

[0065] Wherein, abs is the absolute value function of a number, mean_left is the mean of the line loss rates in the first left time series, mean_right is the mean of the line loss rates in the first right time series, std_right is the standard deviation of the line loss rates in the first right time series, and std_left is the standard deviation of the line loss rates in the first left time series;

[0066] K2 = abs(mean_left_edge - mean_right_edge) / max(mean_left_edge, mean_right_edge);

[0067] Among them, abs is the absolute value function of a number, mean_left_edge is the mean of the second left time-series line loss rate, and mean_right_edge is the mean of the second right time-series line loss rate.

[0068] Step S13: Calculate the first mutation index corresponding to each line loss rate in the line loss rate time series, where the first mutation index is the product of the first mutation degree and the second mutation degree; determine the first mutation date of the line loss rate of the target substation area according to the first mutation indexes corresponding to all line loss rates.

[0069] As an embodiment, the determining the first mutation date of the line loss rate of the target substation area according to the first mutation indexes corresponding to all line loss rates includes:

[0070] Step S131: Obtain the maximum value among all the first mutation indexes greater than the preset threshold, and use the time date corresponding to the maximum value as the first mutation date of the line loss rate of the target substation area.

[0071] Specifically, the determination method for the abnormal substation area in the target substation area is as follows:

[0072] If there is a first mutation index corresponding to a line loss rate in the line loss rate time series that is greater than the preset threshold, it is determined that the target substation area is an abnormal substation area.

[0073] Step S2: Obtain the electricity consumption data of each user in the abnormal substation area within a preset time period, and determine the second mutation date of the electricity consumption data of each user in the abnormal substation area according to the electricity consumption data;

[0074] The determining the second mutation date of the electricity consumption data of each user in the abnormal substation area according to the electricity consumption data includes:

[0075] Step S21: After obtaining the electricity consumption data of each user in the abnormal substation area within a preset time period, preprocess the electricity consumption data to obtain a preprocessed electricity consumption time series;

[0076] Step S22: Calculate the third mutation degree and the fourth mutation degree corresponding to each electricity consumption data in the electricity consumption time series; among them, the third mutation degree corresponding to each electricity consumption data is calculated according to the third left sequence and the third right sequence corresponding to each electricity consumption data, and the fourth mutation degree corresponding to each electricity consumption data is calculated according to the fourth left sequence and the fourth right sequence corresponding to each electricity consumption data; based on a certain electricity consumption data, the electricity consumption time series is divided to obtain the third left sequence and the third right sequence corresponding to the electricity consumption data, the fourth left sequence is obtained based on the third left sequence, and the fourth right sequence is obtained based on the third right sequence;

[0077] Step S23: Calculate the second mutation index corresponding to each power consumption data in the power consumption time series. The second mutation index is the product of the third mutation degree and the fourth mutation degree. Determine the second mutation date of the abnormal substation area's power consumption data based on the second mutation indices corresponding to all power consumption data.

[0078] Step S3: Obtain the line loss value data of the abnormal substation area in a preset time period, and determine the third mutation date of the target substation area's line loss value based on the line loss value data.

[0079] The determining the third mutation date of the target substation area's line loss value based on the line loss value data includes:

[0080] Step S31: After obtaining the line loss value data of the abnormal substation area in a preset time period, preprocess the line loss value data to obtain a preprocessed line loss value time series.

[0081] Step S32: Calculate the fifth mutation degree and the sixth mutation degree corresponding to each line loss rate in the line loss value time series. Among them, the fifth mutation degree corresponding to each line loss value is calculated based on the fifth left sequence and the fifth right sequence corresponding to each line loss value, and the second mutation degree corresponding to each line loss value is calculated based on the sixth left sequence and the sixth right sequence corresponding to each line loss value. Divide the line loss value time series based on a certain line loss value to obtain the fifth left sequence and the fifth right sequence corresponding to the line loss value. Obtain the sixth left sequence based on the fifth left sequence, and obtain the sixth right sequence based on the fifth right sequence.

[0082] Step S33: Calculate the third mutation index corresponding to each line loss value in the line loss value time series. The third mutation index is the product of the fifth mutation degree and the sixth mutation degree. Determine the third mutation date of the target substation area's line loss value based on the third mutation indices corresponding to all line loss values.

[0083] Step S4: If both the first mutation date and the third mutation date are the same as the second mutation date of a certain user's power consumption data, then determine that this user is an abnormal user.

[0084] For example, the mutation date of the substation area's line loss rate (i.e., the first mutation date), the mutation date of the substation area's line loss value (the third mutation date), and the mutation date of the user's power consumption (i.e., the second mutation date) are all determined to be the eighth date in the time series. (For example, they are all December 17, 2022. The exclusion due to abnormal data does not affect the actual date corresponding to the data), which is the same day. Therefore, this user becomes an abnormal user.

[0085] Step S5: Calculate the similarity between the line loss mutation value corresponding to the abnormal user and the power consumption mutation value of the abnormal user to obtain the similarity of each abnormal user.

[0086] As a specific implementation manner, the time series of the line loss values is divided into a seventh left sequence and a seventh right sequence according to the third mutation date, and the line loss mutation value corresponding to the abnormal user is obtained based on the seventh left sequence and the seventh right sequence;

[0087] As a specific implementation manner, the mutation value of the electricity consumption is obtained based on an eighth left sequence and an eighth right sequence; the time series of the electricity consumption is divided into an eighth left sequence and an eighth right sequence according to the second mutation date.

[0088] Wherein, the similarity between the line loss mutation value corresponding to the abnormal user and the electricity consumption mutation value of the abnormal user satisfies the following formula:

[0089] loss_sim = min(delta_mean_lqp, delta_mean_kwh) / max(delta_mean_lqp, delta_mean_kwh); wherein, delta_mean_lpq is the line loss mutation value corresponding to the abnormal user, delta_mean_kqh is the electricity consumption mutation value of the abnormal user, and loss_sim is the similarity between the line loss mutation value corresponding to the abnormal user and the electricity consumption mutation value of the abnormal user.

[0090] Step S6: If the similarity corresponding to a certain abnormal user is greater than the similarity threshold, it is determined that the abnormal user is the final abnormal user.

[0091] For example, if the similarity threshold is defined as 80%, and the similarity corresponding to a certain abnormal user is greater than 80%, it is determined that the similarity meets the requirements, and the abnormal user is the final abnormal user.

[0092] In specific implementation, the similarity threshold can be adjusted according to the actual situation.

[0093] In the above implementation process, the mutation conditions of the substation area line loss rate, the substation area line loss value, and the user electricity consumption are calculated through the mutation degree indexes (i.e., the first mutation index, the second mutation index, and the third mutation index); specifically, it includes first circularly calculating the mutation degree indexes corresponding to each date of various time series data, and then further comparing the mutation degree indexes between different dates to further confirm the mutation date.

[0094] Then, the abnormal user results are output by combining the similarity between the line loss value of the substation and the change value of the user's electricity consumption, and the sudden change day of the substation line loss rate, the line loss value of the substation, and the user's electricity consumption (for example, the abnormal user is determined according to the first sudden change date, the second sudden change date, and the third sudden change date). The final abnormal user results can be output only by relying on the electricity consumption data of the electric energy meter and the line loss data between substations. This is aimed at a wide range and can cover the electricity consumption characteristics of most users who continuously steal electricity, avoiding the problem of failure of the analysis method due to the zero-fire current data being equal or the zero-fire current data being empty when identifying the uneven zero-fire current data.

[0095] The method of the present invention is described below with a specific embodiment. It should be noted that, in this specific embodiment, the mutation index is no longer distinguished by the first mutation index, the second mutation index and the third mutation index, but is distinguished by distinguishing the mutation index in the line loss rate, the line loss value and the power consumption. The mutation date is also not distinguished by the first mutation date, the second mutation date and the third mutation date, but is distinguished by distinguishing the mutation date in the line loss rate, the line loss value and the power consumption; similarly, the time series and the mutation degree are distinguished by distinguishing the time series and the mutation degree in the line loss rate, the line loss value and the power consumption;

[0096] 1. Identification of mutation points of line loss rate in the substation area

[0097] 1.1 Line loss rate data preprocessing

[0098] First, obtain the line loss rate data of the substation area in the last N days [N≥30], and remove the abnormal data with a line loss rate less than or equal to 0% or greater than or equal to 30%. If the number of days of remaining data after screening is still greater than n (the default is N*66%, and the threshold can be adjusted according to actual needs), the filtered data is regenerated into a line loss rate time series in the original time order:

[0099] LR=[lr 1 , lr 2 , …, lr k ];(1)

[0100] Among them, lr k Represents the kth data in the line loss rate time series sequence that is regenerated after screening.

[0101] If the number of days remaining after screening is not greater than n, the area will be regarded as an area with abnormal data and no subsequent calculation will be performed.

[0102] For example: the original 30-day antenna loss rate data is [3.1, 2.6, 2.4, 2.4, 2.5, 2.6, 2.9, 3.2, 3.5, 3.6, 3.2, 3.5, 3.5, 3.5, 3.8, 3.9, 3.4, 3.2, 3.0, 3.4, 4.9, 4.0, -17.7, -9.5, -8.6, 4.7, 4.2, 4.9, 4.1, 4.9], then the data of the three days [-17.7, -9.5, -8.6] do not meet the threshold and are removed. , the line loss rate time series regenerated after elimination is: LR = [3.1, 2.6, 2.4, 2.4, 2.5, 2.6, 2.9, 3.2, 3.5, 3.6, 3.2, 3.5, 3.5, 3.5, 3.8, 3.9, 3.4, 3.2, 3.0, 3.4, 4.9, 4.0, 4.7, 4.2, 4.9, 4.1, 4.9], and the remaining 27 days of data meet more than 66% of the total days, so it is not an abnormal data area and can be used for subsequent mutation calculations.

[0103] 1.2 Calculate the mutation degree corresponding to the date

[0104] In the online loss rate time series LR, the positive order m (the default 2m is N*33%, which can be adjusted according to needs) dates and the reverse order m dates do not participate in the corresponding mutation degree calculation (because selecting the beginning or end of the few days as the mutation date will cause the segmented time series to be too short, affecting the calculation result).

[0105] For the remaining dates, we traverse each date and use each date as the assumed mutation date for subsequent calculations. We divide the time series into the left time series (dates less than the mutation date) and the right time series (dates greater than or equal to the mutation date), calculate the mutation degree between the two time series, and keep them for subsequent comparison.

[0106] In this embodiment, each date is used as a mutation date to participate in a calculation, and the index corresponding to each date is judged to determine whether it meets the conditions of being a mutation date.

[0107] Example: After processing, LR = [3.1, 2.6, 2.4, 2.4, 2.5, 2.6, 2.9, 3.2, 3.5, 3.6, 3.2, 3.5, 3.5, 3.5, 3.8, 3.9, 3.4, 3.2, 3.0, 3.4, 4.9, 4.0, 4.7, 4.2, 4.9, 4.1, 4.9]. At this time, the eighth date (corresponding data 3.2) is selected as the mutation date. Then, according to this date, LR is divided into the left time series LR_LEFT = [3.1, 2.6, 2.4, 2.4, 2.5, 2.6, 2.9] and the right time series LR_RIGHT = [3.2, 3.5, 3.6, 3.2, 3.5, 3.5, 3.5, 3.8, 3.9, 3.4, 3.2, 3.0, 3.4, 4.9, 4.0, 4.7, 4.2, 4.9, 4.1, 4.9]. Subsequent index calculations are performed on these two time series.

[0108] 1.2.1 Calculate the mutation degree K1

[0109] Divide the line loss rate time series into the left time series and the right time series according to the mutation date, calculate the mean line loss rate mean_left of the left time series, the standard deviation std_left of the left time series line loss rate, the mean line loss rate mean_right of the right time series, and the standard deviation std_right of the right time series line loss rate. The mutation degree K1 can be expressed as: K1 = abs(mean_left - mean_right) / max(std_left, std_right) (2)

[0110] Example: For the left time series LR_LEFT = [3.1, 2.6, 2.4, 2.4, 2.5, 2.6, 2.9], the mean mean_left is 2.65 and the standard deviation std_left is 0.24. For the right time series LR_RIGHT = [3.2, 3.5, 3.6, 3.2, 3.5, 3.5, 3.5, 3.8, 3.9, 3.4, 3.2, 3.0, 3.4, 4.9, 4.0, 4.7, 4.2, 4.9, 4.1, 4.9], the mean mean_right is 3.82 and the standard deviation std_right is 0.60. Based on this, K1 = (3.82 - 2.65) / 0.60 = 1.95 can be calculated.

[0111] 1.2.2 Calculate the mutation degree K2

[0112] The time series of line loss rate is divided into the left time series and the right time series according to the mutation date. Select the k (default is 5, which can be adjusted according to the overall time length of the data to reflect the data differences before and after the data change) dates closest to the mutation date in the two time series to generate the left edge time series and the right edge time series. Calculate the mean of the left edge time series mean_left_edge and the mean of the right edge time series mean_right_edge. The mutation degree K2 can be expressed as:

[0113] K2 = abs(mean_left_edge - mean_right_edge) / max(mean_left_edge, mean_right_edge) (3)

[0114] Example: LR = [3.1, 2.6, 2.4, 2.4, 2.5, 2.6, 2.9, 3.2, 3.5, 3.6, 3.2, 3.5, 3.5, 3.5, 3.8, 3.9, 3.4, 3.2, 3.0, 3.4, 4.9, 4.0, 4.7, 4.2, 4.9, 4.1, 4.9]. Currently, the assumed mutation date is the eighth date (corresponding data 3.2). The k = 5 dates closest to this in the left and right time series form the left and right edge time series. Then the left edge time series is LR_LEFT_EDGE = [2.4, 2.4, 2.5, 2.6, 2.9], and the corresponding mean of the left edge time series mean_left_edge is 2.57. The right edge time series is LR_RIGHT_EDGE = [3.2, 3.5, 3.6, 3.2, 3.5], and the corresponding mean of the right edge time series mean_right_edge is 3.40. Based on this, K2 = (3.40 - 2.57) / 3.40 = 0.24 can be calculated.

[0115] 1.2.3 Calculate the mutation degree index K

[0116] If K1 and K2 meet the threshold requirements (for the judgment of the line loss rate of the transformer substation area, by default, K1 needs to be greater than 1.4 and K2 needs to be greater than 0.2, which can be adjusted according to the situation), then calculate the mutation degree index K by combining K1 and K2. The mutation index K can be expressed as:

[0117] K = K1 * K2;

[0118] If K1 and K2 do not meet the threshold requirements, then do not calculate the mutation degree index K corresponding to this mutation date.

[0119] Example: For the above example data, it is calculated that K1 = 1.95 > 1.4 and K2 = 0.24 > 0.2, indicating that it is feasible to use the eighth date of this section of data as the mutation date. Therefore, record this date and the corresponding K = K1 * K2 = 1.95 * 0.24 = 0.468.

[0120] 1.2.4 Determine the mutation date

[0121] The mutation day with the largest mutation index K value is selected as the time series mutation point of the line loss value of the substation.

[0122] Example: The same substation may have multiple dates as mutation dates, and the corresponding K1 and K2 can all meet the threshold requirements. In this case, the date with the largest K is selected as the final mutation date of the line loss value of this substation. In the above example, 0.882 is the largest K value among all K values ​​that meet the K1 and K2 threshold conditions, so the corresponding eighth date is the final mutation date of the line loss value of this substation.

[0123] 2. Identification of mutation points of line loss value in the substation area

[0124] 2.1 Line loss value data preprocessing

[0125] First, obtain the line loss value data of the abnormal area in the last N days [N≥30], and remove the abnormal data with line loss values ​​less than 0. If the number of days of remaining data after screening is still greater than n (the default is N*66%, and the threshold can be adjusted according to actual needs), regenerate the line loss value time series sequence from the filtered data in the original time order:

[0126] LPQ=[lpq1, lpq2, …, lpqk]; (4)

[0127] Among them, lpqk represents the kth data in the line loss value time series sequence regenerated after screening.

[0128] If the number of days remaining after screening is not greater than n, the area will be regarded as an area with abnormal data and no subsequent calculation will be performed.

[0129] For example: the original 30-day line loss value data is [65,54,46,43,46,56,62,73,85,86,72,80,77,79,92,93,78,69,64,72,76,93,-182,-196,-190,104,85,108,87,108], then the data of the three days [-182,-196,-190] are removed because they do not meet the threshold. The line loss rate time series regenerated after elimination is: LPQ = [65, 54, 46, 43, 46, 56, 62, 73, 85, 86, 72, 80, 77, 79, 92, 93, 78, 69, 64, 72, 76, 93, 104, 85, 108, 87, 108]. The remaining 27 days of data meet more than 66% of the total days, so it is not an abnormal data area and subsequent mutation calculations can be performed.

[0130] 2.2 Calculate the mutation degree corresponding to the date

[0131] In the online loss value time series LPQ, the first m (by default, 2m is N * 33%, which can be adjusted according to requirements) dates in the forward order and the last m dates in the reverse order do not participate in the corresponding mutation degree calculation (because selecting a few days at the beginning or end as mutation days will result in an overly short time series after segmentation, affecting the calculation results).

[0132] Among the remaining dates, traverse each date. Treat each daily date as a mutation day, divide the time series into a left time series (dates less than the mutation day) and a right time series (dates greater than or equal to the mutation day), calculate the mutation degree between the two time series, and retain it for subsequent comparison.

[0133] Example: After processing, LPQ = [65, 54, 46, 43, 46, 56, 62, 73, 85, 86, 72, 80, 77, 79, 92, 93, 78, 69, 64, 72, 76, 93, 104, 85, 108, 87, 108]. At this time, the eighth date (corresponding data 73) is selected as the mutation day. Then, according to this date, LPQ is divided into the left time series LPQ_LEFT = [65, 54, 46, 43, 46, 56, 62] and the right time series LPQ_RIGHT = [73, 85, 86, 72, 80, 77, 79, 92, 93, 78, 69, 64, 72, 76, 93, 104, 85, 108, 87, 108]. Subsequent index calculations are performed on these two time series.

[0134] 2.3 Calculate the mutation degree K1

[0135] Divide the line loss value time series into a left time series and a right time series according to the mutation day. Calculate the mean value mean_left of the line loss values in the left time series, the standard deviation std_left of the line loss values in the left time series, the mean value mean_right of the line loss values in the right time series, and the standard deviation std_right of the line loss values in the right time series. The mutation degree K1 can be expressed as: K1 = abs(mean_left - mean_right) / max(std_left, std_right) (5)

[0136] Example: For the left time series LPQ_LEFT = [65, 54, 46, 43, 46, 56, 62], the mean value mean_left is 53, and the standard deviation std_left is 7.95. For the right time series LPQ_RIGHT = [73, 85, 86, 72, 80, 77, 79, 92, 93, 78, 69, 64, 72, 76, 93, 104, 85, 108, 87, 108], the mean value mean_right is 84, and the standard deviation std_right is 12.28. Based on this, K1 = (84 - 53) / 12.28 = 2.52 can be calculated.

[0137] 2.4 Calculate the mutation degree K2

[0138] Divide the time series of line loss values into a left time series and a right time series according to the mutation date. Select the k (default is 5, which can be adjusted according to the overall time length of the data to reflect the data differences before and after the data change) dates closest to the mutation date in the two time series to generate a left edge time series and a right edge time series. Calculate the mean of the left edge time series mean_left_edge and the mean of the right edge time series mean_right_edge. The mutation degree K2 can be expressed as: K2 = abs(mean_left_edge - mean_right_edge) / max(mean_left_edge, mean_right_edge) (6);

[0139] Example: LPQ = [65, 54, 46, 43, 46, 56, 62, 73, 85, 86, 72, 80, 77, 79, 92, 93, 78, 69, 64, 72, 76, 93, 104, 85, 108, 87, 108]. Currently, the assumed mutation date is the eighth date (corresponding to the data 73). The k = 5 dates closest to this in the left and right time series form the left and right edge time series. Then the left edge time series is LPQ_LEFT_EDGE = [46, 43, 46, 56, 62], and the corresponding mean of the left edge time series mean_left_edge is 51. The right edge time series is LPQ_RIGHT_EDGE = [73, 85, 86, 72, 80], and the corresponding mean of the right edge time series mean_right_edge is 79. Then, based on this, K2 = (79 - 51) / 79 = 0.35 can be calculated.

[0140] 2.5 Calculate the mutation degree index K

[0141] If K1 and K2 meet the threshold requirements (for the judgment of the line loss value of the transformer substation area, by default, K1 needs to be greater than 1.4 and K2 needs to be greater than 0.2, which can be adjusted according to the situation), then calculate the mutation degree index K by combining K1 and K2. The mutation index K can be expressed as:

[0142] K = K1 * K2

[0143] If K1 and K2 do not meet the threshold requirements, then do not calculate the mutation degree index K corresponding to this mutation date.

[0144] Example: For the above example data, it is calculated that K1 = 2.52 > 1.4 and K2 = 0.35 > 0.2, indicating that it is feasible to use the eighth date of this section of data as the mutation date. Therefore, record this date and the corresponding K = K1 * K2 = 1.95 * 0.24 = 0.882.

[0145] 2.6 Determine the mutation date

[0146] The mutation day with the largest mutation index K value is selected as the time series mutation point of the line loss value of the substation.

[0147] Example: The same substation may have multiple dates as mutation dates, and the corresponding K1 and K2 can all meet the threshold requirements. In this case, the date with the largest K is selected as the final mutation date of the line loss value of this substation. In the above example, 0.882 is the largest K value among all K values ​​that meet the K1 and K2 threshold conditions, so the corresponding eighth date is the final mutation date of the line loss value of this substation.

[0148] 2.7 Calculate the line loss mutation value

[0149] The line loss value time series is divided into left time series and right time series based on the determined mutation point, and their means mean_left and mean_right are calculated respectively. The absolute error between the two is calculated as the line loss value mutation value, which can be expressed as:

[0150] delta_mean_lpq=abs(mean_left-mean_right)(7)

[0151] For example: If the mutation date of the substation area is determined to be the eighth date, the corresponding left sequence is LPQ_LEFT = [65, 54, 46, 43, 46, 56, 62], the mean value mean_left is 53, and the right sequence is LPQ_RIGHT = [73, 85, 86, 72, 80, 77, 79, 92, 93, 78, 69, 64, 72, 76, 93, 104, 85, 108, 87, 108], the mean value mean_right is 84. The line loss mutation value can be calculated as delta_mean_lpq = 84-53 = 31

[0152] 3. Identification of sudden changes in user electricity consumption

[0153] 3.1 Electricity consumption data preprocessing

[0154] First, obtain the power consumption data of users in abnormal areas for the last N days [N≥30], and remove the abnormal data with power consumption less than 0. If the number of remaining data days after screening is still greater than n (the default is N*66%, and the threshold can be adjusted according to actual needs), then regenerate the power consumption time series sequence from the filtered data in the original time order: KWH=[kwh 1 ,kwh 2 ,…,kwh k ];(8)

[0155] Among them, kwh k Represents the kth data in the electricity consumption time series sequence that is regenerated after filtering.

[0156] If the remaining number of days after screening is not greater than n, then this user is regarded as a user with abnormal data and no subsequent calculations are performed.

[0157] Example: The original electricity consumption data for 30 days is [28.1, 29.2, 25.4, 28.8, 30.6, 31.5, 17.8, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]. All the data meet the threshold and no data is excluded. Then the electricity consumption time series data is: KWH = [28.1, 29.2, 25.4, 28.8, 30.6, 31.5, 17.8, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]. The remaining 30-day data meets more than 66% of the total number of days. Therefore, this user is not a user with abnormal data and subsequent mutation calculations can be performed.

[0158] 3.2 Calculate the mutation degree corresponding to the date

[0159] In the electricity consumption time series KWH, m dates in the forward order (by default, 2m is N * 33%, which can be adjusted according to requirements) and m dates in the reverse order do not participate in the calculation of the corresponding mutation degree (because selecting a few days at the beginning or end as mutation days will result in a too short time series being segmented, affecting the calculation results).

[0160] Among the remaining dates, traverse each date. Take each daily date as the mutation day, divide the time series into a left time series (date less than the mutation day) and a right time series (date greater than or equal to the mutation day), calculate the mutation degree between the two time series, and retain it for subsequent comparison.

[0161] Example: After processing, KWH = [28.1, 29.2, 25.4, 28.8, 30.6, 31.5, 17.8, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]. At this time, the eighth date (corresponding data 0) is selected as the mutation day. Then, according to this date, KWH is divided into a left time series KWH_LEFT = [28.1, 29.2, 25.4, 28.8, 30.6, 31.5, 17.8] and a right time series KWH_RIGHT = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]. Subsequent index calculations are performed on these two time series.

[0162] 3.3 Calculate the mutation degree K1

[0163] The electricity consumption time series is divided into a left time series and a right time series according to the mutation date. Calculate the mean electricity consumption of the left time series mean_left, the standard deviation of the left time series std_left, the mean electricity consumption of the right time series mean_right, and the standard deviation of the right time series std_right. The mutation degree K1 can be expressed as: K1 = abs(mean_left - mean_right) / max(std_left, std_right) (9)

[0164] Example: The left time series KWH_LEFT = [28.1, 29.2, 25.4, 28.8, 30.6, 31.5, 17.8], the mean mean_left is 27.3, and the standard deviation std_left is 4.31. The right time series KWH_RIGHT = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], the mean mean_right is 0, and the standard deviation std_right is 0. Based on this, K1 = (27.3 - 0) / 4.31 = 6.33 can be calculated.

[0165] 3.4 Calculate the mutation degree K2

[0166] The electricity consumption time series is divided into a left time series and a right time series according to the mutation date. Take the k (default is 5, which can be adjusted according to the overall time length of the data to reflect the data difference before and after the data change) dates closest to the mutation date in the two time series to generate a left edge time series and a right edge time series. Calculate the mean of the left edge time series mean_left_edge and the mean of the right edge time series mean_right_edge. The mutation degree K2 can be expressed as: K2 = abs(mean_left_edge - mean_right_edge) / (max(mean_left_edge, mean_right_edge) (10)

[0167] Example: KWH = [28.1, 29.2, 25.4, 28.8, 30.6, 31.5, 17.8, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]. Currently, the assumed mutation date is the eighth date (corresponding data 0). The k = 5 dates closest to this in the left and right time series form the left and right edge time series. Then the left edge time series is KWH_LEFT_EDGE = [25.4, 28.8, 30.6, 31.5, 17.8], and the corresponding left edge time series mean mean_left_edge is 26.8. The right edge time series is KWH_RIGHT_EDGE = [0, 0, 0, 0, 0], and the corresponding right edge time series mean mean_right_edge is 0. Based on this, K2 = (26.8 - 0) / 26.8 = 1 can be calculated.

[0168] 3.5 Calculate the mutation degree index K

[0169] If K1 and K2 meet the threshold requirements (for judging the user's electricity consumption, by default, K1 needs to be greater than 2.5 and K2 needs to be greater than 0.7, which can be adjusted according to the situation), then calculate the mutation degree index K by combining K1 and K2. The mutation index K can be expressed as:

[0170] K = K1 * K2

[0171] If K1 and K2 do not meet the threshold requirements, then do not calculate the mutation degree index K corresponding to this mutation date.

[0172] Example: For the above example data, it is calculated that K1 = 6.33 > 2.5 and K2 = 1 > 0.7, indicating that it is feasible to use the eighth date of this section of data as the mutation date. Therefore, record this date and the corresponding K = K1 * K2 = 6.33 * 1 = 6.33.

[0173] 3.6 Determine the mutation date

[0174] Select the mutation date corresponding to the largest mutation index K value as the power consumption time series mutation point of this substation area.

[0175] Example: For the same user, there may be multiple dates for which the corresponding K1 and K2 of the mutation date can both meet the threshold requirements. In this case, select the date with the largest K as the final power consumption mutation date of this user. In the above example, 6.33 is the largest K value among all K values that meet the K1 and K2 threshold conditions. Therefore, the corresponding eighth date is the final mutation date of this user's power consumption.

[0176] 3.7 Calculate the power consumption mutation value

[0177] The power consumption time series is segmented into a left time series and a right time series based on the determined mutation points, and their means mean_left and mean_right are calculated respectively. The absolute error between the two is calculated as the power consumption mutation value, which can be expressed as: delta_mean_kwh = abs(mean_left - mean_right) (11)

[0178] Example: If the determined mutation date of the user's power consumption is the eighth date, then the corresponding left time series is KWH_LEFT = [28.1, 29.2, 25.4, 28.8, 30.6, 31.5, 17.8], the mean mean_left is 27.3, and the right time series is KWH_RIGHT = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], and the mean mean_right is 0. The power consumption mutation value can be calculated as delta_mean_kqh = 27.3 - 0 = 27.3. After calculating the value of this indicator, it is reserved for later use.

[0179] IV. Evaluate the impact of the user's power consumption mutation degree on the line loss of the transformer substation area

[0180] 4.1 Compare mutation dates

[0181] Select users whose mutation dates of the line loss rate of the transformer substation area, mutation dates of the line loss value of the transformer substation area, and mutation dates of the user's power consumption are all on the same day as abnormal users.

[0182] Example: After calculation, the mutation date of the line loss rate of the transformer substation area, the mutation date of the line loss value of the transformer substation area, and the mutation date of the user's power consumption are all determined to be the eighth date (in this example, they are all December 17, 2022. The elimination due to abnormal data does not affect the actual date corresponding to the data), which is the same day. Therefore, this user becomes an abnormal user.

[0183] 4.2 Calculate the similarity between the mutation amount of the user's power consumption and the mutation amount of the line loss value of the transformer substation area

[0184] Calculate the similarity between the mutation amount of the line loss value of the transformer substation area of the abnormal user and the mutation amount of the user's power consumption, which can be expressed as:

[0185] loss_sim = min(delta_mean_lqp, delta_mean_kwh) / max(delta_mean_lqp, delta_mean_kwh)

[0186] Output the users corresponding to the threshold whose similarity meets the requirements (the threshold is 80%, which can be adjusted according to the situation) as the final abnormal users.

[0187] Example: It has been calculated that delta_mean_lpq = 31 and delta_mean_kqh = 27.3. Therefore, loss_sim = 27.3 / 31 * 100% = 88% > 80%. Thus, the similarity requirement is met, and this user is output as the final abnormal user.

[0188] Embodiment 2:

[0189] Please refer to Figure 2 as shown in Figure 2 which shows a device for identifying inaccurate metering of low-voltage meters, including:

[0190] The first mutation date determination unit is used to obtain the line loss rate data of the target substation area in a preset time period, and determine the first mutation date of the line loss rate of the target substation area and the abnormal substation areas in the target substation area according to the line loss rate data;

[0191] The second mutation date determination unit is used to obtain the power consumption data of each user in the abnormal substation area in a preset time period, and determine the second mutation date of the power consumption data of each user in the abnormal substation area according to the power consumption data;

[0192] The third mutation date determination unit is used to obtain the line loss value data of the abnormal substation area in a preset time period, and determine the third mutation date of the line loss value of the target substation area according to the line loss value data;

[0193] The first determination unit is used to determine that a user is an abnormal user when both the first mutation date and the third mutation date are the same as the second mutation date of the power consumption data of a certain user;

[0194] The similarity determination unit is used to calculate the similarity between the line loss mutation value corresponding to the abnormal user and the power consumption mutation value of the abnormal user, and obtain the similarity of each abnormal user;

[0195] The second determination unit is used to determine that an abnormal user is the final abnormal user when the similarity corresponding to a certain abnormal user is greater than the similarity threshold.

[0196] Embodiment 3:

[0197] Figure 3 This is a schematic structural diagram of an electronic device provided by an embodiment of the present application. In the present application, it can be Figure 3 described by the schematic diagram shown in

[0198] a schematic structural diagram of an electronic device as shown in Figure 3. The electronic device 100 includes one or more processors 102 and one or more storage devices 104, and these components are interconnected through a bus system and / or other forms of connection mechanisms (not shown). It should be noted thatFigure 3 The components and structures of the electronic device 100 shown are merely exemplary and not restrictive. As needed, the electronic device may have Figure 3 some of the components shown, or may also have Figure 3 other components and structures not shown.

[0199] The processor 102 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 100 to perform desired functions.

[0200] The storage device 104 may include one or more computer program products, and the computer program products may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory, etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage media, and the processor 102 may run the program instructions to implement the functions (implemented by the processor) in the embodiments of the present application described below and / or other desired functions. Various application programs and various data may also be stored in the computer-readable storage media, such as various data used and / or generated by the application programs, etc.

[0201] The present invention also provides a computer storage medium, on which a computer program is stored. If the method of the present invention is implemented in the form of software functional units and sold or used as an independent product, it can be stored in this computer storage medium. Based on such an understanding, to implement all or part of the processes in the above-described embodiment methods, the present invention can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer storage medium. When the computer program is executed by a processor, the steps of the above-described various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer storage medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer storage medium can be appropriately increased or decreased according to the requirements of legislation and patent practice within the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer storage medium does not include electrical carrier signals and telecommunication signals.

[0202] For those skilled in the art, various corresponding changes and deformations can be made according to the technical solutions and concepts described above, and all such changes and deformations should fall within the protection scope of the claims of the present invention.

Claims

1. A method for identifying inaccurate metering of low-voltage meters, characterized in that, it includes the following steps: Obtain the line loss rate data of the target substation area in a preset time period, and determine the first mutation date of the line loss rate of the target substation area and the abnormal substation areas in the target substation area according to the line loss rate data; Obtain the power consumption data of each user in the abnormal substation area in a preset time period, and determine the second mutation date of the power consumption data of each user in the abnormal substation area according to the power consumption data; Obtain the line loss value data of the abnormal substation area in a preset time period, and determine the third mutation date of the line loss value of the target substation area according to the line loss value data; If both the first mutation date and the third mutation date are the same as the second mutation date of the power consumption data of a certain user, then determine that this user is an abnormal user; Calculate the similarity between the line loss mutation value corresponding to the abnormal user and the power consumption mutation value of the abnormal user to obtain the similarity of each abnormal user; If the similarity corresponding to an abnormal user is greater than the similarity threshold, then determine that this abnormal user is the final abnormal user; The determining the first mutation date of the line loss rate of the target substation area according to the line loss rate data includes: After obtaining the line loss rate data of the target substation area in a preset time period, preprocess the line loss rate data to obtain a preprocessed line loss rate time series; Calculate the first mutation degree and the second mutation degree corresponding to each line loss rate in the line loss rate time series; among them, the first mutation degree corresponding to each line loss rate is calculated according to the first left sequence and the first right sequence corresponding to each line loss rate, and the second mutation degree corresponding to each line loss rate is calculated according to the second left sequence and the second right sequence corresponding to each line loss rate; based on a certain line loss rate, the line loss rate time series is divided to obtain the first left sequence and the first right sequence corresponding to this line loss rate; based on the first left sequence, a second left sequence is obtained, and based on the first right sequence, a second right sequence is obtained; specifically including: obtaining the time point corresponding to each line loss rate in the line loss rate time series, arranging the line loss rates in the line loss rate time series in chronological order, taking the sequence composed of the line loss rates before a certain line loss rate as the first left sequence, and taking the sequence composed of a certain line loss rate and the line loss rates after a certain line loss rate as the first right sequence; in the first left sequence, taking the sequence composed of m line loss rates before a certain line loss rate as the second left sequence; in the first right sequence, taking the sequence composed of a certain line loss rate and m-1 line loss rates after a certain line loss rate as the second right sequence; where m is a natural number; The first mutation degree K1 and the second mutation degree respectively satisfy the following formulas: K 1 = abs(mean_left - mean_right) / max(std_left, std_right); Among them, abs is the absolute value function of a number, mean_left is the mean of the first left time series line loss rate, mean_right is the mean of the first right time series line loss rate, std_right is the standard deviation of the first right time series line loss rate, and std_left is the standard deviation of the first left time series line loss rate; K 2 = abs(mean_left_edge - mean_right_edge) / max(mean_left_edge, mean_right_edge); Among them, abs is the absolute value function of a number, mean_left_edge is the mean of the second left time series line loss rate, and mean_right_edge is the mean of the second right time series line loss rate; Calculate the first mutation index corresponding to each line loss rate in the time series of the line loss rate. The first mutation index is the product of the first mutation degree and the second mutation degree. Determine the first mutation date of the line loss rate of the target substation area according to the first mutation indexes corresponding to all line loss rates.

2. The method for identifying inaccurate metering of low-voltage meters according to claim 1, wherein, the determination method of the abnormal substation area in the target substation area is as follows: If there is a first mutation index corresponding to the line loss rate in the time series of the line loss rate of the target substation area that is greater than a preset threshold, it is determined that the target substation area is an abnormal substation area.

3. The method for identifying inaccurate metering of low-voltage meters according to claim 1, wherein, the determination of the first mutation date of the line loss rate of the target substation area according to the first mutation indexes corresponding to all line loss rates includes: Obtain the maximum value among all the first mutation indexes greater than the preset threshold, and use the time and date corresponding to the maximum value as the first mutation date of the line loss rate of the target substation area.

4. The method for identifying inaccurate metering of low-voltage meters according to claim 1, wherein, the determination of the second mutation date of the power consumption data of each user in the abnormal substation area according to the power consumption data includes: After obtaining the power consumption data of each user in the preset time period of the abnormal substation area, preprocess the power consumption data to obtain a preprocessed time series of power consumption; Calculate the third mutation degree and the fourth mutation degree corresponding to each power consumption data in the time series of power consumption. Among them, the third mutation degree corresponding to each power consumption data is calculated according to the third left sequence and the third right sequence corresponding to each power consumption data, and the fourth mutation degree corresponding to each power consumption data is calculated according to the fourth left sequence and the fourth right sequence corresponding to each power consumption data. Based on a certain power consumption data, the time series of power consumption is divided into a third left sequence and a third right sequence corresponding to the power consumption data. The fourth left sequence is obtained based on the third left sequence, and the fourth right sequence is obtained based on the third right sequence. Specifically, the time series of power consumption is divided into a third left sequence and a third right sequence according to the mutation date, and k dates closest to the mutation date in the two time series are taken, where k is a natural number, to generate a fourth left sequence and a fourth right sequence. The third mutation degree K1' = abs(mean_left'-mean_right') / max(std_left',std_right'); The fourth mutation degree K2' = abs(mean_left_edge'-mean_right_edge') / max(mean_left_edge',mean_right_edge'); where mean_left’ is the mean of the third left sequence, std_left’ is the standard deviation of the third left sequence, mean_right’ is the mean of the third right sequence, std_right’ is the standard deviation of the third right sequence; mean_left_edge’ is the mean of the fourth left sequence, and mean_right_edge’ is the mean of the fourth right sequence. Calculate the second mutation index corresponding to each power consumption data in the time series of power consumption, where the second mutation index is the product of the third mutation degree and the fourth mutation degree; determine the second mutation date of the abnormal substation area power consumption data according to the second mutation indexes corresponding to all power consumption data.

5. The method for identifying inaccurate metering of low-voltage meters according to claim 1, wherein, the determining the third mutation date of the line loss value of the target substation area according to the line loss value data includes: After obtaining the line loss value data of the abnormal substation area in a preset time period, preprocess the line loss value data to obtain a preprocessed time series of line loss values; Calculate the fifth mutation degree and the sixth mutation degree corresponding to each line loss rate in the time series of line loss values; among them, the fifth mutation degree corresponding to each line loss value is calculated according to the fifth left sequence and the fifth right sequence corresponding to each line loss value, and the second mutation degree corresponding to each line loss value is calculated according to the sixth left sequence and the sixth right sequence corresponding to each line loss value; based on a certain line loss value, divide the time series of line loss values to obtain the fifth left sequence and the fifth right sequence corresponding to the line loss value; obtain the sixth left sequence based on the fifth left sequence, and obtain the sixth right sequence based on the fifth right sequence; specifically, divide the time series of line loss values into the fifth left sequence and the fifth right sequence according to the mutation date; take the k dates closest to the mutation date in the two time series, where k is a natural number, to generate the sixth left sequence and the sixth right sequence, and the fifth mutation degree K1” = abs(mean_left” - mean_right”) / max(std_left”, std_right”); The sixth mutation degree K2” = abs(mean_left_edge” - mean_right_edge”) / max(mean_left_edge”, mean_right_edge”); where, mean_left” is the mean of the fifth left sequence, std_left” is the standard deviation of the fifth left sequence, mean_right” is the mean of the fifth right sequence, std_right” is the standard deviation of the fifth right sequence; mean_left_edge” is the mean of the sixth left sequence, and mean_right_edge” is the mean of the sixth right sequence; Calculate the third mutation index corresponding to each line loss value in the time series of line loss values, where the third mutation index is the product of the fifth mutation degree and the sixth mutation degree; determine the third mutation date of the line loss value of the target substation area according to the third mutation indexes corresponding to all line loss values.

6. The method for identifying inaccurate metering of low-voltage meters according to claim 1, wherein, The similarity between the line loss mutation value corresponding to the abnormal user and the power consumption mutation value of the abnormal user satisfies the following formula: loss_sim = min(delta_mean_lqp, delta_mean_kwh) / max(delta_mean_lqp, delta_mean_kwh); Among them, delta_mean_lpq is the line loss mutation value corresponding to the abnormal user, delta_mean_kqh is the electricity consumption mutation value of the abnormal user, and loss_sim is the similarity between the line loss mutation value corresponding to the abnormal user and the electricity consumption mutation value of the abnormal user.

7. The method for identifying the inaccurate metering of low-voltage meters according to claim 6, characterized in that, the line loss mutation value corresponding to the abnormal user is obtained based on the seventh left sequence and the seventh right sequence, and the line loss value mutation value satisfies delta_mean_lpq = abs(mean-left”'-mean-right”'); mean-left”' is the mean value of the seventh left sequence, and mean-right”' is the mean value of the seventh right sequence; the line loss value time series is divided into the seventh left sequence and the seventh right sequence according to the third mutation date; the mutation value of the electricity consumption of the abnormal user is obtained based on the eighth left sequence and the eighth right sequence, and the electricity consumption mutation value satisfies: delta_mean_kwh = abs(mean_left””-mean_right””); where mean_left”” is the mean value of the eighth left sequence, and mean_right”” is the mean value of the eighth right sequence; the time series of the electricity consumption is divided into the eighth left sequence and the eighth right sequence according to the second mutation date.

8. A device for identifying the inaccurate metering of low-voltage meters, characterized in that, comprising: A first mutation date determination unit, configured to obtain the line loss rate data of the target substation area in a preset time period, and determine the first mutation date of the line loss rate of the target substation area and the abnormal substation area in the target substation area according to the line loss rate data; The first mutation date for determining the line loss rate of the target substation area based on the line loss rate data includes: after obtaining the line loss rate data of the target substation area in a preset time period, preprocessing the line loss rate data to obtain a preprocessed line loss rate time series; calculating a first mutation degree and a second mutation degree corresponding to each line loss rate in the line loss rate time series; wherein, the first mutation degree corresponding to each line loss rate is calculated based on a first left sequence and a first right sequence corresponding to each line loss rate, and the second mutation degree corresponding to each line loss rate is calculated based on a second left sequence and a second right sequence corresponding to each line loss rate; dividing the line loss rate time series based on a certain line loss rate to obtain a first left sequence and a first right sequence corresponding to the line loss rate; obtaining a second left sequence based on the first left sequence and obtaining a second right sequence based on the first right sequence; specifically including: obtaining the time point corresponding to each line loss rate in the line loss rate time series, arranging the line loss rates in the line loss rate time series in chronological order, taking the sequence composed of the line loss rates before a certain line loss rate as the first left sequence, and taking the sequence composed of the certain line loss rate and the line loss rates after the certain line loss rate as the first right sequence; in the first left sequence, taking the sequence composed of m line loss rates before a certain line loss rate as the second left sequence; in the first right sequence, taking the sequence composed of the certain line loss rate and m - 1 line loss rates after the certain line loss rate as the second right sequence; wherein, m is a natural number, and the first mutation degree K1 and the second mutation degree respectively satisfy the following formulas: K 1 = abs(mean_left - mean_right) / max(std_left, std_right); where abs is the absolute value function of a number, mean_left is the mean of the first left time series line loss rate, mean_right is the mean of the first right time series line loss rate, std_right is the standard deviation of the first right time series line loss rate, and std_left is the standard deviation of the first left time series line loss rate; K 2 = abs(mean_left_edge - mean_right_edge) / max(mean_left_edge, mean_right_edge); Wherein, abs is the absolute value function of a number, mean_left_edge is the average value of the line loss rate of the second left time series, and mean_right_edge is the average value of the line loss rate of the second right time series; calculating a first mutation index corresponding to each line loss rate in the line loss rate time series, and the first mutation index is the product of the first mutation degree and the second mutation degree; determining the first mutation date of the line loss rate of the target substation area according to the first mutation indexes corresponding to all line loss rates; A second mutation date determination unit, configured to obtain the power consumption data of each user in the abnormal substation area in a preset time period, and determine the second mutation date of the power consumption data of each user in the abnormal substation area according to the power consumption data; A third mutation date determination unit, configured to obtain the line loss value data of the abnormal substation area in a preset time period, and determine the third mutation date of the line loss value of the target substation area according to the line loss value data; A first determination unit, configured to determine that the user is an abnormal user when both the first mutation date and the third mutation date are the same as the second mutation date of the power consumption data of a certain user; A similarity determination unit, configured to calculate the similarity between the line loss mutation value corresponding to the abnormal user and the power consumption mutation value of the abnormal user, and obtain the similarity of each abnormal user; A second determination unit, configured to determine that the abnormal user is a final abnormal user when the similarity corresponding to a certain abnormal user is greater than the similarity threshold; 9. A computer storage medium, on which a computer program is stored, characterized in that, when the computer program is executed by a processor, the method for identifying the inaccurate measurement of the low-voltage meter as claimed in any one of claims 1 - 7 is implemented.

Images