Electricity stealing detection method and system of smart meter

Abstract

The application relates to the technical field of smart meters, in particular to a power stealing detection method and system of a smart meter, which are used for improving the accuracy of power stealing detection. The main scheme is as follows: current meter data and historical meter data of a target smart meter are acquired, an electrical abnormality score, a magnetic field interference score, a behavior mode score and a line loss abnormality score are calculated according to the current meter data and the historical meter data, a power stealing behavior detection result of the target smart meter is determined according to the electrical abnormality score, the magnetic field interference score, the behavior mode score and the line loss abnormality score, if the power stealing behavior detection result is a suspected power stealing behavior, the target smart meter sends power stealing behavior verification request information to a concentrator, the concentrator calculates a power stealing behavior detection score according to meter data of other smart meters in the area where the target smart meter is located, and whether the suspected power stealing behavior exists is verified according to the power stealing behavior detection score.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of smart meters, and particularly relates to a method and system for detecting electricity stealing of a smart meter. BACKGROUND

[0002] A smart meter is a meter used to measure electric energy, which is usually installed in a user's home and at the output end of a power distribution network, and is a power metering device for electric energy metering.

[0003] However, the existing smart meter usually only has the function of metering and displaying the amount of electricity, and cannot detect electricity stealing when electricity stealing occurs, so that timely measures cannot be taken, and electricity stealing cannot be quickly located in time, thereby resulting in poor electricity stealing prevention effect and failing to regulate electricity use. SUMMARY

[0004] Therefore, the present application provides a method and system for detecting electricity stealing of a smart meter, which can improve the accuracy of electricity stealing detection of a smart meter.

[0005] In a first aspect, the present application provides a method for detecting electricity stealing of a smart meter, which is applied to a smart meter in a smart meter electricity stealing detection system, the smart meter is connected to a concentrator in the smart meter electricity stealing detection system, and the method comprises the following steps.

[0006] Obtaining current meter data and historical meter data of a target smart meter, wherein the meter data comprises meter electrical parameters, meter electricity data, Hall magnetic sensor data, and meter state data;

[0007] Determining a data feature vector according to the current meter data and the historical meter data of the target smart meter, and inputting the data feature vector into an electricity anomaly prediction model to obtain an electricity anomaly prediction result, wherein the electricity anomaly prediction model is a neural network model trained according to meter sample data and corresponding electricity labels;

[0008] If the electricity anomaly prediction result is a probability of normal electricity use that is lower than a preset probability value, then calculating an electrical anomaly score, a magnetic field interference score, a behavior pattern score, and a line loss anomaly score according to the current meter data and the historical meter data;

[0009] Determining an electricity stealing behavior detection result of the target smart meter according to the electrical anomaly score, the magnetic field interference score, the behavior pattern score, and the line loss anomaly score;

[0010] If the electricity stealing behavior detection result is a suspected electricity stealing behavior, then the target smart meter sends electricity stealing behavior verification request information to the concentrator, so that the concentrator calculates an electricity stealing behavior detection score according to meter data of other smart meters in the area where the target smart meter is located.

[0011] The verification module verifies whether the suspected electricity stealing behavior exists according to the electricity stealing behavior detection score.

[0012] In a second aspect, the embodiments of the present application further provide a system for detecting electricity stealing of a smart meter, which comprises:

[0013] The acquisition module is configured to acquire current meter data and historical meter data of a target smart meter, wherein the meter data comprises meter electrical parameters, meter power consumption data, Hall magnetic sensor data and meter state data.

[0014] The prediction module is configured to determine a data feature vector according to the current meter data and the historical meter data of the target smart meter, and input the data feature vector into an electricity consumption anomaly prediction model to obtain an electricity consumption anomaly prediction result, wherein the electricity consumption anomaly prediction model is a neural network model trained according to meter sample data and corresponding electricity consumption labels.

[0015] The calculation module is configured to calculate an electrical anomaly score, a magnetic field interference score, a behavior pattern score and a line loss anomaly score according to the current meter data and the historical meter data, if the electricity consumption anomaly prediction result is a probability of normal electricity consumption that is lower than a preset probability value.

[0016] The determination module is configured to determine a result of electricity stealing behavior detection of the target smart meter according to the electrical anomaly score, the magnetic field interference score, the behavior pattern score and the line loss anomaly score.

[0017] The verification module is configured to send, if the result of electricity stealing behavior detection is a suspected electricity stealing behavior, electricity stealing behavior verification request information to a concentrator, so that the concentrator calculates an electricity stealing behavior detection score according to meter data of other smart meters in a region where the target smart meter is located.

[0018] The verification module is further configured to verify whether the suspected electricity stealing behavior exists according to the electricity stealing behavior detection score.

[0019] In a third aspect, the embodiments of the present application further provide an electronic device, which comprises a processor, a memory and a bus, the memory stores machine readable instructions executable by the processor, when the electronic device is running, the processor and the memory communicate through the bus, and the machine readable instructions are executed by the processor to perform the steps of the electricity stealing detection method of the smart meter in the first aspect.

[0020] In a fourth aspect, the embodiments of the present application further provide a computer readable storage medium, which stores a computer program, and the computer program is executed by the processor to perform the steps of the electricity stealing detection method of the smart meter in the first aspect.

[0021] The method and system for stealing electricity detection of the smart meter provided by the embodiment of the application first acquire current meter data and historical meter data of a target smart meter, the meter data including meter electrical parameters, meter power consumption data, Hall magnetic sensor data and meter state data; then determine a data feature vector according to the current meter data and the historical meter data of the target smart meter, and input the data feature vector into a power consumption anomaly prediction model to obtain a power consumption anomaly prediction result; the power consumption anomaly prediction model is a neural network model trained according to meter sample data and corresponding power consumption labels; if the probability of the power consumption anomaly prediction result being normal power consumption is lower than a preset probability value, then calculate electrical anomaly scores, magnetic field interference scores, behavior pattern scores and line loss anomaly scores according to the current meter data and the historical meter data; then determine a stealing electricity behavior detection result of the target smart meter according to the electrical anomaly scores, the magnetic field interference scores, the behavior pattern scores and the line loss anomaly scores; if the stealing electricity behavior detection result is a suspected stealing electricity behavior, the target smart meter sends a stealing electricity behavior verification request information to a concentrator, so that the concentrator calculates a stealing electricity behavior detection score according to meter data of other smart meters in the area where the target smart meter is located, and finally verifies whether the suspected stealing electricity behavior exists according to the stealing electricity behavior detection score. Compared with the prior art which cannot quickly locate stealing electricity, the smart meter in the application calculates various anomaly scores based on meter data, and then determines a stealing electricity behavior detection result according to the various anomaly scores; if the stealing electricity behavior detection result cannot directly determine the corresponding stealing electricity behavior, the concentrator further determines whether the smart meter has a stealing electricity behavior, so that the application realizes automatic detection of the stealing electricity behavior and improves the detection efficiency and accuracy of the stealing electricity behavior.

[0022] In order to make the above objectives, features and advantages of the application more apparent, the following will describe a preferred embodiment in detail, and the accompanying drawings will be described as follows. BRIEF DESCRIPTION OF DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the application, and therefore should not be considered as a limitation to the scope. For those skilled in the art, other related drawings can also be obtained without creative labor.

[0024] Figure 1 A flowchart of a method for stealing electricity detection of a smart meter provided by the embodiment of the application is shown;

[0025] Figure 2 A structural block diagram of a system for stealing electricity detection of a smart meter provided by the embodiment of the application is shown;

[0026] Figure 3 A schematic diagram of a computer device is shown. DETAILED DESCRIPTION

[0027] The terms "first", "second", and "third" and the like in the specification and claims of the application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the terms so used are interchangeable under appropriate circumstances and embodiments.

[0028] In the embodiments of the present application, the words "exemplary" and "for example" are used to mean "an example of" or "an example, only. Any implementation or design solution described as "exemplary" or "for example" in the embodiments of the present application should not be construed as being preferred or advantageous over other implementations or design solutions. In fact, any implementation or design solution described as "exemplary" or "for example" is intended to present concepts in a concrete manner to aid in understanding the present application.

[0029] In the description of the present application, unless otherwise specified, " / " means that the objects associated in front and back are in a "or" relationship, for example, A / B can represent A or B; "and / or" in the present application is only a description of the association of the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent: A alone, A and B exist at the same time, and B alone, where A and B can be singular or plural. And in the description of the present application, unless otherwise specified, "multiple" means two or more than two. "At least one of the following" or the like means any combination of the items, including any combination of single item or multiple items. For example, at least one of a, b, or c, can represent: a, b, c, a-b, a-c, b-c, or a-b-c, where a, b, and c can be single or multiple.

[0030] In the embodiments of the present application, at least one can also be described as one or more, and multiple can be two, three, four or more, which is not limited by the present application.

[0031] As Figure 1 As shown in the figure, the present application provides a method for detecting electricity stealing of a smart meter, which is applied to a smart meter in a smart meter electricity stealing detection system, the smart meter is connected with a concentrator in the smart meter electricity stealing detection system, and the method for detecting electricity stealing of the smart meter provided by the present application can include:

[0032] S10, obtaining the current meter data and the historical meter data of the target smart meter.

[0033] The meter data includes meter electrical parameters, meter power consumption data, Hall magnetic sensor data, and meter state data. Specifically, the meter electrical parameters can include voltage, current, active power, reactive power, apparent power, power factor, frequency, phase angle, harmonic content, etc.; the meter power consumption data can include cumulative power consumption, load curve, event record (such as voltage loss, current loss, open phase, voltage reverse phase sequence, current reverse polarity, etc.), fee control data (rate, remaining amount, power purchase record, etc.); the Hall magnetic sensor data can include magnetic field strength, magnetic field change rate, magnetic field direction, etc., and the specific form of the meter data is not limited in the embodiment.

[0034] It should be noted that if the meter electrical parameters are abnormal (such as current imbalance, power factor abnormality) and the magnetic field interference is detected at the same time, the electricity stealing possibility is high; if the power consumption data is abnormal (such as sudden decrease in power consumption) and the electrical parameters are abnormal, but there is no magnetic field interference, it may be bypass electricity stealing; and only the magnetic field interference and the normal electrical parameters, it may be that the electricity stealing is not successful or is being attempted. Therefore, in order to comprehensively and accurately detect the electricity stealing behavior, the current meter data and the historical meter data of the target smart meter are obtained, so as to comprehensively analyze and detect the electricity stealing behavior according to the current meter data and the historical meter data of the target smart meter in the subsequent steps, thereby improving the detection accuracy of the electricity stealing behavior.

[0035] S20, determining a data feature vector according to the current meter data and the historical meter data of the target smart meter, and inputting the data feature vector into an electricity consumption anomaly prediction model to obtain an electricity consumption anomaly prediction result.

[0036] The electricity consumption anomaly prediction model is a neural network model trained according to meter sample data and corresponding electricity consumption labels. After obtaining the data feature vector, the data feature vector is input into the electricity consumption prediction model, and the electricity consumption anomaly prediction result can be obtained. The electricity consumption anomaly prediction result includes normal electricity consumption and abnormal electricity consumption, i.e., the electricity consumption anomaly prediction model is a binary classification model, and the result obtained by the model is normal electricity consumption, abnormal electricity consumption, and their respective probability values.

[0037] S30, if the probability of normal electricity consumption is lower than a preset probability value, calculating electrical anomaly scores, magnetic field interference scores, behavior pattern scores, and line loss anomaly scores according to the current meter data and the historical meter data.

[0038] The preset probability value can be 60%, 36%, etc., i.e., if the probability of normal electricity consumption is lower than 60%, the electricity consumption is suspicious, and the electrical anomaly scores, the magnetic field interference scores, the behavior pattern scores, and the line loss anomaly scores are calculated according to the current meter data and the historical meter data, so as to further analyze and determine whether the target smart meter has electricity stealing behavior.

[0039] In an optional embodiment provided in the present application, the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score are calculated according to the current electric meter data and the historical electric meter data, including:

[0040] S301. An electrical abnormality score is calculated according to an electric meter electrical parameter in the current electric meter data and the historical electric meter data.

[0041] Since the electricity stealing behavior can cause the abnormality of the electrical parameter, for example, the current shunting can make the current measurement value smaller, the voltage tampering can make the voltage measurement value abnormal, and the power factor abnormality can indicate the phase tampering, etc. Therefore, the electrical abnormality score needs to be calculated, which is used to quantify the abnormality degree of the electric meter electrical parameter, i.e., the electrical abnormality score is used to represent whether the electric meter electrical parameter deviates from the normal range; including the deviation range degree of the voltage, current, power, and power factor parameters from the normal value.

[0042] In the present embodiment, the relative deviation degrees of each parameter in the electric meter electrical parameter can be calculated first, such as the voltage deviation degree, the current deviation degree, the power deviation degree, and the power factor deviation degree. The calculation method of the deviation degree of each parameter is that the parameter value corresponding to the current electric meter data is subtracted from the absolute value of the average value of the corresponding parameter in the historical electric meter data divided by the average value of the corresponding parameter in the historical electric meter data. Then, the relative deviation degrees of each parameter calculated are weighted to obtain the electrical abnormality score.

[0043] S302. A magnetic field interference score is calculated according to the Hall magnetic sensor data in the current electric meter data and the historical electric meter data.

[0044] It needs to be noted that the strong magnet close to the electric meter can interfere with the magnetic field inside the electric meter, resulting in inaccurate measurement, and therefore the magnetic field interference score is direct evidence of the electricity stealing behavior (strong magnet electricity stealing). The magnetic field interference score is used to represent the interference degree of the electric meter by the external magnetic field, and the deviation degree of the magnetic field strength measured by the Hall magnetic sensor from the normal geomagnetic field.

[0045] Specifically, the embodiment can first calculate the magnetic field intensity anomaly degree, the magnetic field direction anomaly degree, and the magnetic field stability anomaly degree, and then calculate the magnetic field interference score according to the magnetic field intensity anomaly degree, the magnetic field direction anomaly degree, and the magnetic field stability anomaly degree. The magnetic field intensity anomaly degree is calculated according to the current three-dimensional magnetic field intensity measurement value and the historical reference three-dimensional magnetic field intensity. The magnetic field direction anomaly degree is obtained by calculating the cosine similarity of the current magnetic field vector and the reference magnetic field vector. The closer the cosine similarity value is to 0, the more consistent the direction is, and the closer the cosine similarity value is to 1, the greater the difference in direction is. The magnetic field stability anomaly degree is calculated by the current magnetic field change rate and the reference magnetic field change rate. In the embodiment, the reference three-dimensional magnetic field intensity, the reference magnetic field vector, and the reference magnetic field vector are all average values calculated by the Hall magnetic sensor in the historical electric meter data.

[0046] S303, calculating a behavior mode score according to the current electric meter data and the electric meter power consumption data in the historical electric meter data.

[0047] The behavior mode score is used to represent the consistency of the user's power consumption behavior with the historical mode. Stealing electricity behavior often leads to abnormal decrease in power consumption or abnormal shape of load curve. The behavior mode score can capture such abnormal changes. The behavior mode score is used to quantify the deviation of the user's power consumption behavior from the historical mode, including changes in power consumption, load curve shape, etc.

[0048] Specifically, the embodiment can calculate the load curve shape similarity and the power consumption mutation degree according to the current electric meter data and the electric meter power consumption data in the historical electric meter data, and then calculate the behavior mode score by weighting the load curve shape similarity and the power consumption mutation degree. The power consumption mutation degree is obtained by dividing the absolute value of the difference between the current daily total power consumption and the historical same period daily average power consumption by the historical same period daily average power consumption.

[0049] S304, calculating a line loss anomaly score according to the current electric meter data and the electric meter state data in the historical electric meter data.

[0050] Since stealing electricity behavior will lead to an increase in actual line loss, the line loss anomaly score can reflect such changes. Therefore, the embodiment needs to calculate the line loss anomaly score, which is used to represent the degree of power line loss, i.e., the line loss anomaly score is used to quantify the abnormality of the power distribution line loss, which can be obtained by comparing the difference between the theoretical line loss and the actual line loss.

[0051] In the embodiment, firstly, the deviation of the current line loss rate and the theoretical line loss rate, the deviation of the current line loss rate and the historical average line loss rate are calculated, and then the line loss abnormal score is calculated according to the weighted deviation of the current line loss rate and the theoretical line loss rate, and the deviation of the current line loss rate and the historical average line loss rate. Wherein, the current line loss rate is obtained by calculating the sum of the total table active power and the active power of all user meters divided by the total table active power.

[0052] S40, according to the electrical abnormal score, the magnetic field interference score, the behavior mode score and the line loss abnormal score, the electricity stealing behavior detection result of the target smart meter is determined.

[0053] In the embodiment, after the electrical abnormal score, the magnetic field interference score, the behavior mode score and the line loss abnormal score are calculated and normalized, the weighted average method can be used to calculate the comprehensive abnormal score, and then the electricity stealing behavior detection result is determined according to the comprehensive abnormal score. Specifically, the comprehensive abnormal score can be matched with the score range of the corresponding detection result to determine the corresponding electricity stealing behavior detection result. Wherein, the weight can be obtained according to the expert experience or the historical data training, such as the magnetic field interference score is more critical in the electricity stealing behavior, and the weight value can be set higher. Wherein, the electricity stealing behavior detection result can include three kinds: suspected electricity stealing behavior, electricity stealing behavior and normal behavior.

[0054] In an optional embodiment provided in the application, the electricity stealing behavior detection result of the target smart meter is determined according to the electrical abnormal score, the magnetic field interference score, the behavior mode score and the line loss abnormal score, comprising:

[0055] S401, the reliability factor corresponding to the electrical abnormal score, the magnetic field interference score, the behavior mode score and the line loss abnormal score is determined.

[0056] Wherein, the reliability factor is used to represent the stability, consistency and predictability of the score in the historical data. Specifically, the reliability factor corresponding to the electrical abnormal score, the magnetic field interference score, the behavior mode score and the line loss abnormal score is determined, comprising:

[0057] S4011, the historical score sequence corresponding to the electrical abnormal score, the magnetic field interference score, the behavior mode score and the line loss abnormal score is obtained.

[0058] Wherein, the historical score sequence in the embodiment can be represented as , The abnormal score at time t is represented, and N is the sequence length.

[0059] S4012, the stability score, the consistency score and the predictability score are calculated according to the historical score sequence.

[0060] The stability score is used to measure the fluctuation degree of the scores in the historical score sequence, based on the smoothness of the adjacent changes. The calculation process can be: first, calculate the adjacent change amount , Then calculate the average change amount according to the adjacent change amount, and then obtain the stability score by calculating 1 minus the average change amount. The greater the stability score, the more stable it is. In this embodiment, if the sequence is smooth and the fluctuation is small, the average change amount is small, the stability score is close to 1, indicating high stability; if the sequence fluctuates greatly, the stability score is large, and the stability score is close to 0, indicating low stability. The short-term noise level of the sequence can be reflected through the stability score, and the sequence with high stability is more reliable, and high stability indicates that the sequence is less affected by random noise, so the stability score in this embodiment can reduce false alarms caused by fluctuations in electricity stealing detection.

[0061] The consistency score is used to measure whether the scores in the historical score sequence conform to the distribution, that is, to measure the concentration of the sequence values around the central value, based on the average absolute deviation. The calculation process can be: first, calculate the sequence mean, then calculate the average absolute deviation (i.e. the average of the cumulative sum of all time abnormal scores and the sequence mean) according to the sequence mean, and then obtain the consistency score by calculating 1 minus the average absolute deviation multiplied by the preset multiple. The preset multiple is 2, which is used to adjust the range of the average absolute deviation (the average absolute deviation value range is [0, 0.5]), to ensure that the value range of the consistency score is [0, 1]. The average absolute deviation represents the average degree of deviation of the sequence values from the mean. If the sequence values are closely around the sequence mean, the average absolute deviation is small, the consistency score is close to 1, indicating high consistency. High consistency indicates that the behavior of the electricity meter is stable, and the abnormal scores are concentrated, which helps to distinguish between normal and abnormal states.

[0062] The predictability score is used to measure whether the historical score sequence is easy to predict, which can be calculated based on the first-order autocorrelation coefficient. Specifically, first calculate the lag covariance and sequence variance, then calculate the absolute value of the ratio of the lag covariance and the sequence variance to obtain the predictability score, and the value range of the predictability score is [0, 1], and the greater the value, the stronger the predictability. The lag covariance is calculated by the following formula:

[0063]

[0064] wherein, is the sequence mean. In this embodiment, the ratio of the lag covariance and the sequence variance can measure the linear relationship between the current value and the previous value of the sequence. If the predictability score is close to 1, it indicates that the sequence is highly predictable (strong positive or negative correlation); if the predictability score is close to 0, it indicates that it is random and unpredictable.

[0065] S4013, determining an electrical abnormality score, a magnetic field interference score, a behavior pattern score, a line loss abnormality score, respectively corresponding to a reliability factor, by the stability score, the consistency score, and the predictability score.

[0066] wherein the reliability factor corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score is obtained by calculating the average of the stability score, the consistency score, the predictability score of each sequence.

[0067] S402, determining a corresponding weight value according to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score and the reliability factor corresponding thereto.

[0068] Specifically, the determination of the corresponding weight value according to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score and the reliability factor corresponding thereto comprises: calculating a first weight value corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score according to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score; calculating a second weight value corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score according to the reliability factor corresponding thereto; and calculating the weight value corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score according to the first weight value and the second weight value. The first weight value and the second weight value are averaged to obtain the weight value corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score.

[0069] In this embodiment, the weighted sum value of the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score and the corresponding reliability factor is calculated, and the first weight value is determined by the ratio of the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score to the weighted sum value. In this embodiment, the abnormality score is multiplied by the reliability factor, so that the index with high abnormality score and high reliability obtains a higher weight.

[0070] The reliability factor corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score is The calculation formula of the second weight value is i represents E, M, B, L.

[0071] S403, calculate an initial score of the electricity stealing behavior by the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score, and the respective corresponding weight values.

[0072] In this embodiment, the calculation of the initial score of the electricity stealing behavior by the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score, and the respective corresponding weight values comprises: weighting calculation of the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score, and the respective corresponding weight values to obtain the initial score of the electricity stealing behavior.

[0073] S404, determine the electricity stealing behavior detection result of the target smart meter according to the initial score of the electricity stealing behavior.

[0074] Specifically, if the initial score of the electricity stealing behavior is greater than a certain value, such as greater than or equal to 0.75, it is determined that the target smart meter has the electricity stealing behavior, if the initial score of the electricity stealing behavior is less than or equal to 0.3, it is determined that the target smart meter does not have the electricity stealing behavior, and if 0.3 < initial score of the electricity stealing behavior < 0.75, it is determined that the target smart meter has the suspected electricity stealing behavior.

[0075] S50, if the electricity stealing behavior detection result is the suspected electricity stealing behavior, the target smart meter sends electricity stealing behavior verification request information to the concentrator, so that the concentrator calculates an electricity stealing behavior detection score according to the meter data of other smart meters in the area where the target smart meter is located.

[0076] In an optional embodiment provided in the application, the calculation of the electricity stealing behavior detection score by the concentrator according to the meter data of other smart meters in the area where the target smart meter is located comprises: calculation of the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score, and the respective corresponding abnormality mean value and abnormality standard deviation by the concentrator according to the meter data of other smart meters in the area where the target smart meter is located; calculation of the respective corresponding weighting deviation of the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score of the target smart meter according to the abnormality mean value and the abnormality standard deviation; and calculation of the electricity stealing behavior detection score according to the weighting deviation.

[0077] Specifically, the weighting deviation is calculated by the following formula:

[0078]

[0079] wherein, is the weighting deviation corresponding to the i-th type score, i is E, M, B, and L, which respectively represent the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score, is the i-th type score of the target smart meter. is the abnormal mean value of the i-th type score of the region where the target smart meter is located, is the standard deviation of the i-th type score of the region where the target smart meter is located, is a constant coefficient, which is a very small positive number, used to avoid the denominator being zero.

[0080] The embodiment quantifies the abnormality degree of the target smart meter by calculating the standardized positive deviation of the abnormal score of the target smart meter relative to the average level of the region. The overall level of different regions can be adapted, avoiding the inadaptability of the global fixed threshold, and the natural fluctuation in the region is considered through standardization.

[0081] In another optional embodiment provided in the application, the concentrator calculates the electricity stealing behavior detection score according to the meter data of other smart meters in the region where the target smart meter is located, comprising: the concentrator calculates the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score in the region according to the meter data of other smart meters in the region where the target smart meter is located; calculates the average value of the region electricity stealing behavior according to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score in the region calculated according to the meter data of other smart meters in the region where the target smart meter is located; and calculates the electricity stealing behavior detection score by the average value of the region electricity stealing behavior and the electricity stealing behavior detection result.

[0082] S60, verifying whether the suspected electricity stealing behavior exists electricity stealing behavior according to the electricity stealing behavior detection score.

[0083] This application provides a method for detecting electricity theft using a smart meter. First, it acquires the current and historical data of the target smart meter, including electrical parameters, electricity consumption data, Hall effect sensor data, and meter status data. Then, it determines a data feature vector based on the current and historical data and inputs this feature vector into an electricity consumption anomaly prediction model to obtain an anomaly prediction result. The electricity consumption anomaly prediction model is a neural network model trained based on meter sample data and corresponding electricity consumption tags. If the probability of the predicted electricity consumption anomaly is lower than a preset probability... The system calculates electrical anomaly scores, magnetic field interference scores, behavioral pattern scores, and line loss anomaly scores based on current and historical meter data. These scores are then used to determine the target smart meter's electricity theft detection result. If the detection result indicates suspected electricity theft, the target smart meter sends a verification request to the concentrator. The concentrator then calculates the electricity theft detection score based on data from other smart meters in the target smart meter's area. Finally, the detection score is used to verify whether the suspected electricity theft actually occurs. Compared to existing technologies that cannot quickly locate electricity theft, this application's smart meter calculates various anomaly scores based on meter data. If the initial detection result based on these scores cannot directly determine the corresponding electricity theft, the concentrator further determines whether the smart meter is involved in electricity theft. This application achieves automatic detection of electricity theft, improving detection efficiency and accuracy.

[0084] When dividing each function into modules according to its corresponding function. Figure 2 This diagram illustrates a possible composition of the smart meter electricity theft detection system described above and in the embodiments, as shown below. Figure 2 As shown, the electricity theft detection system of this smart meter may include:

[0085] The acquisition module 21 is used to acquire the current meter data and historical meter data of the target smart meter. The meter data includes meter electrical parameters, meter electricity consumption data, Hall magnetic sensor data, and meter status data.

[0086] Prediction module 23 is used to determine a data feature vector based on the current and historical data of the target smart meter, and input the data feature vector into the electricity consumption anomaly prediction model to obtain the electricity consumption anomaly prediction result; the electricity consumption anomaly prediction model is a neural network model trained based on meter sample data and corresponding electricity consumption tags;

[0087] The computing module 23 is configured to calculate an electrical abnormality score, a magnetic field interference score, a behavior pattern score and a line loss abnormality score according to the current meter data and the historical meter data if the probability of the electricity use abnormality prediction result being normal electricity use is lower than the preset probability value.

[0088] The determining module 24 is configured to determine a stealing electricity behavior detection result of the target smart meter according to the electrical abnormality score, the magnetic field interference score, the behavior pattern score and the line loss abnormality score.

[0089] The verifying module 25 is configured to send a stealing electricity behavior verification request information to the concentrator if the stealing electricity behavior detection result is a suspected stealing electricity behavior, so that the concentrator calculates a stealing electricity behavior detection score according to meter data of other smart meters in a region where the target smart meter is located.

[0090] The verifying module 25 is further configured to verify whether the suspected stealing electricity behavior exists according to the stealing electricity behavior detection score.

[0091] In an optional embodiment of the present application, the computing module 23 is specifically configured to:

[0092] calculate an electrical abnormality score according to electrical parameters of the meter in the current meter data and the historical meter data; the electrical abnormality score is used to indicate whether the electrical parameters of the meter deviate from a normal range;

[0093] calculate a magnetic field interference score according to Hall magnetic sensor data in the current meter data and the historical meter data; the magnetic field interference score is used to indicate a degree of interference of the meter by an external magnetic field;

[0094] calculate a behavior pattern score according to meter electricity use data in the current meter data and the historical meter data; the behavior pattern score is used to indicate a consistency degree of the user's electricity use behavior with a historical pattern;

[0095] calculate a line loss abnormality score according to meter state data in the current meter data and the historical meter data; the line loss abnormality score is used to indicate a degree of power line loss.

[0096] In an optional embodiment of the present application, the determining module 24 is specifically configured to:

[0097] determine a reliability factor corresponding to each of the electrical abnormality score, the magnetic field interference score, the behavior pattern score and the line loss abnormality score; the reliability factor is used to indicate stability, consistency and predictability of the score in historical data;

[0098] determine a corresponding weight value according to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score and the reliability factor corresponding thereto respectively;

[0099] calculate an electricity stealing behavior initial score according to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score and the weight value corresponding thereto respectively;

[0100] determine the electricity stealing behavior detection result of the target smart meter according to the electricity stealing behavior initial score.

[0101] In an optional embodiment provided by the present application, the determining module 24 is specifically configured to:

[0102] obtain a historical score sequence corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score and the line loss abnormality score respectively;

[0103] calculate a stability score, a consistency score and a predictability score according to the historical score sequence, the stability score being used to measure the fluctuation degree of the scores in the historical score sequence, the consistency score being used to measure whether the scores in the historical score sequence conform to a distribution, and the predictability score being used to measure whether the historical score sequence is easy to predict;

[0104] determine the reliability factor corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score and the line loss abnormality score according to the stability score, the consistency score and the predictability score.

[0105] In an optional embodiment provided by the present application, the determining module 24 is specifically configured to:

[0106] calculate a first weight value corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score and the line loss abnormality score according to the electrical abnormality score, the magnetic field interference score, the behavior pattern score and the line loss abnormality score;

[0107] calculate a second weight value corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score and the line loss abnormality score according to the reliability factor corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score and the line loss abnormality score respectively;

[0108] calculate a weight value corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score and the line loss abnormality score according to the first weight value and the second weight value.

[0109] In an optional embodiment provided by the present application, the determining module 24 is specifically configured to:

[0110] The electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score, and the respective corresponding weight values are weighted to obtain the initial score of the electricity stealing behavior.

[0111] In an optional embodiment provided by the present application, the verification module 25 is specifically configured to:

[0112] The concentrator calculates the abnormality mean value and the abnormality standard deviation of the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score in the region according to the meter data of the other smart meters in the region where the target smart meter is located.

[0113] The concentrator calculates the weighted deviation of the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score of the target smart meter according to the abnormality mean value and the abnormality standard deviation.

[0114] The concentrator calculates the electricity stealing behavior detection score according to the weighted deviation.

[0115] In an optional embodiment provided by the present application, the verification module 25 is specifically configured to:

[0116] The concentrator calculates the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score according to the meter data of the other smart meters in the region where the target smart meter is located.

[0117] The concentrator calculates the average value of the electricity stealing behavior in the region according to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score calculated according to the meter data of the other smart meters in the region where the target smart meter is located.

[0118] The concentrator calculates the electricity stealing behavior detection score according to the average value of the electricity stealing behavior in the region and the electricity stealing behavior detection result.

[0119] In an optional embodiment provided by the present application, the weighted deviation is calculated by the following formula:

[0120]

[0121] wherein, is the weighted deviation corresponding to the i-th type score, i is E, M, B, and L, which respectively represent the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score, is the i-th type score of the target smart meter, is the abnormality mean value of the i-th type score in the region, is the standard deviation of the i-th type score in the region, is a constant coefficient.

[0122] The specific limitations of the system can refer to the above limitations of the electricity stealing detection method of the smart meter, which will not be repeated here. Each module in the above device can be implemented by software, hardware and their combination in whole or in part. The above modules can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory in the computer device in software form, so that the processor calls and executes the operations corresponding to the above modules.

[0123] In one embodiment, a computer device, which can be a server, is provided, and its internal structure diagram can be as shown in Figure 3 The computer device includes a processor, a memory, a network interface and a database connected by a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with external terminals through network connection. The computer program is executed by the processor to implement an electricity stealing detection method for a smart meter.

[0124] In one embodiment, a computer device is provided, including a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements the following steps when executing the computer program:

[0125] obtaining current meter data and historical meter data of a target smart meter, the meter data including meter electrical parameters, meter power consumption data, Hall magnetic sensor data and meter state data;

[0126] determining a data feature vector according to the current meter data and the historical meter data of the target smart meter, and inputting the data feature vector into a power consumption anomaly prediction model to obtain a power consumption anomaly prediction result; the power consumption anomaly prediction model is a neural network model trained according to meter sample data and corresponding power consumption labels;

[0127] if the probability of the power consumption anomaly prediction result being normal power consumption is lower than a preset probability value, calculating electrical anomaly scores, magnetic field interference scores, behavior pattern scores and line loss anomaly scores according to the current meter data and the historical meter data;

[0128] determining a power stealing behavior detection result of the target smart meter according to the electrical anomaly scores, the magnetic field interference scores, the behavior pattern scores and the line loss anomaly scores;

[0129] If the electricity stealing behavior detection result is a suspected electricity stealing behavior, the target smart meter sends electricity stealing behavior verification request information to the concentrator, so that the concentrator calculates an electricity stealing behavior detection score according to the meter data of other smart meters in the area where the target smart meter is located;

[0130] According to the electricity stealing behavior detection score, it is verified whether the suspected electricity stealing behavior exists.

[0131] In one embodiment, a computer readable storage medium is provided, and a computer program is stored on the computer readable storage medium. The computer program is executed by a processor to implement the following steps:

[0132] Obtaining current meter data and historical meter data of a target smart meter, the meter data including meter electrical parameters, meter power consumption data, Hall magnetic sensor data, and meter state data;

[0133] According to the current meter data and the historical meter data of the target smart meter, a data feature vector is determined, and the data feature vector is input into a power consumption anomaly prediction model to obtain a power consumption anomaly prediction result. The power consumption anomaly prediction model is a neural network model trained according to meter sample data and corresponding power consumption labels;

[0134] If the power consumption anomaly prediction result is that the probability of normal power consumption is lower than a preset probability value, then an electrical anomaly score, a magnetic field interference score, a behavior pattern score, and a line loss anomaly score are calculated according to the current meter data and the historical meter data;

[0135] According to the electrical anomaly score, the magnetic field interference score, the behavior pattern score, and the line loss anomaly score, a stealing behavior detection result of the target smart meter is determined;

[0136] If the electricity stealing behavior detection result is a suspected electricity stealing behavior, the target smart meter sends electricity stealing behavior verification request information to the concentrator, so that the concentrator calculates an electricity stealing behavior detection score according to the meter data of other smart meters in the area where the target smart meter is located;

[0137] According to the electricity stealing behavior detection score, it is verified whether the suspected electricity stealing behavior exists.

[0138] In one embodiment, a computer program product is provided, and the computer program product includes a computer program. The computer program is executed by a processor to implement the following steps:

[0139] Obtaining current meter data and historical meter data of a target smart meter, the meter data including meter electrical parameters, meter power consumption data, Hall magnetic sensor data, and meter state data;

[0140] determine a data feature vector according to current meter data and historical meter data of the target smart meter, and input the data feature vector into a power consumption anomaly prediction model to obtain a power consumption anomaly prediction result; the power consumption anomaly prediction model is a neural network model trained according to meter sample data and corresponding power consumption labels;

[0141] if the power consumption anomaly prediction result is a probability of normal power consumption lower than a preset probability value, calculate an electrical anomaly score, a magnetic field interference score, a behavior pattern score, and a line loss anomaly score according to the current meter data and the historical meter data;

[0142] determine a power stealing behavior detection result of the target smart meter according to the electrical anomaly score, the magnetic field interference score, the behavior pattern score, and the line loss anomaly score;

[0143] if the power stealing behavior detection result is a suspected power stealing behavior, the target smart meter sends a power stealing behavior verification request information to a concentrator, so that the concentrator calculates a power stealing behavior detection score according to meter data of other smart meters in a region where the target smart meter is located;

[0144] verify whether the suspected power stealing behavior exists according to the power stealing behavior detection score.

[0145] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware, and the computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments. Any reference to memory, storage, database or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0146] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the system is divided into different functional units or modules to complete all or part of the functions described above.

[0147] The above examples are only used to illustrate the technical solutions of the present application, but not to limit it; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that the technical solutions recorded in the foregoing examples can still be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for detecting electricity theft of a smart meter, characterized in that, The method is applied to a smart meter in a power stealing detection system of the smart meter, the smart meter is connected with a concentrator in the power stealing detection system of the smart meter, and the method comprises: obtaining current meter data and historical meter data of a target smart meter, the meter data comprising meter electrical parameters, meter power consumption data, Hall magnetic sensor data and meter state data; determining a data feature vector according to the current meter data and the historical meter data of the target smart meter, and inputting the data feature vector into a power consumption anomaly prediction model to obtain a power consumption anomaly prediction result; the power consumption anomaly prediction model is a neural network model trained according to meter sample data and corresponding power consumption labels; if the probability of the power consumption anomaly prediction result being normal power consumption is lower than a preset probability value, calculating electrical anomaly scores, magnetic field interference scores, behavior pattern scores and line loss anomaly scores according to the current meter data and the historical meter data; determining a power stealing behavior detection result of the target smart meter according to the electrical anomaly scores, the magnetic field interference scores, the behavior pattern scores and the line loss anomaly scores; if the power stealing behavior detection result is a suspected power stealing behavior, the target smart meter sends a power stealing behavior verification request information to the concentrator, so that the concentrator calculates a power stealing behavior detection score according to meter data of other smart meters in the area where the target smart meter is located; verifying whether the suspected power stealing behavior exists according to the power stealing behavior detection score.

2. The method of claim 1, wherein, The method comprises: calculating electrical anomaly scores according to meter electrical parameters in the current meter data and the historical meter data; the electrical anomaly scores are used to indicate whether the meter electrical parameters deviate from the normal range; calculating magnetic field interference scores according to Hall magnetic sensor data in the current meter data and the historical meter data; the magnetic field interference scores are used to indicate the degree of interference of the meter by external magnetic fields; calculating behavior pattern scores according to meter power consumption data in the current meter data and the historical meter data; the behavior pattern scores are used to indicate the consistency of the user's power consumption behavior with the historical pattern; calculating line loss anomaly scores according to meter state data in the current meter data and the historical meter data; the line loss anomaly scores are used to indicate the degree of power line loss.

3. The method of claim 2, wherein, The method comprises: determining reliability factors corresponding to the electrical anomaly scores, the magnetic field interference scores, the behavior pattern scores and the line loss anomaly scores respectively; the reliability factors are used to indicate the stability, consistency and predictability of the scores in the historical data; determining corresponding weight values according to the electrical anomaly scores, the magnetic field interference scores, the behavior pattern scores, the line loss anomaly scores and the reliability factors corresponding thereto respectively. The electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score, and the corresponding weight values thereof are used to calculate an initial electricity stealing behavior score. The electricity stealing behavior detection result of the target smart meter is determined according to the initial electricity stealing behavior score.

4. The method of claim 3, wherein, The reliability factors corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score are determined. A historical score sequence corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score is obtained. A stability score, a consistency score, and a predictability score are calculated according to the historical score sequence. The stability score is used to measure the fluctuation degree of the historical score sequence. The consistency score is used to measure whether the historical score sequence conforms to a distribution. The predictability score is used to measure whether the historical score sequence is easy to predict. The reliability factors corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score are determined according to the stability score, the consistency score, and the predictability score.

5. The method of claim 3, wherein, The corresponding weight values are determined according to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score, and the corresponding reliability factors thereof. A first weight value corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score is calculated according to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score. A second weight value corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score is calculated according to the reliability factors corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score. The weight values corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score are calculated according to the first weight value and the second weight value.

6. The method of claim 3, wherein, The initial electricity stealing behavior score is calculated by weighting the electrical abnormality score, the magnetic field interference score, the behavior pattern score, the line loss abnormality score, and the corresponding weight values thereof. The concentrator calculates an electricity stealing behavior detection score according to the meter data of other smart meters in the area where the target smart meter is located.

7. The method according to any one of claims 1 to 6, characterized in that, The concentrator calculates an abnormality mean value and an abnormality standard deviation corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score in the area according to the meter data of other smart meters in the area where the target smart meter is located. A weighted deviation degree corresponding to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score of the target smart meter is calculated according to the abnormality mean value and the abnormality standard deviation. ​ According to the weighted deviation, a stealing electricity behavior detection score is calculated.

8. The method according to any one of claims 1 to 6, characterized in that, The concentrator calculates a stealing electricity behavior detection score according to the meter data of other smart meters in the area where the target smart meter is located. The concentrator calculates an electrical abnormality score, a magnetic field interference score, a behavior pattern score, and a line loss abnormality score in the area according to the meter data of other smart meters in the area where the target smart meter is located. According to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score in the area calculated according to the meter data of other smart meters in the area where the target smart meter is located, an average value of regional stealing electricity behavior is calculated. According to the average value of regional stealing electricity behavior and the stealing electricity behavior detection result, a stealing electricity behavior detection score is calculated.

9. The method of claim 7, wherein, According to the average value of regional stealing electricity behavior and the stealing electricity behavior detection result, a stealing electricity behavior detection score is calculated. The system comprises: wherein, is the weighted deviation degree corresponding to the i-th type score, is the i-th type score of the target smart meter, i takes values of E, M, B, L, representing the electrical anomaly score, the magnetic field interference score, the behavior pattern score, and the line loss anomaly score, respectively, is the anomaly mean of the i-th type score of the region, is the standard deviation of the i-th type score of the region, is a constant coefficient.

10. A power theft detection system for a smart meter, the system comprising: An acquisition module is configured to acquire current meter data and historical meter data of a target smart meter, wherein the meter data comprises meter electrical parameters, meter power consumption data, Hall magnetic sensor data, and meter state data. A prediction module is configured to determine a data feature vector according to the current meter data and the historical meter data of the target smart meter, and input the data feature vector into a power consumption anomaly prediction model to obtain a power consumption anomaly prediction result, wherein the power consumption anomaly prediction model is a neural network model trained according to meter sample data and corresponding power consumption labels. A calculation module is configured to calculate an electrical abnormality score, a magnetic field interference score, a behavior pattern score, and a line loss abnormality score according to the current meter data and the historical meter data if a probability of normal power consumption is lower than a preset probability value. A determination module is configured to determine a stealing electricity behavior detection result of the target smart meter according to the electrical abnormality score, the magnetic field interference score, the behavior pattern score, and the line loss abnormality score. A verification module is configured to send a stealing electricity behavior verification request information to a concentrator if the stealing electricity behavior detection result is a suspected stealing electricity behavior, so that the concentrator calculates a stealing electricity behavior detection score according to the meter data of other smart meters in the area where the target smart meter is located. The verification module is further configured to verify whether the suspected stealing electricity behavior exists according to the stealing electricity behavior detection score. ​

Images