System and method for diagnosing abnormal power consumption behavior of user

Abstract

The invention discloses a system and a method for diagnosing abnormal power consumption behaviors of a user, relates to the technical field of power data analysis, and solves the problem that normal power consumption fluctuation caused by external interference factors in a region is neglected when the power consumption behaviors of the user are diagnosed. And the technical problem of inaccurate power consumption data analysis caused by neglect of potential default conditions of default users in different areas to normal users is solved. According to the method, the regional fluctuation factor is obtained by analyzing the total power consumption of the regional target; analyzing the geographic position to obtain a position influence factor; analyzing the historical reference data to obtain a power consumption factor, and calculating the power consumption factor based on the regional fluctuation factor and the position influence factor to obtain a first calculation value and a second calculation value; and establishing a behavior diagnosis model, and inputting the power utilization factor, the first calculation value and the second calculation value into the behavior diagnosis model to obtain a diagnosis result, thereby improving the accuracy of user power utilization behavior analysis.

Description

Technical Field

[0001] The present invention belongs to the field of smart grids and relates to power data analysis technology, specifically a diagnosis system and method for abnormal power consumption behavior of users. Background Art

[0002] With the development of social economy, the demand for electricity is increasing, while the supply capacity of power resources is limited by various factors. In order to meet the growing demand for electricity, power companies need to take effective measures to prevent users from abnormal electricity consumption and ensure the rational use and safe and stable supply of power resources. Abnormal electricity consumption may affect the safe and stable operation of the power grid. For example, overload may cause local power grid paralysis, and electricity theft may cause economic losses to power grid operations and affect the normal electricity consumption of the majority of users. Therefore, timely detection and handling of these abnormal situations is crucial to ensure the overall health of the power system.

[0003] The prior art (invention patent application publication number CN113988723A) discloses a user behavior locking method and system based on abnormal analysis of electricity consumption data. The method performs abnormal analysis based on the user's electricity consumption data within a predetermined time period. When there is an abnormality in the electricity consumption data, the abnormal electricity consumption data change in electricity consumption is calculated, and then the user's behavior type is determined based on the change in electricity consumption. The above scheme solves the problem of accurately locking user behavior based on abnormal electricity consumption data, but when diagnosing the user's electricity consumption behavior, it ignores the normal electricity consumption fluctuations caused by external interference factors in the region, and ignores the potential default caused by the behavior of default users in different regions on normal users, resulting in inaccurate analysis of electricity consumption data.

[0004] Therefore, the present invention discloses a system and method for diagnosing abnormal electricity usage behavior of users, which are used to solve the above problems. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a diagnosis system and method for abnormal electricity consumption behavior of users, which are used to solve the technical problem that when diagnosing the electricity consumption behavior of users, normal electricity consumption fluctuations caused by external interference factors in the region are ignored, and potential default situations caused by the behavioral driving effect of default users in different regions on normal users are ignored, resulting in inaccurate analysis of electricity consumption data. The present invention obtains a regional fluctuation factor by analyzing the total target electricity consumption in the region; obtains a location influence factor by analyzing the geographical location; obtains an electricity consumption factor by analyzing historical reference data, calculates the electricity consumption factor once based on the regional fluctuation factor to obtain a first calculated value, and calculates the electricity consumption factor twice based on the location influence factor to obtain a second calculated value; establishes a behavior diagnosis model, and inputs the electricity consumption factor, the first calculated value, and the second calculated value into the behavior diagnosis model to obtain a diagnosis result to solve the above problems.

[0006] To achieve the above-mentioned object, a first aspect of the present invention provides a user abnormal electricity consumption behavior diagnosis system, comprising: a data processing module, and a data collection module, a behavior diagnosis module and a database connected thereto;

[0007] The data collection module is used to obtain basic information, including power consumption information, current time and user's geographical location; power consumption information includes user target total power consumption and regional target total power consumption;

[0008] The data processing module is used to extract historical power consumption from the user's historical data from the database, analyze the historical power consumption to obtain historical reference data; analyze the regional target total power consumption to obtain the regional fluctuation factor; analyze the geographical location to obtain the location influence factor; obtain the power consumption factor by analyzing the historical reference data, calculate the power consumption factor once based on the regional fluctuation factor to obtain a first calculated value, and calculate the power consumption factor twice based on the location influence factor to obtain a second calculated value;

[0009] The behavior diagnosis module is used to establish a behavior diagnosis model, input the power consumption factor, the first calculated value and the second calculated value into the behavior diagnosis model to obtain a diagnosis result; wherein the diagnosis result includes normal power consumption, power theft and excess power consumption.

[0010] Preferably, the obtaining of basic information includes:

[0011] The current time is extracted from the computer, and the time period in which the current time is located is obtained, and the time period before the current time period is marked as the target time period; the user's electricity consumption in the target time period is obtained through a smart meter, and the electricity consumption is marked as the user's target total electricity consumption; the target total electricity consumption of all users in the area is added up to obtain the regional target total electricity consumption; wherein one hour is a time period, and the regional range is set manually.

[0012] It is worth noting that the purpose of setting the target time period in the present invention is to obtain the power consumption in a complete unit of time. Fixing the time can better analyze the user's power consumption behavior and improve the accuracy of the analysis.

[0013] Preferably, the analyzing of historical power consumption to obtain historical reference data includes:

[0014] A1: extract the current time and determine whether the date of the current time is a statutory holiday; if yes, obtain the historical power consumption corresponding to the statutory holiday in the target time period, mark the average value of the historical power consumption as historical power consumption 1 LD1, set the value of historical power consumption 2 LD2 to 0, and jump to A4; if no, set the value of historical power consumption 1 LD1 to 0, and jump to A2;

[0015] A2: Get the position of the current date in the current week, mark the days at the same position in the historical data as target days, get a number of historical power consumptions corresponding to the target time period in the target days, and establish a historical power consumption database based on the historical power consumptions;

[0016] A3: Obtain the variance of the historical electricity consumption in the historical electricity consumption database, and determine whether the variance is less than the variance threshold; if yes, calculate the average value of the historical electricity consumption database to obtain the historical electricity consumption 2 LD2; if no, remove the historical electricity consumption in the historical electricity consumption database that is most different from the mode of the historical electricity consumption database, and re-judge the variance until the variance of the historical electricity consumption database is less than the variance threshold, and calculate the average value of the remaining historical electricity consumption in the historical electricity consumption database to obtain the historical electricity consumption 2 LD2; wherein the variance threshold is obtained through experience;

[0017] A4: Based on the formula LCS=α1×LD1+α2×LD2, the historical reference data LCS corresponding to the current user at the current time is obtained; wherein α1 and α2 are both proportional adjustment factors greater than 0, and α1+α2=1.

[0018] It is worth noting that the present invention incorporates statutory holidays into the analysis of historical reference data. This is because users in different regions have different electricity consumption during different statutory holidays. The present invention takes these different situations into consideration to obtain a more accurate reference data, providing data support for the subsequent calculation of electricity consumption factors.

[0019] It should be noted that the electricity consumption of users in different areas is different during different statutory holidays. For example: if most of the residents living in a certain area are migrant workers, when the Spring Festival comes, these migrant workers will choose to return to their hometowns, which will cause the electricity consumption in the current area to decrease during the Spring Festival; if most of the residents living in a certain area are local residents, when the Spring Festival comes, the families of local residents will choose to return to the current area, which will cause the electricity consumption in the current area to increase during the Spring Festival.

[0020] Preferably, the analyzing of the regional target total electricity consumption to obtain the regional fluctuation factor includes:

[0021] B1: Extract the target total power consumption QMZ of the region and the number of households H corresponding to the target total power consumption of the region, and determine whether the number of households is less than the set number of households BH; if yes, jump to B2; if no, jump to B3;

[0022] B2: Based on the formula QMZP=QMZ / H, the regional average QMZP of the regional target total electricity consumption is obtained, and based on the formula QBZ=QMZP / BZP, the regional fluctuation factor QBZ is obtained; where BZP is the standard regional average value set manually;

[0023] B3: The regional fluctuation factor QBZ is obtained based on the formula QBZ=(1+ln((H-BH) / BH+1))×(QMZP / BZP).

[0024] It is worth noting that the present invention conducts an overall electricity consumption analysis on all electricity users in the area to obtain a reference value that can represent the electricity consumption fluctuation in the current area corresponding to the target time period at the current time. Through this reference value and subsequent electricity consumption factors, it can be analyzed whether the current area needs to use more or less electricity at this time due to human or natural influences, thereby avoiding errors in the analysis of users' electricity consumption behavior due to external interference factors.

[0025] Preferably, the analyzing the geographical location to obtain the location influence factor includes:

[0026] The geographical location is extracted, the number of electricity users DH in the area is extracted, the number of abnormal electricity users YDH in the area is extracted, and the location influence factor WYZ is obtained based on the formula WYZ=exp(YDH / DH); wherein abnormal electricity consumption includes electricity theft and excess electricity consumption.

[0027] It is worth noting that when the number of abnormal electricity users in the area is high, it indicates that the electricity consumption in the current area is low, and people's electricity consumption behavior will be affected by the surrounding credit environment. When there are more default electricity use cases around, it will have a behavioral driving effect on users, resulting in an increased probability of users defaulting on electricity use. The present invention calculates the probability of the number of default electricity users in the current area, and obtains a default influence factor of a default user affecting the electricity use behavior of other users, thereby increasing the accuracy of system analysis.

[0028] Preferably, obtaining the electricity consumption factor by analyzing the historical reference data includes:

[0029] The user's target total electricity consumption HML and historical reference data LCS corresponding to the current time are extracted, and the electricity consumption factor DYZ is obtained based on the formula DYZ=HML / LCS.

[0030] Preferably, the step of calculating the power consumption factor based on the regional fluctuation factor to obtain the first calculated value includes:

[0031] C1: extract the regional fluctuation factor QBZ and the electricity consumption factor DYZ, and judge whether the value of QBZ-DYZ is greater than the maximum value of the judgment range; if yes, obtain the first calculated value JZ1 based on the formula JZ1=β×ln((DYZ-QBZ) / QBZ+1); if no, jump to C2;

[0032] C2: Determine whether the value of QBZ-DYZ is less than the maximum value of the judgment range; if yes, obtain the first calculated value JZ1 based on the formula JZ1=β×(-ln(-(DYZ-QBZ) / QBZ+1)); if no, obtain the first calculated value JZ1 based on the formula JZ1=δ×(DYZ-QBZ) / QBZ; wherein β and δ are both amplitude adjustment factors greater than 0.

[0033] It is worth noting that when obtaining the first calculated value, the present invention uses three different formulas for calculation according to the difference between the regional fluctuation factor and the electricity consumption factor. When the difference between the regional fluctuation factor and the electricity consumption factor is within the judgment range, the first calculated value changes linearly with the difference. This is because when the difference is within the judgment range, it indicates that the current user may be in normal electricity consumption. Linear change can provide better stability, which means that even in the case of large fluctuations, the first calculated value will not fluctuate greatly, thereby maintaining the stability of the analysis results; when the difference between the regional fluctuation factor and the electricity consumption factor is not within the judgment range, it indicates that the current user may be in abnormal electricity consumption. At this time, in order to ensure the safety of the system, when the difference exceeds the normal range, nonlinear calculation is required to avoid the situation where the calculation result is too large and may cause system instability.

[0034] Preferably, the second calculation value is obtained by performing a secondary calculation on the power consumption factor based on the location influence factor, including:

[0035] The location influence factor WYZ and the power consumption factor DYZ are extracted, and a second calculated value JZ2 is obtained based on the formula JZ2=μ×(1+WYZ)×DYZ; wherein μ is a proportional adjustment factor greater than 0.

[0036] Preferably, the establishing of the behavioral diagnosis model comprises:

[0037] Acquire a historical electricity consumption factor, and a first calculated value and a second calculated value corresponding to the historical electricity consumption factor through a database;

[0038] Integrate historical electricity consumption factors, and the first calculated value, the second calculated value and the diagnostic results corresponding to the historical electricity consumption factors into several groups of training data and test data; use the training data to train the artificial intelligence model; use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally obtain a behavioral diagnosis model with the electricity consumption factor, the first calculated value and the second calculated value as input and the diagnostic result as output; wherein the artificial intelligence model includes a BP neural network model or an RBF neural network model.

[0039] A second aspect of the present invention provides a method for diagnosing abnormal power usage behavior of a user, comprising the following steps:

[0040] S1: Obtain basic information;

[0041] S2: extracting historical electricity consumption from the user's historical data from the database, analyzing the historical electricity consumption to obtain historical reference data; analyzing the regional target total electricity consumption to obtain the regional fluctuation factor; analyzing the geographical location to obtain the location influence factor; obtaining the electricity consumption factor by analyzing the historical reference data, calculating the electricity consumption factor once based on the regional fluctuation factor to obtain a first calculated value, and calculating the electricity consumption factor twice based on the location influence factor to obtain a second calculated value;

[0042] S3: Establish a behavior diagnosis model, input the power consumption factor, the first calculated value and the second calculated value into the behavior diagnosis model to obtain a diagnosis result.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. The present invention obtains a regional fluctuation factor by analyzing the target total power consumption of the region; obtains a location influence factor by analyzing the geographical location; obtains a power factor by analyzing historical reference data, calculates the power factor once based on the regional fluctuation factor to obtain a first calculated value, and calculates the power factor twice based on the location influence factor to obtain a second calculated value; establishes a behavior diagnosis model, inputs the power factor, the first calculated value, and the second calculated value into the behavior diagnosis model to obtain a diagnosis result, thereby solving the technical problem that when diagnosing the power consumption behavior of users, the normal power consumption fluctuation caused by external interference factors in the region is ignored, and the potential default caused by the behavior driving effect of default users in different regions on normal users is ignored, resulting in inaccurate power consumption data analysis. The present invention provides a more sophisticated and comprehensive power consumption behavior analysis method, which helps power companies better understand and predict users' power consumption behavior, thereby improving the efficiency and accuracy of power services.

[0045] 2. The present invention performs an overall electricity consumption analysis on all electricity users in the area to obtain a reference value that can represent the electricity consumption fluctuation in the current area corresponding to the target time period at the current time. Through this reference value and subsequent electricity consumption factors, it can be analyzed whether the current area needs to use more or less electricity at this time due to human or natural influences, thereby avoiding errors in the analysis of users' electricity consumption behavior due to external interference factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0047] Figure 1 It is a schematic diagram of the operation steps of the present invention;

[0048] Figure 2 It is a schematic diagram of the system module of the present invention;

[0049] Figure 3 A schematic diagram of the operation steps for obtaining a first calculated value according to the present invention. DETAILED DESCRIPTION

[0050] The technical scheme of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] See also Figure 1-Figure 2 , the first aspect of the present invention provides a user abnormal electricity consumption behavior diagnosis system, including: a data processing module, and a data collection module, a behavior diagnosis module and a database connected thereto;

[0052] Data collection module: used to obtain basic information; the basic information includes power consumption information, current time and user's geographical location; power consumption information includes user target total power consumption and regional target total power consumption;

[0053] Data processing module: used to extract historical power consumption from user historical data from the database, analyze the historical power consumption to obtain historical reference data; analyze the regional target total power consumption to obtain the regional fluctuation factor; analyze the geographical location to obtain the location influence factor; obtain the power consumption factor by analyzing the historical reference data, calculate the power consumption factor once based on the regional fluctuation factor to obtain a first calculated value, and calculate the power consumption factor twice based on the location influence factor to obtain a second calculated value;

[0054] Behavior diagnosis module: used to establish a behavior diagnosis model, input the power consumption factor, the first calculated value and the second calculated value into the behavior diagnosis model to obtain a diagnosis result; wherein the diagnosis result includes normal power consumption, power theft and excess power consumption.

[0055] It should be noted that the time point for data collection can be: manually set to a fixed location every hour for data collection.

[0056] Exemplarily, in this embodiment, the current time is 19:30 on September 10, 2024;

[0057] Get the current time period as 18:00-19:00;

[0058] Since September 10, 2024 is not a legal holiday, the value of historical power consumption LD1 is set to 0;

[0059] Get the position of September 10, 2024 as the second day in the current week, mark the second day of each week in the historical data as the target day, obtain several historical power consumptions corresponding to 18:00-19:00 in the target day, and establish a historical power consumption database based on the historical power consumption;

[0060] Obtain the variance of historical electricity consumption in the historical electricity consumption database, and determine whether the variance is less than the variance threshold; if yes, calculate the average value of the historical electricity consumption database to obtain the historical electricity consumption 2 LD2; if no, remove the historical electricity consumption in the historical electricity consumption database that has the largest difference from the mode of the historical electricity consumption database, and re-judge the variance until the variance of the historical electricity consumption database is less than the variance threshold, and then calculate the average value of the remaining historical electricity consumption in the historical electricity consumption database to obtain the historical electricity consumption 2 LD2;

[0061] Based on the formula LCS = α1 × LD1 + α2 × LD2, the historical reference data LCS corresponding to the current user at 19:30 is obtained;

[0062] Extract the regional target total electricity consumption QMZ and the statistical number of households H corresponding to the regional target total electricity consumption, because the statistical number of households is not less than the set number of households BH; therefore, the regional fluctuation factor QBZ is obtained based on the formula QBZ = (1 + ln ((H-BH) / BH+1)) × (QMZP / BZP);

[0063] Extract the area where the geographical location is located, extract the number of electricity users DH in the area, extract the number of abnormal electricity users YDH in the area, and obtain the location influence factor WYZ based on the formula WYZ=exp(YDH / DH);

[0064] Extract the user target total electricity consumption HML and historical reference data LCS corresponding to 19:30, and obtain the electricity consumption factor DYZ based on the formula DYZ = HML / LCS;

[0065] Extract the regional fluctuation factor QBZ and the power consumption factor DYZ. Since the value of QBZ-DYZ is greater than the maximum value of the judgment range, the first calculated value JZ1 is obtained based on the formula JZ1=β×ln((DYZ-QBZ) / QBZ+1);

[0066] Extract the location influence factor WYZ and the power consumption factor DYZ, and obtain the second calculated value JZ2 based on the formula JZ2=μ×(1+WYZ)×DYZ;

[0067] A behavior diagnosis model is established, and the power consumption factor, the first calculated value and the second calculated value are input into the behavior diagnosis model to obtain a diagnosis result.

[0068] Basic information is obtained in this application, including:

[0069] The current time is extracted from the computer, and the time period in which the current time is located is obtained, and the time period before the current time period is marked as the target time period; the user's electricity consumption in the target time period is obtained through the smart meter, and the electricity consumption is marked as the user's target total electricity consumption; the target total electricity consumption of all users in the area is added up to obtain the regional target total electricity consumption; wherein, one hour is a time period, and the regional range is set manually.

[0070] It should be noted that this application divides a day into 24 time periods.

[0071] It should be noted that the target time period is the hour before the current time, and the user's target total electricity consumption is the total electricity consumption of the current user within the hour before the current time.

[0072] It should be noted that if the current time is 0:00-1:00, the target time period of the current time is 23:00-24:00 of the previous day; for example: if the current time is 0:20 on the 6th, the target time period is 23:00-24:00 on the 5th.

[0073] In this application, historical electricity consumption is analyzed to obtain historical reference data, including:

[0074] A1: Extract the current time and determine whether the date of the current time is a statutory holiday; if yes, obtain the historical power consumption corresponding to the statutory holiday in the target time period, mark the average value of the historical power consumption as historical power consumption 1 LD1, set the value of historical power consumption 2 LD2 to 0, and jump to A4; if no, set the value of historical power consumption 1 LD1 to 0, and jump to A2;

[0075] A2: Get the position of the current date in the current week, mark the days at the same position in the historical data as target days, get the historical power consumption corresponding to the target time period in the target days, and establish a historical power consumption database based on the historical power consumption;

[0076] A3: Obtain the variance of historical electricity consumption in the historical electricity consumption database, and determine whether the variance is less than the variance threshold; if yes, calculate the average value of the historical electricity consumption database to obtain the historical electricity consumption 2 LD2; if no, remove the historical electricity consumption in the historical electricity consumption database that is the most different from the mode of the historical electricity consumption database, and re-judge the variance until the variance of the historical electricity consumption database is less than the variance threshold, and calculate the average value of the remaining historical electricity consumption in the historical electricity consumption database to obtain the historical electricity consumption 2 LD2; wherein the variance threshold is obtained through experience;

[0077] A4: Based on the formula LCS=α1×LD1+α2×LD2, the historical reference data LCS corresponding to the current user at the current time is obtained; wherein α1 and α2 are both proportional adjustment factors greater than 0, and α1+α2=1.

[0078] It should be noted that the electricity consumption of users in different areas is different during different statutory holidays. For example: if most of the residents living in a certain area are migrant workers, when the Spring Festival comes, these migrant workers will choose to return to their hometowns, which will cause the electricity consumption in the current area to decrease during the Spring Festival; if most of the residents living in a certain area are local residents, when the Spring Festival comes, the families of local residents will choose to return to the current area, which will cause the electricity consumption in the current area to increase during the Spring Festival.

[0079] It should be noted that the specific explanation for obtaining several historical electricity consumptions corresponding to the target time period during statutory holidays is: if the target time period is 14:00-15:00, and the statutory holiday on the current date is Mid-Autumn Festival, then the historical electricity consumption in the time period of 14:00-15:00 on Mid-Autumn Festival in the historical data of the current user is obtained.

[0080] It should be noted that the position of the date where the current time is located in the current week is: Monday is the first day, Tuesday is the second day, Wednesday is the third day, Thursday is the fourth day, Friday is the fifth day, Saturday is the sixth day, and Sunday is the seventh day.

[0081] In this application, the regional target total electricity consumption is analyzed to obtain regional fluctuation factors, including:

[0082] B1: Extract the target total power consumption QMZ of the region and the number of households H corresponding to the target total power consumption of the region, and judge whether the number of households is less than the set number of households BH; if yes, jump to B2; if no, jump to B3;

[0083] B2: Based on the formula QMZP=QMZ / H, the regional average QMZP of the regional target total electricity consumption is obtained, and based on the formula QBZ=QMZP / BZP, the regional fluctuation factor QBZ is obtained; where BZP is the standard regional average value set manually;

[0084] B3: The regional fluctuation factor QBZ is obtained based on the formula QBZ=(1+ln((H-BH) / BH+1))×(QMZP / BZP).

[0085] It should be noted that human or natural influences may cause users in the area to use more or less electricity at this time. Human influences include changes in regional electricity consumption caused by road construction, line inspections, etc., and natural influences include changes in regional electricity consumption caused by typhoons, fires, etc.

[0086] In this application, the geographical location is analyzed to obtain location influencing factors, including:

[0087] The geographical location is extracted, the number of electricity users DH in the area is extracted, the number of abnormal electricity users YDH in the area is extracted, and the location influence factor WYZ is obtained based on the formula WYZ=exp(YDH / DH); wherein abnormal electricity consumption includes electricity theft and excess electricity consumption.

[0088] In this application, the electricity consumption factor is obtained by analyzing the historical reference data, including:

[0089] The user's target total electricity consumption HML and historical reference data LCS corresponding to the current time are extracted, and the electricity consumption factor DYZ is obtained based on the formula DYZ=HML / LCS.

[0090] See also Figure 3 In this application, the power consumption factor is calculated once based on the regional fluctuation factor to obtain a first calculated value, including:

[0091] C1: extract the regional fluctuation factor QBZ and the electricity consumption factor DYZ, and judge whether the value of QBZ-DYZ is greater than the maximum value of the judgment range; if yes, obtain the first calculated value JZ1 based on the formula JZ1=β×ln((DYZ-QBZ) / QBZ+1); if no, jump to C2;

[0092] C2: Determine whether the value of QBZ-DYZ is less than the maximum value of the judgment range; if yes, obtain the first calculated value JZ1 based on the formula JZ1=β×(-ln(-(DYZ-QBZ) / QBZ+1)); if no, obtain the first calculated value JZ1 based on the formula JZ1=δ×(DYZ-QBZ) / QBZ; wherein β and δ are both amplitude adjustment factors greater than 0.

[0093] It should be noted that when the value of QBZ-DYZ is greater than the maximum value of the judgment range, it means that the current user's electricity consumption has exceeded the standard electricity consumption of the area, which means that the current user may be in over-consumption of electricity;

[0094] When the value of QBZ-DYZ is within the judgment range, it means that the fluctuation of the current user's electricity consumption is consistent with the normal electricity consumption, which means that the current user may be in normal electricity consumption;

[0095] When the value of QBZ-DYZ is less than the maximum value of the judgment range, it means that the current user's electricity consumption is lower than the standard electricity consumption in the area, which means that the current user may be stealing electricity.

[0096] In the present application, the second calculation value is obtained by performing a secondary calculation on the power consumption factor based on the location influence factor, including:

[0097] The location influence factor WYZ and the power consumption factor DYZ are extracted, and a second calculated value JZ2 is obtained based on the formula JZ2=μ×(1+WYZ)×DYZ; wherein μ is a proportional adjustment factor greater than 0.

[0098] A behavioral diagnostic model is established in this application, including:

[0099] Acquire a historical electricity consumption factor, and a first calculated value and a second calculated value corresponding to the historical electricity consumption factor through a database;

[0100] Integrate historical electricity consumption factors, and the first calculated value, the second calculated value and the diagnostic results corresponding to the historical electricity consumption factors into several groups of training data and test data; use the training data to train the artificial intelligence model; use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally obtain a behavioral diagnosis model with the electricity consumption factor, the first calculated value and the second calculated value as input and the diagnostic result as output; wherein the artificial intelligence model includes a BP neural network model or an RBF neural network model.

[0101] Specifically, the trained artificial intelligence model is tested using the test data, and the specific operations for adjusting the artificial intelligence model according to the test results are: input the historical power consumption factor, the first calculated value, and the second calculated value in the test data into the trained artificial intelligence model to obtain the corresponding diagnostic result, and compare the diagnostic result with the corresponding diagnostic result in the test data. When the two are the same, there is no need to adjust the parameters and perform the next set of test data detection; when the two are different, adjust the corresponding parameters until they are the same, and then perform the next set of test data detection; when the number of diagnostic results obtained from all test data that are the same as the corresponding diagnostic results in the test data accounts for 90% or more of the total test data, a behavioral diagnostic model with the power consumption factor, the first calculated value, and the second calculated value as input and the diagnostic result as output is obtained.

[0102] It should be noted that before model training and testing, the test results are obtained by manually analyzing the power consumption factor, the first calculated value and the second calculated value.

[0103] A second aspect of the present invention provides a method for diagnosing abnormal power usage behavior of a user, comprising the following steps:

[0104] S1: Obtain basic information;

[0105] S2: extracting historical electricity consumption from the user's historical data from the database, analyzing the historical electricity consumption to obtain historical reference data; analyzing the regional target total electricity consumption to obtain the regional fluctuation factor; analyzing the geographical location to obtain the location influence factor; obtaining the electricity consumption factor by analyzing the historical reference data, calculating the electricity consumption factor once based on the regional fluctuation factor to obtain a first calculated value, and calculating the electricity consumption factor twice based on the location influence factor to obtain a second calculated value;

[0106] S3: Establish a behavior diagnosis model, input the power consumption factor, the first calculated value and the second calculated value into the behavior diagnosis model to obtain a diagnosis result.

[0107] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0108] Working principle of the present invention:

[0109] The present invention first obtains the current time and the geographical location of the user, and sets up the target time period to obtain the user's target total electricity consumption and the regional target total electricity consumption. In this way, the electricity consumption in a complete unit of time can be obtained. Fixing the time can better analyze the user's electricity consumption behavior and improve the accuracy of the analysis; extract the historical electricity consumption in the user's historical data from the database; analyze the historical electricity consumption to obtain historical reference data. The electricity consumption of the user on different dates is different. Considering this, a more accurate reference data can be obtained, which provides data support for the subsequent calculation of the electricity consumption factor; analyze the regional target total electricity consumption to obtain the regional fluctuation factor, which can avoid the error of the user's electricity consumption behavior analysis caused by external interference factors; analyze the geographical location to obtain an influence factor of a default user affecting the electricity consumption behavior of other users, which increases the accuracy of the system analysis; obtain the electricity factor by analyzing the historical reference data; calculate the electricity factor once based on the regional fluctuation factor to obtain a first calculated value, and calculate the electricity factor twice based on the location influence factor to obtain a second calculated value; establish a behavior diagnosis model, and input the electricity factor, the first calculated value and the second calculated value into the behavior diagnosis model to obtain a diagnosis result.

[0110] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A diagnostic system for abnormal power consumption behavior of users, characterized in that: include: A data processing module, and a data collection module, a behavior diagnosis module and a database connected thereto; The data collection module is used to obtain basic information, including power consumption information, current time and user's geographical location; power consumption information includes user target total power consumption and regional target total power consumption; The data processing module is used to extract historical power consumption from the user's historical data from the database, analyze the historical power consumption to obtain historical reference data; analyze the regional target total power consumption to obtain the regional fluctuation factor; analyze the geographical location to obtain the location influence factor; obtain the power consumption factor by analyzing the historical reference data, calculate the power consumption factor once based on the regional fluctuation factor to obtain a first calculated value, and calculate the power consumption factor twice based on the location influence factor to obtain a second calculated value; The behavior diagnosis module is used to establish a behavior diagnosis model, input the power consumption factor, the first calculated value and the second calculated value into the behavior diagnosis model to obtain a diagnosis result; wherein the diagnosis result includes normal power consumption, power theft and excess power consumption.

2. A diagnostic system for abnormal power consumption behavior of a user according to claim 1, characterized in that: The obtaining of basic information includes: The current time is extracted from the computer, and the time period in which the current time is located is obtained, and the time period before the current time period is marked as the target time period; the user's electricity consumption in the target time period is obtained through a smart meter, and the electricity consumption is marked as the user's target total electricity consumption; the target total electricity consumption of all users in the area is added up to obtain the regional target total electricity consumption; wherein one hour is a time period.

3. A diagnostic system for abnormal power consumption behavior of a user according to claim 1, characterized in that: The historical reference data obtained by analyzing the historical power consumption includes: A1: extract the current time and determine whether the date of the current time is a statutory holiday; if yes, obtain the historical power consumption corresponding to the statutory holiday in the target time period, mark the average value of the historical power consumption as historical power consumption 1 LD1, set the value of historical power consumption 2 LD2 to 0, and jump to A4; if no, set the value of historical power consumption 1 LD1 to 0, and jump to A2; A2: Get the position of the current date in the current week, mark the days at the same position in the historical data as target days, get a number of historical power consumptions corresponding to the target time period in the target days, and establish a historical power consumption database based on the historical power consumptions; A3: Obtain the variance of the historical electricity consumption in the historical electricity consumption database, and determine whether the variance is less than the variance threshold; if yes, calculate the average value of the historical electricity consumption database to obtain the historical electricity consumption 2 LD2; if no, remove the historical electricity consumption in the historical electricity consumption database that is the most different from the mode of the historical electricity consumption database, and re-judge the variance until the variance of the historical electricity consumption database is less than the variance threshold, and then calculate the average value of the remaining historical electricity consumption in the historical electricity consumption database to obtain the historical electricity consumption 2 LD2; A4: Based on the formula LCS=α1×LD1+α2×LD2, the historical reference data LCS corresponding to the current user at the current time is obtained; wherein α1 and α2 are both proportional adjustment factors greater than 0, and α1+α2=1.

4. A diagnostic system for abnormal power consumption behavior of a user according to claim 1, characterized in that: The regional fluctuation factor obtained by analyzing the regional target total electricity consumption includes: B1: Extract the target total power consumption QMZ of the region and the number of households H corresponding to the target total power consumption of the region, and determine whether the number of households is less than the set number of households BH; if yes, jump to B2; if no, jump to B3; B2: Based on the formula QMZP=QMZ / H, the regional average value QMZP of the regional target total electricity consumption is obtained, and based on the formula QBZ=QMZP / BZP, the regional fluctuation factor QBZ is obtained; where BZP is the standard regional average value; B3: The regional fluctuation factor QBZ is obtained based on the formula QBZ=(1+ln((H-BH) / BH+1))×(QMZP / BZP).

5. A diagnostic system for abnormal power consumption behavior of a user according to claim 1, characterized in that: The analysis of the geographical location to obtain the location influencing factor includes: The geographical location is extracted, the number of electricity users DH in the area is extracted, the number of abnormal electricity users YDH in the area is extracted, and the location influence factor WYZ is obtained based on the formula WYZ=exp(YDH / DH); wherein abnormal electricity consumption includes electricity theft and excess electricity consumption.

6. A diagnostic system for abnormal power consumption behavior of a user according to claim 3, characterized in that: The electricity consumption factor obtained by analyzing the historical reference data includes: The user's target total electricity consumption HML and historical reference data LCS corresponding to the current time are extracted, and the electricity consumption factor DYZ is obtained based on the formula DYZ=HML / LCS.

7. A diagnostic system for abnormal power consumption behavior of a user according to claim 6, characterized in that: The step of calculating the power consumption factor based on the regional fluctuation factor to obtain a first calculated value includes: C1: extract the regional fluctuation factor QBZ and the electricity consumption factor DYZ, and judge whether the value of QBZ-DYZ is greater than the maximum value of the judgment range; if yes, obtain the first calculated value JZ1 based on the formula JZ1=β×ln((DYZ-QBZ) / QBZ+1); if no, jump to C2; C2: Determine whether the value of QBZ-DYZ is less than the maximum value of the judgment range; if yes, obtain the first calculated value JZ1 based on the formula JZ1=β×(-ln(-(DYZ-QBZ) / QBZ+1)); if no, obtain the first calculated value JZ1 based on the formula JZ1=δ×(DYZ-QBZ) / QBZ; wherein β and δ are both amplitude adjustment factors greater than 0.

8. A diagnostic system for abnormal power consumption behavior of a user according to claim 7, characterized in that: The second calculation value is obtained by performing a secondary calculation on the power consumption factor based on the location influence factor, including: The location influence factor WYZ and the power consumption factor DYZ are extracted, and a second calculated value JZ2 is obtained based on the formula JZ2=μ×(1+WYZ)×DYZ; wherein μ is a proportional adjustment factor greater than 0.

9. A diagnostic system for abnormal power consumption behavior of a user according to claim 8, characterized in that: The establishment of a behavioral diagnosis model comprises: Acquire a historical electricity consumption factor, and a first calculated value and a second calculated value corresponding to the historical electricity consumption factor through a database; Integrate historical electricity consumption factors, and the first calculated value, the second calculated value and the diagnostic results corresponding to the historical electricity consumption factors into several groups of training data and test data; use the training data to train the artificial intelligence model; use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally obtain a behavioral diagnosis model with the electricity consumption factor, the first calculated value and the second calculated value as input and the diagnostic result as output; wherein the artificial intelligence model includes a BP neural network model or an RBF neural network model.

10. A method for diagnosing abnormal power usage behavior of a user, based on the operation of a diagnostic system for abnormal power usage behavior of a user according to any one of claims 1 to 9, characterized in that: S1: Obtain basic information; S2: extract the historical electricity consumption in the user's historical data from the database, analyze the historical electricity consumption to obtain historical reference data; analyze the regional target total electricity consumption to obtain the regional fluctuation factor; Analyze the geographical location to obtain the location influence factor; analyze the historical reference data to obtain the power consumption factor, calculate the power consumption factor once based on the regional fluctuation factor to obtain a first calculation value, and calculate the power consumption factor twice based on the location influence factor to obtain a second calculation value; S3: Establish a behavior diagnosis model, input the power consumption factor, the first calculated value and the second calculated value into the behavior diagnosis model to obtain a diagnosis result.

Images