CN111784379B

CN111784379B - Estimation method and device for electric charge after-payment and screening method and device for abnormal cases

Info

Publication number: CN111784379B
Application number: CN202010424910.8A
Authority: CN
Inventors: 万泉; 陈雁; 刘俊玲; 袁葆; 欧阳红; 张文; 姜涛; 刘俊恺; 白琳; 闫富荣
Original assignee: State Grid Corp of China SGCC; State Grid Information and Telecommunication Co Ltd; Beijing China Power Information Technology Co Ltd
Current assignee: State Grid Corp of China SGCC; State Grid Information and Telecommunication Co Ltd; Beijing China Power Information Technology Co Ltd
Priority date: 2020-05-19
Filing date: 2020-05-19
Publication date: 2023-09-15
Anticipated expiration: 2040-05-19
Also published as: CN111784379A

Abstract

The application provides an estimation method and device of a charge for collecting electricity and a screening method and device of abnormal cases, which are used for calculating the charge for collecting electricity corresponding to the electricity larceny amount of a user in an electricity larceny time interval by combining data of the electricity larceny method, the electricity larceny time interval, the electricity larceny amount of the user and the like. The method solves the problems that when the electric charge is calculated and the confirmed electric larceny user pays the electric charge, if the electric charge to be paid is calculated directly by the existing electric charge calculating mode, the calculated electric charge is inaccurate, and the phenomenon of excessive or insufficient electric charge is caused.

Description

Estimation method and device for electric charge after-payment and screening method and device for abnormal cases

Technical Field

The application relates to the technical field of power systems, in particular to an estimation method and device for electric charge collection and a screening method and device for abnormal cases.

Background

With the continuous increase of regional economy, the living standard of residents is continuously improved, and the electricity stealing phenomenon is more serious when the electric power demand is increased. The electricity stealing method is various, so that the electricity stealing checking work is more difficult. Aiming at the confirmed electricity stealing users, how to accurately and reasonably calculate the electric charge to be paid after-charge becomes a great difficulty in the field of electricity stealing prevention.

At present, when the after-charge of the confirmed electricity larceny user is calculated, the existing calculating mode of the after-charge electricity fee is mainly calculated according to an electric power method and a power supply business rule, and the calculating mode has a cut-off problem. Because of different types of electricity larceny users and different electricity larceny methods of the electricity larceny users, if the electric charge to be paid is calculated directly by using the existing electric charge calculation mode, the calculated electric charge to be paid is easy to be inaccurate, and the phenomenon of excessive or insufficient electric charge to be paid often exists. This can lead to inaccurate handling of the anti-theft work and also to loss of the relevant power supply units.

Disclosure of Invention

In view of this, the present application provides a method and apparatus for estimating the electric charge to be paid and a method and apparatus for screening abnormal cases, so as to solve the problem that when calculating the electric charge to be paid by the confirmed electricity larceny user, if the electric charge to be paid is directly calculated by the existing electric charge calculation method, the calculated electric charge to be paid is not accurate enough, and there is often a phenomenon that the electric charge to be paid is too much or too little.

In order to achieve the above purpose, the present application provides the following technical solutions:

the first aspect of the application discloses an estimation method for electric charge collection, which comprises the following steps:

Aiming at a user with electricity stealing behavior, acquiring electricity utilization characteristic data of the user; the electricity consumption characteristic data comprise the electricity consumption of the user, the line loss electricity of the station area and the time when the user opens the ammeter table cover;

obtaining the electricity stealing time interval of the user according to the electricity utilization characteristic data;

identifying the electricity stealing method of the user, and obtaining the average electricity stealing quantity of the user according to the electricity stealing method of the user;

calculating the electricity stealing time interval and the average electricity stealing quantity to obtain the electricity stealing quantity of the user in the electricity stealing time interval;

and calculating the electric charge to be paid after the user steals the electric quantity in the electricity stealing time interval according to a preset electric charge calculation method.

Optionally, in the above method, the obtaining the electricity stealing time interval of the user according to the electricity consumption characteristic data includes:

inquiring the electricity utilization characteristic data to obtain an electricity utilization data sequence from the last time of opening the meter cover to the time of stealing electricity check by the user; the power consumption data sequence comprises a line loss electric quantity sequence of a station area of each day and a power consumption sequence of a user in the period from the last time of opening a meter cover to the time of stealing electricity check by the user;

Searching a time interval meeting the condition that the line loss electric quantity of the station area is suddenly increased and the simultaneous occurrence frequency of the user power consumption sudden drop is higher than a preset threshold value in the power consumption data sequence, and taking the time interval as a first suspected power stealing time interval;

screening to obtain a time interval from the electricity larceny checking time to the last time of opening the meter cover from the electricity larceny checking time of the user;

sliding window scanning is carried out in a time interval from the electricity stealing checking time to the last meter cover opening time from the electricity stealing checking time to obtain a time interval in which the Person correlation between the line loss data of all the station areas and the electricity consumption of the user is higher than a preset threshold value, and the time interval is used as a second suspected electricity stealing time interval;

and merging the first suspected electricity larceny time and the second suspected electricity larceny time to obtain an electricity larceny time interval of the user.

Optionally, in the above method, the obtaining average electricity larceny amount data of the user according to an electricity larceny method of the user includes:

if the electricity stealing method of the user is equal-ratio electricity stealing, judging the proportion of electricity stealing according to an electricity stealing instrument of the user;

Calculating the user electricity consumption data according to the proportion to obtain average electricity stealing quantity data of the user;

if the electricity stealing method of the user is bypass electricity stealing, acquiring an electricity stealing application mode of the user;

and inquiring the average electricity consumption corresponding to the electricity stealing application mode to obtain the average electricity stealing electricity data of the user.

The application discloses a screening method for abnormal cases, which comprises the following steps:

aiming at the electricity stealing case samples for which the electric charge service is paid, analyzing the data of each electricity stealing case sample to obtain sample characteristic data of each electricity stealing case sample; the sample characteristic data comprise electricity stealing methods, electricity consumption, line loss, average electricity stealing data, the ratio of live line current to zero line current, the interval time between power failure events and power up events and the interval time between meter cover opening events and meter cover closing events of each sample case;

dividing the electricity stealing case samples by using a fuzzy clustering algorithm according to the sample characteristic data to form clustering results of one or more electricity stealing categories;

Obtaining standard additional charge of each electricity stealing category according to the clustering result;

screening out an electric charge tracing service abnormal case by using the electric charge tracing sample of the electric charge tracing case of the same electric charge tracing type and the standard electric charge tracing; the additional charge of the electricity larceny case sample is calculated by the method according to any one of the first aspect of the application.

Optionally, in the above method, the method for screening the additional charge service abnormal case by using the additional charge and the standard additional charge of the electricity stealing case sample of the same electricity stealing category includes:

the electric charge collected by each electric charge collecting case sample of the same electric charge collecting class is respectively differed from the standard electric charge collected by the same electric charge collecting class, and a result value is obtained;

if the result value is larger than a preset threshold value, the electricity larceny case sample corresponding to the result value belongs to the additional charge business abnormal case.

Optionally, in the above method, the method uses the electric charge collected by the electric charge collection case sample of the same electric charge collection category and the standard electric charge collected by the electric charge collection, and screens out abnormal cases of electric charge collected business, including

Inputting the standard additional charge of each electricity larceny category and sample characteristic data of each electricity larceny case sample into a classifier;

Dividing the electricity stealing case sample into a training set and a testing set;

training a classifier by using the electricity stealing case samples in the training set, and setting the electricity stealing case samples as abnormal cases of the service of the additional charge if the difference value between the additional charge of the electricity stealing case samples in the same electricity stealing category in the training set and the standard additional charge of the same electricity stealing category is larger than the threshold value;

and inputting the electricity stealing case samples in the test set into the classifier for classification processing, and screening out the after-call electricity fee service abnormal cases.

Optionally, the method further comprises:

and evaluating the electric charge service work by using the screened electric charge service abnormality case data.

The third aspect of the present application discloses an estimation device for electric charge collection, comprising:

the device comprises an acquisition unit, a control unit and a control unit, wherein the acquisition unit is used for acquiring electricity utilization characteristic data of a user aiming at the user with electricity stealing behavior; the electricity consumption characteristic data comprise the electricity consumption of the user, the line loss electricity of the station area and the time when the user opens the ammeter table cover;

the first processing unit is used for obtaining the electricity stealing time interval of the user according to the electricity utilization characteristic data;

The second processing unit is used for identifying the electricity stealing method of the user and obtaining the average electricity stealing quantity of the user according to the electricity stealing method of the user;

the first calculation unit is used for calculating the electricity stealing time interval and the average electricity stealing quantity to obtain the electricity stealing quantity of the user in the electricity stealing time interval;

and the second calculation unit is used for calculating the additional charge corresponding to the electricity larceny quantity of the user in the electricity larceny time interval according to a preset electricity charge calculation method.

Optionally, in the foregoing apparatus, the first processing unit includes:

the first inquiring subunit is used for inquiring the electricity utilization characteristic data to obtain an electricity utilization data sequence from the last time of opening the meter cover to the time of stealing electricity check by the user; the power consumption data sequence comprises a line loss electric quantity sequence of a station area of each day and a power consumption sequence of a user in the period from the last time of opening a meter cover to the time of stealing electricity check by the user;

the first searching subunit is used for searching a time interval meeting the condition that the line loss electric quantity of the station area is suddenly increased and the simultaneous occurrence frequency of the user power consumption is higher than a preset threshold value in a preset time period from the power consumption data sequence, and taking the time interval as a first suspected power stealing time interval;

The second searching subunit is used for screening and obtaining a time interval from the electricity stealing checking time to the last time of opening the meter cover from the electricity stealing checking time of the user in the time of opening the meter cover of the user;

the scanning subunit is used for carrying out sliding window scanning in a time interval from the electricity stealing checking time to the last time of opening the meter cover time from the electricity stealing checking time to obtain a time interval in which the pearson correlation between the line loss data of all the station areas and the electricity consumption of the user is higher than a preset threshold value, and the time interval is used as a second suspected electricity stealing time interval;

and the merging subunit is used for merging the first suspected electricity larceny time and the second suspected electricity larceny time to obtain an electricity larceny time interval of the user.

Optionally, in the foregoing apparatus, the second processing unit includes:

a judging subunit, configured to judge the proportion of electricity larceny according to the electricity larceny instrument of the user if the electricity larceny method of the user is equal-ratio electricity larceny;

the calculating subunit is used for calculating the user electricity consumption data according to the proportion to obtain average electricity stealing electricity consumption data of the user;

The acquisition subunit is used for acquiring the electricity stealing application mode of the user if the electricity stealing method of the user is bypass electricity stealing;

and the second inquiry subunit is used for inquiring the average electricity consumption corresponding to the electricity stealing application mode to obtain the average electricity stealing electricity quantity data of the user.

The fourth aspect of the present application discloses a screening device for abnormal cases, including:

the analysis unit is used for analyzing the data of each electricity stealing case sample aiming at the electricity stealing case samples for which the electric charge service is paid, so as to obtain sample characteristic data of each electricity stealing case sample; the sample characteristic data comprise electricity stealing methods, electricity consumption, line loss, average electricity stealing data, the ratio of live line current to zero line current, the interval time between power failure events and power up events and the interval time between meter cover opening events and meter cover closing events of each sample case;

the dividing unit is used for dividing the electricity stealing case samples by using a fuzzy clustering algorithm according to the sample characteristic data to form clustering results of one or more electricity stealing categories;

the acquisition unit is used for acquiring the standard additional charge of each electricity stealing category according to the clustering result;

The screening unit is used for screening out the additional charge service abnormal cases by utilizing the additional charge and the standard additional charge of the electricity stealing case samples of the same electricity stealing category; the additional charge of the electricity larceny case sample is calculated by the method according to any one of the first aspect of the application.

Optionally, in the foregoing apparatus, the screening unit includes:

the calculating subunit is used for respectively differencing the electric charge collected by each electric charge stealing case sample of the same electric larceny type with the standard electric charge collected by the same electric larceny type to obtain a result value;

and the judging subunit is used for judging whether the electricity stealing case sample corresponding to the result value belongs to the additional charge business abnormal case if the result value is larger than a preset threshold value.

Optionally, in the foregoing apparatus, the screening unit includes:

the input subunit is used for inputting the standard additional charge of each electricity larceny type and sample characteristic data of each electricity larceny case sample into the classifier;

the dividing subunit is used for dividing the electricity stealing case sample into a training set and a testing set;

the training subunit is used for training the classifier by using the electricity stealing case samples in the training set, and if the difference value between the electric charge collected by the electricity stealing case samples in the same electricity stealing category in the training set and the standard electric charge collected by the electricity stealing category is larger than the threshold value, the electricity stealing case samples are set to be abnormal electric charge collecting business cases;

And the screening subunit is used for inputting the electricity stealing case samples in the test set into the classifier for classification processing and screening out the after-payment electricity fee business abnormal cases.

Optionally, the above device further includes:

and the evaluation unit is used for evaluating the electric charge pursuing business work by using the screened electric charge pursuing business abnormal case data.

According to the technical scheme, the estimation method for the electric charge after-payment obtains the electricity utilization characteristic data of the user aiming at the user with the electricity stealing behavior. The electricity consumption characteristic data comprise the electricity consumption of the user, the line loss electricity of the station area and the time for opening the ammeter table cover by the user. And then obtaining the electricity stealing time interval of the user according to the electricity utilization characteristic data. And identifying the electricity stealing method of the user, and obtaining the average electricity stealing quantity of the user according to the electricity stealing method of the user. And calculating the electricity stealing time interval and the average electricity stealing quantity to obtain the electricity stealing quantity of the user in the electricity stealing time interval. And finally, calculating the electric charge to be paid after the user steals the electric quantity in the electricity stealing time interval according to a preset electric charge calculation method. Therefore, the method can solve the problems that when the electric charge is calculated and the confirmed electric larceny user pays the electric charge, if the electric charge to be paid is calculated directly by the existing electric charge calculating mode, the calculated electric charge is inaccurate, and the phenomenon of excessive or insufficient electric charge is often caused.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present application, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart of an estimation method for electric charge collection according to an embodiment of the present application;

FIG. 2 is a flow chart of one implementation of step S102 disclosed in another embodiment of the present application;

fig. 3 is a flowchart of a screening method for abnormal cases according to another embodiment of the present application;

FIG. 4 is a flow chart of one implementation of step S304 disclosed in another embodiment of the present application;

FIG. 5 is a schematic diagram of a linear support vector machine in the prior art;

FIG. 6 is a schematic diagram of an estimation device for electric charge collection according to another embodiment of the present application;

fig. 7 is a schematic diagram of a screening apparatus for abnormal cases according to another embodiment of the present application.

Detailed Description

The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.

In the present disclosure, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

Moreover, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.

As known from the background art, when calculating the after-charge of the confirmed electricity larceny user, the existing calculating mode of the after-charge electric charge is mainly calculated according to the electric power method and the power supply business rule, and the calculating mode has a cut-off problem. Because of different types of electricity larceny users and different electricity larceny methods of the electricity larceny users, if the electric charge to be paid is calculated directly by using the existing electric charge calculation mode, the calculated electric charge to be paid is easy to be inaccurate, and the phenomenon of excessive or insufficient electric charge to be paid often exists. This can lead to inaccurate handling of the anti-theft work and also to loss of the relevant power supply units.

In view of this, an embodiment of the present application provides a method for adjusting a server load, as shown in fig. 1, which specifically includes:

s101, aiming at a user with electricity stealing behavior, acquiring electricity utilization characteristic data of the user; the electricity consumption characteristic data comprise the electricity consumption of the user, the line loss electricity of the station area and the time for opening the ammeter table cover by the user.

It should be noted that, for those users who have determined that they have electricity stealing actions, electricity stealing is performed by the electricity consumer with the purpose of reducing the metering of electric energy and reducing the payment of electric charges. The electricity consumption characteristic data comprise the electricity consumption of the user, the line loss electricity of the station area and the time when the user opens the ammeter table cover. Active power loss and electric energy loss generated in the transmission and distribution process of the power network are collectively called line loss, and are abbreviated as line loss. The line loss types can be classified into statistical line loss, theoretical line loss and management line loss types, wherein the line loss electric quantity of the station area refers to statistical line loss, namely actual line loss, and is the difference value between the power supply quantity and the sales electric quantity calculated according to the electric energy meter index.

S102, obtaining a power stealing time interval of the user according to the power utilization characteristic data.

The method is characterized in that the electricity consumption characteristic data such as the electricity consumption of a user, the line loss electricity of a station area, the time when the user opens an ammeter cover and the like are utilized to analyze the data which accords with the electricity stealing characteristics of the user, and the electricity stealing time interval of the user is obtained.

Optionally, in another embodiment of the present application, an implementation manner of step S102, as shown in fig. 2, specifically includes:

s201, inquiring electricity utilization characteristic data to obtain an electricity utilization data sequence from the last time of opening a meter cover to the time of stealing electricity by a user; the power consumption data sequence comprises a line loss power sequence of a station area of each day and a power consumption sequence of a user in the period from the last time of opening the meter cover to the time of stealing electricity.

It should be noted that, the electricity consumption of the user, the line loss of the transformer area, the time when the user opens the ammeter cover, and other electricity consumption characteristic data are queried to obtain the electricity consumption data sequence from the last time when the user opens the ammeter cover to the time when the user steals the ammeter. Wherein the electricity consumption data sequence comprises the time period from the last time of opening the meter cover to the time period of the electricity larceny checkLine loss power sequence of each day of the area in the room and power consumption sequence of the user. For example, assuming that the latest cover opening time is 1/2020 and the latest cover opening time is 1/2020, and the electric larceny check time is 3/2020, the power loss sequence Y (Y ₁ ,y ₂ …y _n ) User A daily electricity quantity sequence X (X) ₁ ,x ₂ …x _n )。

S202, a time interval meeting the condition that the line loss electricity quantity of the station area is suddenly increased and the simultaneous occurrence frequency of the user electricity consumption sudden drop is higher than a preset threshold value in a preset time period is searched out from the electricity consumption data sequence and is used as a first suspected electricity stealing time interval.

It should be noted that, the sudden increase of the line loss power generally indicates that the power load of the users in the area is generally increased or the power stealing behavior is caused, and if the line loss is increased and the power consumption of the users is reduced, abnormal behavior may exist. Therefore, a time interval meeting the condition that the power consumption of the station area is suddenly increased and the simultaneous occurrence frequency of the power consumption of the user is higher than a threshold value in a certain time period is searched from the power consumption data sequence and is used as a first suspected power stealing time interval. For example, in a time zone of 30 days, from 1 in 2020 to 1 in 2020, and in which the power consumption of the station is suddenly increased and the number of simultaneous user power consumption drops is higher than 5, a time zone of 1 in 2020 to 1 in 2020 is taken as a first suspected power theft time zone.

S203, screening to obtain a time interval from the electricity larceny checking time to the last time of opening the meter cover from the electricity larceny checking time from the user to the electricity larceny checking time.

It should be noted that, the time when the user opens the ammeter lid is obtained from the user electricity consumption characteristic data, and then the time interval from the time when the user opens the ammeter lid to the time when the user steals the ammeter to the time when the ammeter lid is opened last time from the time when the user steals the ammeter.

S204, sliding window type scanning is carried out in a time interval from the electricity stealing checking time to the last time of opening the meter cover time from the electricity stealing checking time, so that a time interval in which the line loss data of all the areas and the pearson correlation coefficient of the electricity consumption of the user are higher than a preset threshold value is obtained, and the time interval is used as a second suspected electricity stealing time interval.

It should be noted that, the pearson correlation coefficient (Pearson correlation coefficient) is a linear correlation coefficient, which is used to reflect statistics of the linear correlation degree of two variables, and is also called "simple correlation coefficient", "pearson product moment correlation coefficient", or "linear correlation coefficient", and is generally denoted by r, where n is a sample size, and is an observed value and an average value of the two variables, respectively. r describes the degree of linear correlation between two variables. The value of r is between-1 and +1, if r is more than 0, the two variables are positively correlated, namely, the larger the value of one variable is, the larger the value of the other variable is; if r <0, it indicates that the two variables are inversely related, i.e., the larger the value of one variable, the smaller the value of the other variable. If r=0, this indicates that the two variables are not linearly related, but other ways of relating are possible, such as a curvilinear way. The correlation between the line loss data of the transformer area and the power consumption of the user can be obtained by calculating the pearson correlation coefficient, and when the correlation between the daily power consumption of a single user and the daily line loss power consumption of the transformer area is very high, the power consumption behavior of the single user is similar to the change of the line loss of the transformer area, so that the power stealing behavior is very likely to exist.

Therefore, the sliding window type scanning is performed in a time interval from the time of the electricity larceny check to the time of the last time of opening the meter cover from the time of the electricity larceny check, 15 days can be set as a window, 7 days are set as step sizes, and the cyclic scanning is performed. And obtaining a time interval in which the pearson correlation coefficient of the line loss data of all the areas and the power consumption of the user is higher than a threshold value, wherein the threshold value can be set to be 0.85 and used as a second suspected electricity stealing time interval. For example, in the time interval of 1 st month and 1 st month in 2020, the pearson correlation coefficient between the line loss data of the station and the power consumption of the user in the two time intervals of 1 st month and 15 th month in 2020 and 14 th month and 28 th month in 2020 is higher than 0.85, both the time intervals are regarded as the second suspected electricity larceny time interval.

And S205, combining the first suspected electricity larceny time and the second suspected electricity larceny time to obtain an electricity larceny time interval of the user.

As can be seen from the above examples, the first time period for suspected electricity theft is from 1/2/2020 to 1/3/2020, and the second time period for suspected electricity theft is from 1/2020 to 15/2020 and from 14/2020 to 28/2020. And combining the first suspected electricity larceny time and the second suspected electricity larceny time to obtain two time intervals, namely, 1 month and 1 day in 2020 to 15 days in 2020 and 1 month and 1 day in 2020 to 3 days in 2020, wherein the two time intervals are the electricity larceny time intervals of the user.

S103, identifying the electricity stealing method of the user, and obtaining the average electricity stealing quantity of the user according to the electricity stealing method of the user.

The staff identifies the electricity stealing method of the electricity stealing user according to experience, and analyzes data according to different electricity stealing methods of the electricity stealing user to obtain the average electricity stealing quantity of the user.

Optionally, in another embodiment of the present application, an implementation of obtaining the average electricity larceny amount of the user according to the electricity larceny method of the user in step S103 may include:

if the electricity stealing method of the user is equal-ratio electricity stealing, judging the proportion of electricity stealing according to the electricity stealing instrument of the user.

And calculating the electricity consumption data of the user according to the proportion to obtain the average electricity stealing electricity consumption data of the user.

If the electricity stealing method of the user is bypass electricity stealing, the electricity stealing application mode of the user is obtained.

And inquiring the average electricity consumption corresponding to the electricity stealing application mode to obtain the average electricity stealing data of the user.

It should be noted that if the electricity stealing method of the user is equal ratio electricity stealing, the actual electricity consumption is generally several times of the electricity metering quantity of the meter, so that the proportion of electricity stealing can be determined according to the property of the collecting electricity stealing instrument, and the electricity consumption data of the user is calculated according to the proportion, so as to obtain the average electricity stealing electricity consumption data of the user. If the electricity is stolen around, the electricity stealing user usually goes over metering devices such as an ammeter and the like to steal electricity, and the actual electricity consumption is difficult to judge through the electricity metering, so that the average electricity stealing electricity quantity data of the user can be obtained according to the electricity stealing application mode of the user, for example, the average electricity consumption corresponding to a part of high-power machines used for cultivation or production by the user is inquired.

And S104, calculating the electricity stealing time interval and the average electricity stealing quantity to obtain the electricity stealing quantity of the user in the electricity stealing time interval.

After the electricity stealing time interval and the average electricity stealing amount of the electricity stealing user are obtained, the average electricity stealing amount of the user is multiplied by the electricity stealing time, so that the total electricity stealing amount of the user in the electricity stealing time interval can be obtained.

S105, calculating the electric charge to be paid after the user steals the electric quantity in the electricity stealing time interval according to a preset electric charge calculation method.

After obtaining the total electricity stealing amount of the user in the electricity stealing time interval, calculating the additional charge corresponding to the total electricity stealing amount of the user in the electricity stealing time interval according to the electricity charge calculation method of each power supply unit.

In the estimation method for electric charge collection provided by the embodiment, aiming at the user with electricity stealing behavior, the electricity utilization characteristic data of the user is obtained. The electricity consumption characteristic data comprise the electricity consumption of the user, the line loss electricity of the station area and the time for opening the ammeter table cover by the user. And then obtaining the electricity stealing time interval of the user according to the electricity utilization characteristic data. And identifying the electricity stealing method of the user, and obtaining the average electricity stealing quantity of the user according to the electricity stealing method of the user. And calculating the electricity stealing time interval and the average electricity stealing quantity to obtain the electricity stealing quantity of the user in the electricity stealing time interval. And finally, calculating the electric charge to be paid after the user steals the electric quantity in the electricity stealing time interval according to a preset electric charge calculation method. Therefore, the method can solve the problems that when the electric charge is calculated and the confirmed electric larceny user pays the electric charge, if the electric charge to be paid is calculated directly by the existing electric charge calculating mode, the calculated electric charge is inaccurate, and the phenomenon of excessive or insufficient electric charge is often caused.

In another embodiment of the present application, a method for screening abnormal cases is also disclosed, see fig. 3, which specifically includes:

s301, analyzing data of each electricity stealing case sample aiming at the electricity stealing case sample for which the electric charge service is paid, so as to obtain sample characteristic data of each electricity stealing case sample; the sample characteristic data comprise electricity stealing methods, electricity consumption, line loss, average electricity stealing data, the ratio of live line current to neutral line current, the interval time between power failure events and power up events and the interval time between meter cover opening events and meter cover closing events of each sample case.

It should be noted that, for the electricity stealing case samples for which the electric charge service has been paid, the data of each electricity stealing case sample is analyzed, and the sample characteristic data of each electricity stealing case sample is extracted. The sample characteristic data comprise electricity stealing methods, electricity consumption, line loss, average electricity stealing data, the ratio of live line current to neutral line current, the interval time between power failure events and power up events and the interval time between meter cover opening events and meter cover closing events of each sample case.

S302, dividing the electricity stealing case samples by using a fuzzy clustering algorithm according to the sample characteristic data to form clustering results of one or more electricity stealing categories.

It should be noted that the fuzzy C-means algorithm is a clustering algorithm based on partitioning, and its idea is to maximize the similarity between objects partitioned into the same cluster, and minimize the similarity between different clusters. The fuzzy C-means algorithm is an improvement of the common C-means algorithm, which is hard for data division, and the fuzzy C-means algorithm is a flexible fuzzy division. Let the sample set be x= { X ₁ ，x ₂ ，…，x _n The number of categories is denoted by C, the center of each cluster is denoted by m (j=1, 2, …, C), u _j (x _i ) Is the membership function of the ith sample corresponding to the jth class, then the clustering loss function based on the membership functionCan be written as:

b is a weighted index, also called a smoothing factor, and controls the sharing degree of the modes among fuzzy classes; let J _f For m _j And u _j (x _i ) The bias of (2) is 0, and the minimum value of the formula (1) is obtained. From equation (1):

and (3) solving the formula (2) and the formula (3) by adopting an iterative mode method until a convergence condition is met, so as to obtain an optimal solution.

Among the fuzzy clustering algorithms, the fuzzy C-means algorithm is most widely and successfully applied, and obtains the membership degree of each sample point to all class centers by optimizing an objective function, so that the class of the sample point is determined to achieve the purpose of automatically classifying the sample data. The algorithm is specifically executed with reference to the prior art, and will not be described in detail here.

Therefore, after sample characteristic data of each electricity stealing case sample are obtained, the electricity stealing case samples are divided by using a fuzzy C-means algorithm, and different class clusters are distinguished according to membership degree sizes to form clustering results of one or more electricity stealing classes.

S303, obtaining the standard additional charge of each electricity stealing category according to the clustering result.

It should be noted that, after the clustering result of one or more electricity stealing categories is formed, the standard electric charge for each electricity stealing category is obtained according to the characteristics of the clustering center of each electricity stealing category, and the clustering center is generally the average electric charge for all the case samples in the electricity stealing category.

S304, utilizing the electric charge recollection and standard electric charge recollection of the electric charge stealing case samples of the same electric charge stealing type to screen out abnormal electric charge recollection business cases; the electric charge recollection of the electric charge stealing case sample is calculated by the electric charge recollection estimation method.

After obtaining the standard electric charge collected by each electricity larceny type, comparing the actual electric charge collected by the electricity larceny case sample belonging to the same electricity larceny type with the standard electric charge collected by the electricity larceny type, and if the actual electric charge collected by a certain electricity larceny case sample is different from the standard electric charge collected by the electricity larceny type by more than one standard value, the electric charge collected by the electricity larceny type is an abnormal case, thereby screening out the abnormal case of the electric charge collected by the electricity larceny type. The actual electric charge recollection of the electric charge stealing case sample is calculated by an estimation method such as the electric charge recollection.

Optionally, in another embodiment of the present application, an implementation manner of step S304 specifically includes:

and respectively differencing the electric charge collected by each electric charge stealing case sample of the same electric charge stealing category with the standard electric charge collected by the same electric charge stealing category to obtain a result value.

If the result value is larger than the preset threshold value, the electricity stealing case sample corresponding to the result value belongs to the additional charge business abnormal case.

After obtaining the standard electric charge collected from each electricity larceny type, calculating the difference between the actual electric charge collected from each electricity larceny case sample belonging to the same electricity larceny type and the standard electric charge collected from the same electricity larceny type. If the difference between the actual electric charge collected from a certain electricity-stealing case sample and the standard electric charge collected from the electricity-stealing category is larger than a threshold, the threshold is generally set to be thirty percent of the standard electric charge collected from the corresponding electricity-stealing category, and the electricity-stealing case sample belongs to an abnormal case of the electric charge collecting business.

Optionally, in another embodiment of the present application, an implementation manner of step S304, as shown in fig. 4, specifically includes:

s401, inputting the standard additional charge of each electricity stealing category and the sample characteristic data of each electricity stealing case sample into a classifier.

It should be noted that, in this embodiment, the standard additional charge of each electricity stealing category and the sample feature data of each electricity stealing case sample are input as feature vectors into a Support vector machine (Support VectorMachine, SVM) classifier for data analysis and processing.

It should be noted that, the support vector machine method is generally used for classification, and is a classification method with strong universality. The nature of classification is to take a vector as input and assign a type tag to this vector. For sample space s= { (x) ₁ ,y ₁ ),(x ₂ ,y ₂ ),…(x _n ,y _n ) X represents the sample feature and y is the sample label. In most classification scenarios, the types are mutually exclusive, so each sample is assigned only one type. The input vector space is divided by decision regions, which are determined by decision boundaries or decision planes. As shown in fig. 5, the regular triangle and square represent two different classes of samples, the blue square and gray triangle on the straight line are called support vectors, and D is called decision plane.

The SVM classifier may be referred to as a linear classifier, and may also be referred to as a nonlinear classifier. It is defined to find the model with the largest geometric separation in the two-dimensional space of fig. 5. The essence of the method is to solve the problem of convex quadratic programming. The ultimate goal of a linear SVM is to maximize the distance separation to find the hyperplane (in two dimensions, decision D) and to be the only hyperplane that exists. The linear separable SVM generally adopts an optimal solution, namely a maximum interval method, obtained by optimizing convex quadratic problem. The objective function of the linear separable SVM meets the Lagrangian function solving condition, and the linear separable SVM can be further extended to a nonlinear classification SVM by introducing the Lagrangian function, namely, a kernel function is introduced.

The above is an ideal case where training samples are separable, but often in practice, samples will always have noise or outliers. In order to eliminate the influence of noise point interference, a relaxation variable is introduced into a target formula of the SVM, the objective function is converted into soft interval maximization, and the obtained vector machine is called a linear SVM. When the linear model cannot solve the classification problem, a nonlinear model, such as an elliptic curve, may be used for classification. However, the nonlinear model is very difficult to solve, so the nonlinear model is generally converted into a linear model by a conversion mode.

S402, dividing the electricity stealing case sample into a training set and a testing set.

It should be noted that, from all the electricity stealing case samples, a part of samples are randomly selected as a training set for training the support vector machine classifier. And then taking all the electricity stealing case samples as a test set for classification processing.

S403, training the classifier by using the electricity stealing case samples in the training set, and setting the electricity stealing case samples as abnormal cases of the service of the electric charge to be paid if the difference value between the electric charge to be paid by the electricity stealing case samples in the same electricity stealing category in the training set and the standard electric charge to be paid by the electricity stealing category is larger than a threshold value.

It should be noted that, training the classifier by using the electricity stealing case samples in the training set, setting the threshold value for screening abnormal cases to be thirty percent of the standard additional charge of the electricity stealing type, and if the difference value between the actual additional charge of the electricity stealing case samples in the same electricity stealing type in the training set and the standard additional charge of the same electricity stealing type is greater than the threshold value, setting the electricity stealing case samples to be the additional charge business abnormal cases.

S404, inputting the electricity stealing case samples in the test set into a classifier for classification processing, and screening out the after-payment electricity fee business abnormal cases.

After the classifier is trained by using the electricity stealing case samples in the training set, the support vector machine classifier can realize the classification processing function, and at the moment, the case samples in the training set can be input into the support vector machine classifier for classification processing, so that the after-call electricity fee service abnormal cases can be directly screened out.

In the screening method for abnormal cases provided in this embodiment, the data of each electricity stealing case sample is analyzed by aiming at the electricity stealing case sample for which the electric charge service is paid, so as to obtain the sample characteristic data of each electricity stealing case sample. And then dividing the electricity stealing case samples by using a fuzzy clustering algorithm according to the sample characteristic data to form clustering results of one or more electricity stealing categories. And obtaining the standard additional charge of each electricity stealing category according to the clustering result. The electric charge recollection and standard electric charge recollection of the electric charge stealing case samples of the same electric charge stealing type are utilized to screen out abnormal cases of electric charge recollection service; the electric charge recollection of the electric charge stealing case sample is calculated by the electric charge recollection estimation method. Therefore, the data calculated by the estimation method for the electric charge can be checked, and the accuracy is improved.

Optionally, in another embodiment of the present application, the method for screening abnormal cases may further include:

It should be noted that, by using the screened out after-call electric charge business abnormal case data, the deviation between the reasonable after-call electric quantity of different typical categories is calculated for the site work order of each site processor, as shown in table 1,

TABLE 1

The generated form data is used for evaluating the electric charge collecting business work of the units or individuals, for example, a part of operators commonly have the deviation of the electric charge collecting business for processing the electric charge stealing user cases and the standard electric charge collecting business of the corresponding electric charge stealing types, so that the electric charge collecting business work has problems, the operators or the units of the operators or the individuals should be investigated for finding reasons, and corresponding business training can be carried out on the operators or the units if necessary.

The other embodiment of the present application also provides an estimation device for electric charge collection, as shown in fig. 6, which specifically includes:

an obtaining unit 601, configured to obtain, for a user who has performed electricity stealing actions, electricity usage characteristic data of the user; the electricity consumption characteristic data comprise the electricity consumption of the user, the line loss electricity of the station area and the time for opening the ammeter table cover by the user.

The first processing unit 602 is configured to obtain a power stealing time interval of the user according to the power consumption characteristic data.

The second processing unit 603 is configured to identify a power stealing method of the user, and obtain an average power stealing amount of the user according to the power stealing method of the user.

The first calculating unit 604 is configured to calculate the electricity stealing time interval and the average electricity stealing amount, so as to obtain the electricity stealing amount of the user in the electricity stealing time interval.

The second calculating unit 605 is configured to calculate a charge for additional charge corresponding to the amount of electricity stolen by the user in the electricity stealing time interval according to a preset charge calculating method.

In the estimation device for electric charge collection provided in this embodiment, the obtaining unit 601 obtains the electricity consumption characteristic data of the user for the user who has the electricity larceny. The electricity consumption characteristic data comprise the electricity consumption of the user, the line loss electricity of the station area and the time for opening the ammeter table cover by the user. The first processing unit 602 then obtains a power stealing time interval of the user according to the power consumption characteristic data. The second processing unit 603 identifies the electricity stealing manipulation of the user, and obtains the average electricity stealing amount of the user according to the electricity stealing manipulation of the user. The first calculating unit 604 calculates the electricity stealing time interval and the average electricity stealing amount, so as to obtain the electricity stealing amount of the user in the electricity stealing time interval. Finally, the second calculating unit 605 calculates the electric charge for the user to pay after-charge corresponding to the electric charge amount during the time interval of the electricity theft according to the preset electric charge calculating method. Therefore, the method can solve the problems that when the electric charge is calculated and the confirmed electric larceny user pays the electric charge, if the electric charge to be paid is calculated directly by the existing electric charge calculating mode, the calculated electric charge is inaccurate, and the phenomenon of excessive or insufficient electric charge is often caused.

In this embodiment, the specific execution of the acquiring unit 601, the first processing unit 602, the second processing unit 603, the first computing unit 604, and the second computing unit 605 may refer to the content of the method embodiment corresponding to fig. 1, and will not be described herein.

Optionally, in another embodiment of the present invention, an implementation of the first processing unit 602 includes:

the first inquiring subunit is used for inquiring the electricity utilization characteristic data to obtain an electricity utilization data sequence from the last time of opening the meter cover to the time of stealing electricity inspection by the user; the power consumption data sequence comprises a line loss power sequence of a station area of each day and a power consumption sequence of a user in the period from the last time of opening the meter cover to the time of stealing electricity.

The first searching subunit is configured to search, from the power consumption data sequence, a time interval that satisfies a condition that, in a preset time period, the line loss power of the transformer area is suddenly increased and the number of times of simultaneous occurrence of the power consumption sudden drop of the user is higher than a preset threshold, and use the time interval as a first suspected power stealing time interval.

And the second searching subunit is used for screening and obtaining a time interval from the electricity stealing checking time to the last time of opening the meter cover from the electricity stealing checking time from the time when the user opens the meter cover.

The scanning subunit is used for carrying out sliding window scanning in a time interval from the electricity stealing checking time to the last time of the meter cover opening time from the electricity stealing checking time to obtain a time interval that the pearson correlation of the line loss data of all the station areas and the electricity consumption of the user is higher than a preset threshold value, and the time interval is used as a second suspected electricity stealing time interval.

And the merging subunit is used for merging the first suspected electricity stealing time and the second suspected electricity stealing time to obtain an electricity stealing time interval of the user.

In this embodiment, the specific execution process of the first query subunit, the first search subunit, the second search subunit, the scanning subunit, and the merging subunit may refer to the content of the method embodiment corresponding to fig. 2, which is not described herein again.

Optionally, in another embodiment of the present invention, an implementation manner of the second processing unit 603 includes:

and the judging subunit is used for judging the proportion of electricity larceny according to the electricity larceny instrument of the user if the electricity larceny method of the user is equal proportion electricity larceny.

And the calculating subunit is used for calculating the user electricity consumption data according to the proportion to obtain the average electricity stealing amount data of the user.

And the acquisition subunit is used for acquiring the electricity stealing application mode of the user if the electricity stealing method of the user is bypass electricity stealing.

And the second inquiry subunit is used for inquiring the average electricity consumption corresponding to the electricity stealing application mode to obtain the average electricity stealing amount data of the user.

In this embodiment, the specific execution process of the determining subunit, the calculating subunit, the obtaining subunit, and the second querying subunit may refer to the content corresponding to the foregoing method embodiment, which is not described herein again.

Another embodiment of the present application further provides a screening device for abnormal cases, as shown in fig. 7, which specifically includes:

the analysis unit 701 is configured to analyze data of each electricity stealing case sample for the electricity stealing case sample for which the electric charge service has been paid, so as to obtain sample feature data of each electricity stealing case sample; the sample characteristic data comprise electricity stealing methods, electricity consumption, line loss, average electricity stealing data, the ratio of live line current to neutral line current, the interval time between power failure events and power up events and the interval time between meter cover opening events and meter cover closing events of each sample case.

The dividing unit 702 is configured to divide the electricity stealing case samples by using a fuzzy clustering algorithm according to the sample feature data, so as to form a clustering result of one or more electricity stealing categories.

And the acquiring unit 703 is configured to obtain a standard additional charge of each electricity stealing category according to the clustering result.

The screening unit 704 is configured to screen out an abnormal case of the electric charge service by using the electric charge collected from the electric charge sample of the same electric charge collection type and the standard electric charge collected from the electric charge collection type; the electric charge for the additional charge of the electricity larceny case sample is calculated by any one of the above estimation methods for the additional charge.

In the screening device for abnormal cases provided in this embodiment, the analysis unit 701 analyzes the data of each electricity stealing case sample for the electricity stealing case sample for which the electric charge service has been paid, so as to obtain the sample feature data of each electricity stealing case sample. The dividing unit 702 then divides the electricity stealing case samples according to the sample feature data by using a fuzzy clustering algorithm to form a clustering result of one or more electricity stealing categories. The acquiring unit 703 obtains the standard additional charge of each electricity stealing category according to the clustering result. The screening unit 704 screens out abnormal cases of the electric charge service by using the electric charge recollection and standard electric charge recollection of the electric charge stealing case samples of the same electric charge stealing type; the electric charge recollection of the electric charge stealing case sample is calculated by the electric charge recollection estimation method. Therefore, the data calculated by the estimation method for the electric charge can be checked, and the accuracy is improved.

In this embodiment, the specific execution process of the analysis unit 701, the division unit 702, the acquisition unit 703 and the screening unit 704 can be referred to in the embodiment of the method corresponding to fig. 3, and will not be described herein again.

Optionally, in another embodiment of the present invention, an implementation manner of the screening unit 704 specifically includes:

and the calculating subunit is used for respectively differencing the electric charge collected by each electricity stealing case sample of the same electricity stealing category with the standard electric charge collected by the same electricity stealing category to obtain a first result value.

And the judging subunit is used for judging whether the electricity stealing case sample corresponding to the first result value belongs to the after-payment electricity fee business abnormal case if the first result value is larger than the preset threshold value.

In this embodiment, the specific execution process of the calculating subunit and the determining subunit may refer to the content corresponding to the foregoing method embodiment, which is not described herein again.

and the input subunit is used for inputting the standard additional charge of each electricity larceny type and the sample characteristic data of each electricity larceny case sample into the classifier.

The dividing subunit is used for dividing the electricity stealing case sample into a training set and a testing set.

The training subunit is used for training the classifier by using the electricity stealing case samples in the training set, and if the difference value between the additional charge of the electricity stealing case samples in the same electricity stealing category in the training set and the standard additional charge of the same electricity stealing category is larger than the threshold value, the electricity stealing case samples are set as additional charge business abnormal cases.

In this embodiment, the specific execution process of the input subunit, the dividing subunit, the training subunit, and the screening subunit may refer to the content of the method embodiment corresponding to fig. 4, which is not described herein again.

Optionally, in another embodiment of the present invention, the screening device for abnormal cases may further include:

In this embodiment, the specific execution process of the evaluation unit may refer to the content corresponding to the above-mentioned method embodiment, and will not be described herein again.

In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. In particular, for a system or system embodiment, since it is substantially similar to a method embodiment, the description is relatively simple, with reference to the description of the method embodiment being made in part. The systems and system embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.

Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative elements and steps are described above generally in terms of functionality in order to clearly illustrate the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An estimation method for electric charge collection is characterized by comprising the following steps:

Combining the first suspected electricity larceny time and the second suspected electricity larceny time to obtain an electricity larceny time interval of the user;

2. The method of claim 1, wherein obtaining average electricity larceny data for the user based on electricity larceny techniques for the user comprises:

3. The screening method for the abnormal cases is characterized by comprising the following steps of:

screening out an electric charge tracing service abnormal case by using the electric charge tracing sample of the electric charge tracing case of the same electric charge tracing type and the standard electric charge tracing; wherein the electric charge for the additional charge of the electricity theft case sample is calculated by the method of any one of claims 1 to 2; the electric charge tracing service abnormal case is an electric charge tracing case sample in which the actual electric charge tracing and the standard electric charge tracing of the electric charge tracing class differ by more than one standard value.

4. The method of claim 3, wherein the screening out additional electric charge business anomaly cases using the additional electric charge and the standard additional electric charge for the electricity theft case samples of the same electricity theft category comprises:

5. The method of claim 3, wherein the screening out additional electric charge business anomaly cases using the additional electric charge and the standard additional electric charge for the electricity theft case samples of the same electricity theft category comprises:

training a classifier by using the electricity stealing case samples in the training set, and setting the electricity stealing case samples as abnormal cases of the service of the electric charge to be paid if the difference value between the electric charge to be paid by the electricity stealing case samples in the same electricity stealing category in the training set and the standard electric charge to be paid by the electricity stealing category is larger than a threshold value;

6. The method as recited in claim 4, further comprising:

7. An estimation device for a pay-after-call, comprising:

the second calculation unit is used for calculating the additional charge corresponding to the electricity larceny quantity of the user in the electricity larceny time interval according to a preset electricity charge calculation method;

Wherein the first processing unit includes:

8. An abnormal case screening device, comprising:

the screening unit is used for screening out the additional charge service abnormal cases by utilizing the additional charge and the standard additional charge of the electricity stealing case samples of the same electricity stealing category; wherein the electric charge for the additional charge of the electricity theft case sample is calculated by the method of any one of claims 1 to 2; the electric charge tracing service abnormal case is an electric charge tracing case sample in which the actual electric charge tracing and the standard electric charge tracing of the electric charge tracing class differ by more than one standard value.