Identification method based on basic characteristics of abnormal metering power utilization indexes

Abstract

The invention provides an identification method based on basic characteristics of an abnormal metering power utilization index, and the method comprises the steps: collecting power utilization index data, and forming a time sequence curve; comparing the power utilization index data with the measurement abnormal index basic feature library, and screening out suspected abnormal days; selecting two normal days, drawing current, voltage, load and electric quantity curves, and extracting curve morphological characteristics and three-phase balance characteristics; calculating the relevance between intra-day indexes of the suspected abnormal day; calculating cross-day index relevance between the suspected abnormal day and the normal day; and carrying out secondary judgment. According to the identification method based on the basic characteristics of the abnormal measurement power utilization indexes, daily monitoring indexes and the basic characteristic library of the abnormal measurement indexes are compared and preliminarily screened, the power utilization curve of a suspected abnormal power utilization day is subjected to morphological analysis, and intra-day / cross-day multi-dimensional correlation calculation based on the Pearson's correlation coefficient is carried out; and high-precision, low-false-alarm and high-adaptability intelligent research and judgment on metering abnormity are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of electric quantity metering, in particular to a method for identifying abnormal electric quantity metering based on basic features of abnormal electric quantity metering indexes. BACKGROUND

[0002] Under the background of the accelerated construction of smart grid and the digital transformation of power consumption management, as the core energy supporting the operation of society, the consumption of electricity continues to rise. However, abnormal power consumption behaviors such as electricity theft and metering device failure also occur frequently, which not only causes huge economic losses to power supply enterprises, but also seriously threatens the safe operation of the power grid and hinders the high-quality development of the smart grid. At present, the detection of abnormal power consumption behaviors in the industry mainly relies on single rule threshold or lagging work order and data comparison, and generally lacks a multi-dimensional and multi-index cross-verification mechanism, which is difficult to adapt to the complex and variable actual scenarios of user power consumption behaviors, resulting in low accuracy of abnormal identification, high false positive and false negative rates. Especially for abnormal types with strong concealment and weak features such as reverse current and slow running of electric energy meter, traditional methods often fail to effectively capture their subtle changes.

[0003] The statements herein merely provide background technology related to the present application, and do not necessarily constitute the prior art. SUMMARY

[0004] The purpose of the present application is to provide a method for identifying abnormal electric quantity metering based on basic features of abnormal electric quantity metering indexes, which has the advantage of high accuracy of abnormal identification.

[0005] To achieve the above-mentioned purpose, the present application provides a method for identifying abnormal electric quantity metering based on basic features of abnormal electric quantity metering indexes, which comprises: S10, collecting electric quantity index data of a target user on multiple dates, the electric quantity index data including three-phase current, three-phase voltage, load and active power data, forming a time series curve; S20, comparing the electric quantity index data with a pre-constructed basic feature library of abnormal electric quantity metering indexes, and screening out at least one index feature abnormal date as a suspected abnormal day; S30, selecting two normal days matching the power consumption behavior type of the suspected abnormal day, drawing the current, voltage, load and power curves of the suspected abnormal day and the two normal days, and extracting the curve shape features and three-phase balance features; S40, calculating the intra-day index correlation of the suspected abnormal day; calculating the cross-day index correlation between the suspected abnormal day and the normal day; S50, performing secondary judgment to confirm that the suspected abnormal day has abnormal metering.

[0006] In an embodiment, the correlation between the intraday indexes of the suspected abnormal day in step S40 includes the correlation between three-phase currents, the correlation between current and load, the correlation between load and power, and the correlation between the suspected abnormal day and the normal day includes the correlation between same-phase current-current, the correlation between same-phase voltage-voltage, the correlation between load-load, and the correlation between power-power; in step S50, secondary judgment is performed based on the index characteristics, the curve shape characteristics, the three-phase balance characteristics, and the correlation calculated in step S40.

[0007] In an embodiment, in step S10, the power consumption index data is collected by high-speed power line carrier communication technology.

[0008] In an embodiment, the power consumption index data is collected every 5-20 minutes.

[0009] In an embodiment, the metering abnormal index basic feature library includes at least one of the following abnormal types: “existence of reverse current”, “electric energy meter stop running”, “existence of current and non-existence of power”, “current loss”, “electric energy meter flying”, “current reverse connection or wiring error”, “electric energy meter reverse running”, or “power overflow”.

[0010] In an embodiment, in step S30, the curve shape characteristics include: negative value segment in the current or load curve, platform segment in the power curve, load peak value offset, or curve dramatic fluctuation.

[0011] In an embodiment, in step S40, the Pearson correlation coefficient method is used to calculate the correlation between indexes.

[0012] In an embodiment, in step S40, when the Pearson correlation coefficient between three-phase currents is lower than 0.9, it is determined that three-phase imbalance is abnormal.

[0013] In an embodiment, in step S40, the threshold of the cross-day index correlation is set as: the adaptive threshold of current curve correlation is 0.9, and the adaptive threshold of voltage curve correlation is 0.8.

[0014] In an embodiment, the two normal days are historical normal power consumption days, which are selected from historical data according to user power consumption behavior types.

[0015] In summary, compared with the prior art, the recognition method based on metering abnormal power consumption index basic features provided by the present application has the following beneficial effects: The identification method based on metering abnormal electricity index basic features of the application, by comparing the daily monitoring index with the metering abnormal index basic feature library for preliminary screening, performing morphological analysis on the electricity curve of the suspected abnormal electricity day, and based on the intraday / cross-day multi-dimensional correlation calculation of Pearson correlation coefficient, realizes high-precision, low-false alarm and strong adaptability intelligent research and judgment of metering abnormality; without additional hardware investment, significantly improves the efficiency and accuracy of abnormal identification, especially suitable for complex and variable actual electricity consumption scenarios. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 The flowchart of the identification method based on metering abnormal electricity index basic features of the application.

[0017] Figure 2 The application flowchart of the identification method based on metering abnormal electricity index basic features of the application.

[0018] Figure 3 The schematic diagram of the curve correlation degree calculation method.

[0019] Figure 4 The typical electricity index curve of the suspected abnormal day user.

[0020] Figure 5 The typical electricity index curve of the first normal day user.

[0021] Figure 6 The typical electricity index curve of the second normal day user. DETAILED DESCRIPTION

[0022] The identification method based on metering abnormal electricity index basic features of the application is further described in detail below in combination with the drawings and specific embodiments. The advantages and features of the application will be clearer according to the following description. It should be noted that the drawings are greatly simplified and all use non-precise proportions, only to facilitate, clearly assist in explaining the purpose of the embodiments of the application. In order to make the purpose, features and advantages of the application more obvious and easy to understand, please refer to the drawings. It should be noted that the structure, proportion, size, etc. shown in the drawings attached to the specification are only used to cooperate with the content disclosed in the specification, so that those skilled in the art can understand and read, and are not used to limit the limited conditions of the implementation of the application, so they do not have technical significance. Any modification of structure, change of proportion relationship or adjustment of size, without affecting the effect and purpose that can be achieved by the application, should still fall within the scope of the technical content disclosed by the application.

[0023] As shown in Figure 1 and Figure 2 The application provides an identification method based on metering abnormal electricity index basic features, which comprises: S10, collect the power consumption index data of the target user on multiple dates, in the embodiment, the power consumption index data is collected by HPLC (High-speed Power Line Carrier) technology, and the power consumption index data includes three-phase current data, three-phase voltage data, load data and active power data, and a time sequence curve is formed; The power consumption index data is collected every 5-20 minutes, and in the embodiment, the power consumption index data is collected every 15 minutes, that is, 96 times of power consumption index data are collected in 24 hours.

[0024] S20, compare the power consumption index data with the pre-constructed metering abnormal index basic feature library, and select at least one index feature abnormal date as a suspected abnormal day. By summarizing the index data of different metering abnormal typical users, the metering abnormal index basic feature library can be induced. Before extracting the curve features, the metering abnormal index basic feature library is used for basic judgment of the metering abnormal data. By setting the metering abnormal index basic feature library, the normal and abnormal power consumption indexes are distinguished and compared, the abnormal occurrence date is judged, the suspected abnormal date curve is found, and the research and judgment amount is reduced.

[0025] The metering abnormal index basic feature library includes at least one of the following abnormal types: "existence of reverse current", "electric energy meter stop", "current without power", "current loss", "electric energy meter fly", "current reverse connection or wiring error", "electric energy meter reverse", or "power overflow".

[0026] S30, select two normal days matching the power consumption behavior type of the suspected abnormal day, draw the current, voltage, load and power curve of the suspected abnormal day and the two normal days, and extract the curve shape feature and three-phase balance feature. In the embodiment, the curve shape feature includes: negative value segment in the current or load curve, platform segment in the power curve, load peak value offset or curve sharp fluctuation. The two normal days are historical normal power consumption days, which are selected from the historical data according to the user power consumption behavior type. In actual power consumption process, different users often have different current curves, curve voltages and load curves due to different power consumption behaviors and habits, and the above-mentioned curves present certain regularity, such as obvious repeated power consumption regularity characteristics of current curves in certain time period of working days and double holidays, so that the typical data curves can be induced as the typical curves of normal power consumption period.

[0027] The following table is the basic feature of normal power consumption data curve. In the embodiment, the balance refers to the balance of three-phase current and three-phase voltage in the three-phase circuit, the three-phase current / voltage balance refers to the equal amplitude, same frequency and 120-degree phase difference of the current / voltage of three phases, and the feature in the curve is the high correlation of the three-phase current curve. In the actual power consumption process, when the single-phase load is unevenly distributed or the power supply fails, it may cause three-phase imbalance, increase the overall line loss and transformer loss, and generate a large zero sequence current. When the three-phase is unbalanced, the correlation coefficient between the three-phase currents will also decrease significantly.

[0028] S40, calculate the correlation between the intraday indicators of the suspected abnormal day, including the correlation degrees between the three-phase currents, the current and the load, the load and the power; calculate the cross-day indicator correlation between the suspected abnormal day and the normal day, including the correlation degrees of the same-phase current-current, the same-phase voltage-voltage, the load-load and the power-power. In the embodiment, the Pearson correlation coefficient method is used to calculate the correlation between the indicators. The correlation refers to the closeness of the linear relationship between two variables, which can be calculated by the Pearson correlation coefficient. The closer the absolute value of the Pearson correlation coefficient r is to 1, the stronger the linear relationship is. For example, the current and the load, in the ideal case, the voltage is constant, and according to the load formula, the load is proportional to the current. In this case, the Pearson correlation coefficient between the current and the voltage is close to 1. However, in the actual circuit, due to the existence of inductors, capacitors and other components and nonlinear loads, the voltage is in a small fluctuation state, and the current and the load are not completely proportional, and the correlation coefficient of the two may deviate from 1.

[0029] Specifically, the correlation degree of the feature load curve and the daily load curve at the same time of a phase The calculation formula is: Taking the load curve as an example, in the formula x and y represent the load values of the feature load curve and the selected daily load curve at the same time of a phase, respectively; and represent the load values of the feature load and the daily load at the i th time, respectively; and represent the average load values of the feature load and the daily load at all times, respectively; represents the covariance of the load values of the feature load curve and the daily load curve; and represent the standard deviations of the load values of the feature load curve and the daily load curve, respectively. When the absolute value of the correlation degree is calculated to be between 0.8 and 1.0, it means that the two numerical curves have significant correlation.

[0030] In the embodiment, when the Pearson correlation coefficient between the three-phase currents is lower than 0.9, it is determined that the three-phase imbalance is abnormal. The threshold of the cross-day index correlation is set as: the adaptive threshold of the current curve correlation is 0.9; the adaptive threshold of the voltage curve correlation can be set as 0.8.

[0031] S50, based on the index characteristics, the curve shape characteristics, the three-phase balance characteristics and the correlation degree calculated in step S40, secondary judgment is performed, and if multiple dimensions meet the abnormality determination condition, it is determined that the suspected abnormal day exists metering abnormality. Among them, the curve characteristic dimension calculation in the secondary judgment includes direct observation method and curve characteristic calculation method.

[0032] The curve characteristic direct observation method draws the suspected abnormal day curve and the normal day curve (such as the average curve or multiple normal curves) in the same graph, and directly observes the shape difference, such as peak shift, curve flattening or severe fluctuation.

[0033] The curve characteristic calculation method includes data preprocessing, establishment of normal model and abnormality detection. The data preprocessing includes dividing the historical load curve and the curve to be detected into N segments, calculating the average value of each segment to obtain an N-dimensional vector representing each curve. For each load curve (length n), it is divided into N segments. Each segment contains data points (assuming n can be divided by N). The average value of each segment of data is calculated to obtain an N-dimensional vector , where is the average value of the i-th segment. i

[0034] The establishment of the normal model includes calculating the normal statistical characteristics of each segment (i.e. each position in the vector) based on the historical vectors, including the mean and the standard deviation. Assuming there are M historical load curves, each curve is converted into an N-dimensional vector. Let be the set of historical vectors, where each is an N-dimensional vector.

[0035] The abnormality detection includes comparing the vector of the curve to be detected with the normal model to identify abnormalities. For each segment position i ( i from 1 to N), the mean and the standard deviation of the historical data are calculated.

[0036] Set the threshold coefficient k (usually k = 2 or 3, corresponding to 95% or 99% confidence level), then the normal range of the i-th segment is .

[0037] For each segment i, if v i is outside the normal range​ If the i-th segment is abnormal, then the i-th segment is marked as abnormal. Whether the entire curve is abnormal can be determined according to the number or severity of the abnormal segments. For example, if the proportion of abnormal segments exceeds a threshold (such as 10%), the curve is abnormal.

[0038] The identification method based on the basic characteristics of the metering abnormal electricity index of the present application forms a set of accurate and efficient intelligent research and judgment system of metering abnormality by constructing an intelligent research and judgment model of metering abnormality. By deeply mining user electricity data, using curve correlation analysis technology, and comprehensively using Pearson correlation coefficient method, the correlation between the feature electricity data curve and the daily electricity data curve is calculated to accurately determine whether the user electricity load has changed abnormally and identify the metering abnormality. At the same time, by analyzing the curve characteristics and correlation of electricity data, abnormal electricity data is screened out, the research and judgment calculation amount is reduced, and the efficiency and accuracy of abnormal electricity detection are further improved.

[0039] The following takes a user with a metering abnormal record of "existence of reverse current" as an example to illustrate the identification method based on the basic characteristics of the metering abnormal electricity index of the present application, but does not limit the protection scope of the identification method based on the basic characteristics of the metering abnormal electricity index of the present application to the scope of the illustrated embodiment. First, collect the full amount of electricity data in the time range of one week before and after the abnormal time of the user, perform metering abnormality research and judgment, collect and align the user electricity index data, and collect the 96-point (collected every 15 minutes, 96 points in 24 hours) curve data of daily three-phase current, three-phase voltage, load, and electricity in the time range of one week before and after the user.

[0040] Subsequently, the basic characteristics of the electricity index are analyzed. By comparing the metering abnormal index basic characteristic library (in which the metering abnormal index basic characteristic library is shown in the following table), it is found that "existence of reverse current" exists in a certain current phase. There is a negative reverse current, but the 96-point average value is greater than 0. Based on this feature, the suspected abnormal date curve and the normal date curve in the range of one week before and after are obtained. Metering Abnormal Index Basic Characteristic Library Table Then draw the user abnormal day and normal day index data curve. As shown in Figures 4-6 After drawing the suspected abnormal day and normal day index data curve of the user, the index characteristics can be obtained. After comparing with the first normal day and the second normal day index curve, the abnormal day index characteristics can be obtained. In this example, there are 2-phase currents with negative values in the three-phase current curve of the abnormal day of the user, and there are abnormal fluctuations; the three-phase voltage curve does not show obvious abnormalities; the load curve has a negative load, and then there is a steep rise to recover to normal load; the electric energy curve has a horizontal line without growth.

[0041] Then, the correlation of the daily index curve is analyzed. By calculating the correlation of the daily index curve, the curve characteristics and the three-phase balance related data can be further obtained, and the three-phase unbalance adaptive threshold can be set to 0.9. The daily index correlation analysis based on the example is shown in the following table, and it can be found that the metering anomaly has obvious three-phase current unbalance, the daily Pearson coefficient of current A-current C anomaly is 0.83273, and the daily Pearson coefficient of current B-current C anomaly is 0.86050. Then, the correlation of the daily index curve is analyzed. By calculating the correlation of the daily index curve, the curve characteristics and the three-phase balance related data can be further obtained, and the three-phase unbalance adaptive threshold can be set to 0.9. The daily index correlation analysis based on the example is shown in the following table, and it can be found that the metering anomaly has obvious three-phase current unbalance, the daily Pearson coefficient of current A-current C anomaly is 0.83273, and the daily Pearson coefficient of current B-current C anomaly is 0.86050. Finally, the secondary judgment of the metering anomaly is performed. Based on the metering anomaly sample of "existence of reverse current", the curve characteristics include: from the index characteristics and system rules, there is a negative current index and the 96-point average current is greater than 0; from the curve characteristics, there is three-phase current curve imbalance, abnormal fluctuation of load curve, and horizontal line of power curve; from the curve correlation, the three-phase current correlation degree of the abnormal day is less than 0.9, the current curve correlation degree of the abnormal day compared with the normal day is less than 0.9, and the voltage curve correlation degree of the abnormal day compared with the normal day is less than 0.8. The curve characteristics and the curve correlation degree are the basic characteristics of the metering anomaly, and the metering anomaly can be screened out by the curve characteristics and the correlation degree. The judgment of different data dimensions further supports the accuracy of the judgment.

[0042] It is to be understood that the terminology used herein such as first and second, and the like, is only intended to distinguish one entity or action from another entity or action, without necessarily requiring or implying any such actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises... a" does not, without more constraints, exclude the presence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0043] In the description of the application, it is to be understood that the orientation or positional relationship indicated by the terms "center", "height", "thickness", "upper", "lower", "vertical", "horizontal", "top", "bottom", "inner", "outer", "axial", "radial", "circumferential", and the like is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the application and simplifying the description, and does not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the application. In the description of the application, the meaning of "a plurality of" is two or more, unless otherwise specified and limited.

[0044] In the description of the application, unless otherwise explicitly specified and limited, the terms "mounting", "connecting", "connecting", "fixing" should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or it can be integrated; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium; it can be the internal communication of two elements or the interaction relationship between two elements. For those skilled in the art, the specific meaning of the above terms in the application can be understood according to the specific circumstances.

[0045] In the present application, unless otherwise explicitly specified and limited, "on" or "under" of the first feature to the second feature can include that the first and second features are in direct contact, or can include that the first and second features are not in direct contact but are in contact through another feature between them. Moreover, "on", "above" and "above" of the first feature to the second feature includes that the first feature is directly above and obliquely above the second feature, or only indicates that the horizontal height of the first feature is higher than that of the second feature. "Below", "below" and "below" of the first feature to the second feature includes that the first feature is directly below and obliquely below the second feature, or only indicates that the horizontal height of the first feature is less than that of the second feature.

[0046] While the application has been described in detail by reference to preferred embodiments thereof, it should be recognized that the description set forth herein is by way of example and that modifications of the procedures described can be employed without departing from the scope of the application. Accordingly, the scope of the application should be determined by the appended claims and equivalents thereof.

Claims

1. A method for identifying based on a metering abnormal electricity consumption index basic feature, characterized in that, The identification method comprises: S10, collecting power consumption index data of a target user on multiple dates, the power consumption index data comprising three-phase current, three-phase voltage, load and active power data, to form a time series curve; S20, comparing the power consumption index data with a pre-constructed metering abnormality index basic feature library, and screening out at least one index feature abnormality date as a suspected abnormality day; S30, selecting two normal days matching the power consumption behavior type of the suspected abnormality day, drawing current, voltage, load and power curves of the suspected abnormality day and the two normal days, and extracting curve shape features and three-phase balance features; S40, calculating intra-day index correlation of the suspected abnormality day; calculating cross-day index correlation between the suspected abnormality day and the normal days; S50, performing secondary judgment to confirm that the suspected abnormality day has metering abnormality.

2. The method for identifying based on a metering abnormal power consumption index basic feature according to claim 1, characterized in that, The intra-day index correlation of the suspected abnormality day calculated in step S40 comprises correlation degrees among three-phase currents, between current and load, between load and power; the cross-day index correlation between the suspected abnormality day and the normal days comprises correlation degrees of phase current-current, phase voltage-voltage, load-load and power-power; in step S50, secondary judgment is performed based on the index features, the curve shape features, the three-phase balance features and the correlation degrees calculated in step S40.

3. The method of claim 1, wherein the method comprises: In step S10, the power consumption index data is collected through high-speed power line carrier communication technology.

4. The method for identifying based on a metering abnormal power consumption index basic feature according to claim 3, characterized in that, The power consumption index data is collected every 5-20 minutes.

5. The method of claim 1, wherein the method comprises: The metering abnormality index basic feature library comprises features of at least one of the following abnormality types: "existence of reverse current", "electric energy meter stop", "current without power", "current loss", "electric energy meter fly", "current reverse connection or wiring error", "electric energy meter reverse", or "power overflow".

6. The method of claim 1, wherein the method comprises: In step S30, the curve shape features comprise: negative value segment in current or load curve, platform segment in power curve, load peak value offset or curve sharp fluctuation.

7. The method of claim 2, wherein the method comprises: In step S40, the Pearson correlation coefficient method is used to calculate the correlation between indexes.

8. The method for identifying based on a metering abnormal power consumption index basic feature according to claim 7, characterized in that, In step S40, when the Pearson correlation coefficient between three-phase currents is lower than 0.9, it is determined that three-phase imbalance abnormality exists.

9. The method for identifying based on a metering abnormal power consumption index basis feature according to claim 7, characterized in that, In step S40, the threshold of the cross-day index correlation is set as: the current curve correlation adaptive threshold is 0.9; the voltage curve correlation adaptive threshold can be set as 0.

8.

10. The method of claim 1, wherein, The two normal days are historical normal power consumption days, which are selected from historical data according to user power consumption behavior types as typical days without abnormality.

Images