CN113610409A

CN113610409A - Early warning method and device for electric charge recovery risk

Info

Publication number: CN113610409A
Application number: CN202110925816.5A
Authority: CN
Inventors: 陈雨泽; 袁葆; 张文; 陈雁; 张静; 赵冠东; 张明珠; 李国强; 周春; 刘大鹏; 欧阳红
Original assignee: Beijing China Power Information Technology Co Ltd
Current assignee: Beijing China Power Information Technology Co Ltd
Priority date: 2021-08-12
Filing date: 2021-08-12
Publication date: 2021-11-05

Abstract

The application provides an electric charge recovery risk early warning method and device, and an industry development trend prediction value of the industry where a user is located in the current charging month is obtained based on an industry development trend prediction model; acquiring the electricity charge amount of a previous charge month of a user, historical electricity charge payment data and business expansion transaction data; all the acquired data are used as input of an electric charge recovery risk early warning model, and a probability value of electric charge late payment generated in the current charging month is output; and if the probability value is within the risk early warning range, generating early warning information. According to the scheme, the probability value of the electricity charge late payment generated by the user in the current charging month is predicted by utilizing the industry development trend prediction model and the electricity charge recovery risk early warning model, the industry development trend prediction value of the industry where the user is located, information such as historical electricity charge payment data and industry expansion electricity transaction data of the user are comprehensively considered, the user is objectively and comprehensively identified, and the responsible personnel are reminded to take corresponding measures when the user is identified to be the user with high electricity charge recovery risk.

Description

Early warning method and device for electric charge recovery risk

Technical Field

The invention relates to the technical field of data processing, in particular to an electric charge recovery risk early warning method and device.

Background

For power grid enterprises, the recovery of electric charges is an important work, and is closely related to the normal production and operation of the enterprises. In order to ensure that the electric charge recovery work is better developed, the users with the electric charge recovery risk can be identified in advance, so that the enterprise is warned, the enterprise can conveniently make response measures in advance, and the loss of enterprise operation is reduced.

At present, the purpose of identifying the users with the risk of recovering the electric charge is achieved through the following modes: setting a plurality of judgment conditions for historical electricity consumption and electricity charge payment conditions such as electricity charge, electricity consumption safety, arrearage, default money, power failure and the like of a user according to human experience, and identifying the user with the risk of electricity charge recovery according to the set judgment conditions.

Therefore, the conventional experience-based electric charge recovery risk early warning method is too subjective, and therefore the identification of the user with the electric charge recovery risk is insufficient.

Disclosure of Invention

In view of this, embodiments of the present invention provide an electric charge recycling risk early warning method and apparatus, so as to solve the problem in the prior art that identification of a user with an electric charge recycling risk is not sufficient.

In order to achieve the above purpose, the embodiments of the present invention provide the following technical solutions:

on one hand, the embodiment of the invention provides an electric charge recovery risk early warning method, which comprises the following steps:

acquiring an industry development trend prediction value of the industry where a user is in the current charging month based on a pre-constructed industry development trend prediction model, wherein the industry development trend prediction model is constructed based on industry historical electricity sales development trend characteristic data and industry economic indicators;

acquiring the electricity charge amount of a previous charge month of a user, historical electricity charge payment data and business expansion transaction data;

taking the electric charge amount of the previous charging month of the user, historical electric charge payment data, business expansion electric data and the industry development trend prediction value as the input of a pre-constructed electric charge recovery risk early warning model, predicting the probability value of the electric charge late payment generated in the current charging month in the electric charge recovery risk early warning model, and outputting the probability value;

and if the probability value is within the risk early warning range, generating early warning information for indicating that the electric charge recovery risk is high.

Optionally, the process of constructing the industry development trend prediction model includes:

extracting power selling trend characteristic data in industry historical power selling data in a set period according to a season adjustment algorithm, wherein the set period takes monthly degrees as a unit;

acquiring a leading index and an instantaneity index related to the industry where a user is located, wherein the leading index and the instantaneity index are obtained by analyzing the correlation of power selling trend characteristic data and industry economic indexes of various industries, the leading index at least comprises industry expansion net capacity, a manufacturing Procurement Manager Index (PMI) and a non-manufacturing PMI, and the instantaneity index at least comprises industry domestic total production value (GDP) acceleration and industry added value acceleration;

performing decorrelation calculation on the antecedent indexes and the instantaneity indexes according to a principal component analysis algorithm to obtain decorrelation indexes;

and constructing an industry development trend prediction model by utilizing a Support Vector Machine (SVM) regression algorithm based on the electricity sales trend feature data and the correlation index.

Optionally, the process of obtaining the lead indicator and the instantaneity indicator by analyzing the correlation between the electricity sales trend characteristic data of each industry and the industry economic indicator includes:

extracting power selling trend characteristic data in historical power selling data of each industry in a set period according to a season adjustment algorithm;

acquiring industrial economic indexes of various industries;

analyzing the correlation between the power selling trend characteristic data and the industry economic indexes corresponding to each industry to obtain leading indexes and instantaneity indexes of each industry;

the front guiding indexes are analyzed according to a dynamic time warping algorithm, and monthly numbers influencing different front guiding indexes are obtained.

Optionally, the construction process of the electric charge recycling risk early warning model includes:

determining a sample charging month, and acquiring a sample industry development trend prediction value of the industry where a user is in the sample charging month based on a pre-constructed industry development trend prediction model;

determining monthly electric charge payment sample frequency of the user based on the electric charge payment times of the user in a set period;

calculating the difference value of the user electric charge payment time and the electric charge issuing time of the previous charging month of the sample charging month, and taking the difference value as the user electric charge payment sample time length of the previous charging month;

determining the monthly capacity sample variation of the user in the previous charging month by using the user business expansion capacity completion capacity, the user business expansion capacity reduction completion capacity, the user business expansion pause capacity and the user business expansion recovery capacity in the previous charging month of the sample charging month;

acquiring historical electricity charge late payment generation times of a user before the sample charging month and previous electricity charge late payment generation time;

and taking the electric charge amount generated in the previous charging month of the sample charging months, the sample industry development trend prediction value, the user monthly electric charge payment sample frequency, the user electric charge payment sample duration, the user monthly capacity sample variation, the historical electric charge late payment generation times and the previous electric charge late payment generation time as inputs, taking the probability of generating the electric charge late payment in the sample charging months as an output, and constructing an electric charge recovery risk early warning model by utilizing an SVM (support vector machine) regression algorithm.

Optionally, the method further includes:

and if the probability value is not in the risk early warning range, generating prompt information for indicating that the electric charge recycling risk is low.

On the other hand, the embodiment of the invention provides an electric charge recovery risk early warning device, which comprises:

the industry development trend prediction model is used for predicting an industry development trend prediction value of the industry where the user is located in the current charging month, and is constructed on the basis of industry historical electricity sales development trend characteristic data and industry economic indicators;

the acquisition module is used for acquiring the electric charge amount of the user in the previous charging month, historical electric charge payment data and business expansion transaction data;

the electric charge recovery risk early warning model is used for predicting historical electric charge payment data, business expansion electric data and the industry development trend prediction value based on the received electric charge amount of the user in the previous charging month, and outputting a probability value of electric charge late payment generated in the current charging month;

and the early warning module is used for generating early warning information for indicating that the electric charge recovery risk is high if the probability value is within the risk early warning range.

Optionally, the method further includes: a first building block;

the first building block comprises:

the first extraction unit is used for extracting the power selling trend characteristic data in industry historical power selling data in a set period according to a season adjustment algorithm, wherein the set period takes monthly degrees as a unit;

the system comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring a leading index and an instantaneity index related to the industry where a user is located, the leading index and the instantaneity index are obtained by analyzing the correlation of power selling trend characteristic data and industry economic indexes of various industries, the leading index at least comprises industry net augmentation capacity, a manufacturing industry purchase management index PMI and a non-manufacturing industry PMI, and the instantaneity index at least comprises industry domestic production total value GDP acceleration and industry augmentation value acceleration;

the decorrelation unit is used for performing decorrelation calculation on the antecedent indexes and the instantaneity indexes according to a principal component analysis algorithm to obtain decorrelation indexes;

and the first construction unit is used for constructing an industry development trend prediction model by utilizing a Support Vector Machine (SVM) regression algorithm based on the power selling amount trend characteristic data and the correlation index.

Optionally, the method further includes: an analysis module;

the analysis module includes:

the second extraction unit is used for extracting the power selling trend characteristic data in the historical power selling data of each industry in the set period according to the season adjustment algorithm;

the second acquisition unit is used for acquiring the industrial economic indexes of each industry;

the analysis unit is used for analyzing the correlation between the electricity selling trend characteristic data and the industry economic indicators corresponding to each industry to obtain the leading indicators and the instantaneity indicators of each industry; the front guiding indexes are analyzed according to a dynamic time warping algorithm, and monthly numbers influencing different front guiding indexes are obtained.

Optionally, the method further includes: a second building block;

the second building block comprises:

the third obtaining unit is used for determining a sample charging month and obtaining a sample industry development trend prediction value of the industry where the user is located in the sample charging month based on a pre-constructed industry development trend prediction model; determining monthly electric charge payment sample frequency of the user based on the electric charge payment times of the user in a set period; acquiring historical electricity charge late payment generation times of a user before the sample charging month and previous electricity charge late payment generation time;

the calculating unit is used for calculating the difference value between the user electricity fee payment time and the electricity fee issuing time of the previous charging month of the sample charging month, and taking the difference value as the user electricity fee payment sample time length; determining the monthly capacity sample variation of the user by using the user business expansion capacity completion capacity, the user business expansion capacity reduction completion capacity, the user business expansion pause capacity and the user business expansion recovery capacity of the previous charging month of the sample charging month;

and the second construction unit is used for taking the electric charge amount generated in the previous charging month of the sample charging month, the sample industry development trend prediction value, the monthly electric charge payment sample frequency of the user, the electric charge payment sample duration of the user, the monthly capacity sample variation of the user, the historical electric charge late payment generation times and the previous electric charge late payment generation time as input, taking the probability of generating the electric charge late payment in the sample charging month as output, and constructing an electric charge recovery risk early warning model by utilizing an SVM (support vector machine) regression algorithm.

Optionally, the early warning module is further configured to generate a prompt message for indicating that the electric charge recycling risk is low if the probability value is not within the risk early warning range.

Based on the electric charge recovery risk early warning method and device provided by the embodiment of the invention, the industry development trend prediction value of the industry where the user is located in the current charging month is obtained based on the pre-constructed industry development trend prediction model, and the industry development trend prediction model is constructed based on the industry historical electricity sales development trend characteristic data and the industry economic index; acquiring the electricity charge amount of a previous charge month of a user, historical electricity charge payment data and business expansion transaction data; taking the electric charge amount of the previous charging month of the user, historical electric charge payment data, business expansion electric data and the industry development trend prediction value as the input of a pre-constructed electric charge recovery risk early warning model, predicting the probability value of the electric charge late payment generated in the current charging month in the electric charge recovery risk early warning model, and outputting the probability value; and if the probability value is within the risk early warning range, generating early warning information for indicating that the electric charge recovery risk is high. According to the scheme provided by the embodiment of the invention, the probability value of the electricity charge late payment generated by the user in the current charging month is predicted by utilizing the pre-constructed industry development trend prediction model and the electricity charge recovery risk early warning model, and the information such as the industry development trend prediction value of the industry where the user is located, the historical electricity charge payment data and the industry expansion electricity transaction data of the user are comprehensively considered during prediction, so that the user with the electricity charge recovery risk is objectively and comprehensively identified from multiple dimensions, and when the user is identified as a user with a high electricity charge recovery risk, relevant responsible personnel are reminded, so that the relevant responsible personnel can formulate a corresponding electricity charge recovery scheme, and the loss of the electric power enterprise operation is avoided.

Drawings

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

Fig. 1 is a schematic flow chart of an electric charge recycling risk early warning method according to an embodiment of the present invention;

FIG. 2 is a schematic flow chart of building an industry development trend prediction model according to an embodiment of the present invention;

fig. 3 is a schematic flow chart for analyzing the correlation between the power selling trend characteristic data of each industry and the industry economic indicators to obtain a leading indicator and an instantaneity indicator according to the embodiment of the present invention;

FIG. 4 is a schematic diagram of a hard- ε band hyperplane as defined in an SVM regression algorithm;

FIG. 5 is a diagram illustrating a method for solving a positive point set D in an SVM regression algorithm⁺And negative class point set D^-Schematic diagram of the classification planes involved in the binary problem of (1);

fig. 6 is a schematic flow chart of constructing an electric charge recycling risk early warning model according to an embodiment of the present invention;

fig. 7 is a block diagram of a structure of an electric charge recycling risk early warning apparatus according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

In this application, 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 an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.

The terms "first," "second," "third," "fourth," and the like in the description and in the claims of the present application and in the drawings described above, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein.

According to the background art, the existing experience-based electric charge recovery risk early warning mode is too subjective, so that the identification of users with electric charge recovery risks is insufficient.

Therefore, the embodiment of the invention provides an electric charge recovery risk early warning method and device, and aims to solve the problem that in the prior art, identification of a user with an electric charge recovery risk is insufficient.

The following terms are specifically referred to in the embodiments of the present invention:

risk of electric charge recovery: the method refers to the risks that the power consumption enterprise is shut down, bankruptcy, restructured and converted, the power consumption enterprise has bad operation conditions, the power consumption enterprise has shortage of mobile funds, the power consumption enterprise is converted into leases, governments are removed, the social stability and the like, and the power charge cannot be timely recovered.

Selling electric quantity: the electric quantity sold outside by the electric power enterprise and the sales income can be obtained according to the electric quantity, including the electric quantity sold to the user for direct consumption and the electric quantity sold to other electric power enterprises by wholesale, and the electric quantity supplied by the electric power enterprise to non-electric power production, basic construction and non-production departments and the like of the enterprise.

And E, business expansion and installation: the system is a general name of the service flow of the power supply department in the whole process from power application of the power users to actual power utilization. The method mainly comprises new installation capacity reduction, sales capacity reduction, suspension recovery and the like. The net incremental capacity is equal to the new capacity minus the sales capacity.

The embodiment of the invention provides an electric charge recovery risk early warning method and device, and the detailed description is given through specific embodiments.

Referring to fig. 1, a schematic flow chart of an electric charge recycling risk early warning method according to an embodiment of the present invention is shown.

The electric charge recovery risk early warning method comprises the following steps:

s101: and acquiring an industry development trend prediction value of the industry where the user is in the current charging month based on a pre-constructed industry development trend prediction model.

In S101, an industry development trend prediction model is constructed based on industry historical electricity sales development trend characteristic data and industry economic indicators, and then the industry development trend prediction model can be used for predicting an industry development trend prediction value of the industry where the user is located in the current charging month, so that the condition that the electricity sales of the industry where the user is located in the current charging month is influenced by the industry development trend is obtained, and the electricity sales recovery risk of the user is predicted subsequently.

It should be noted that the historical electricity sales quantity development trend characteristic data of the industry refers to data reflecting the electricity sales quantity development trend characteristic of the industry in the historical electricity sales quantity data of the industry. The industrial economic index refers to an index reflecting the development condition of industrial economy.

In the specific implementation process of S101, historical electricity selling quantity development trend characteristic data of the industry where the user is located before the current charging month and an industry economic index corresponding to the industry where the user is located are obtained, and the historical electricity selling quantity development trend characteristic data and the industry economic index are processed together based on the industry development trend prediction model to obtain an industry development trend prediction value of the industry where the user is located in the current charging month.

S102: and acquiring the electricity charge amount of the user in the previous charging month, historical electricity charge payment data and business expansion transaction data.

In S102, the historical electricity fee payment data includes monthly electricity fee payment frequency of the previous consumption month of the user, electricity fee issuance time of the previous charge month of the user, electricity fee payment time of the previous charge month of the user, historical electricity fee late fee generation times of the user, and previous electricity fee late fee generation time of the user.

The monthly electric charge payment frequency of the previous consumption month of the user refers to the times of the user paying the electric charge in the previous consumption month.

The electricity fee issuance time of the previous charge month of the user refers to the time when the electricity company bills the electricity fee used by the user in the previous charge month.

The user electric charge payment time refers to the time of the user for paying the electric charge in the previous charging month.

The generation times of the historical electric charge late fees of the users refer to the times of the users for paying the electric charge late fees together at present.

The previous generation time of the electric charge late payment of the user refers to the time when the user last paid the electric charge late payment.

The business expansion and power transaction data comprise user business expansion and capacity completion capacity, user business expansion and capacity reduction completion capacity, user business expansion suspension capacity and user business expansion recovery capacity.

The capacity for completing the expansion capacity of the user industry refers to the capacity for increasing the power consumption agreed by contracts applied by the user to the power enterprises; the capacity expansion and reduction completion of the user industry refers to the fact that a user applies for reducing the power consumption capacity agreed by a contract to a power enterprise; the user business expansion capacity refers to the condition that a user applies for temporarily stopping all or part of the power consumption capacity of the powered equipment to the power enterprise; the user expansion recovery capacity refers to the condition that a user applies to a power enterprise to recover all or part of the power utilization capacity of the powered equipment.

In the process of implementing S102 specifically, corresponding data may be acquired from the power marketing internal system, specifically, the amount of the electric charge of the user in a previous charging month of the current charging month, monthly electric charge payment frequency of the user in the previous charging month, electric charge issuance time of the user in the previous charging month, electric charge payment time of the user in the previous charging month, user industry expansion capacity completion capacity, user industry expansion suspension capacity, and user industry expansion recovery capacity of the user before the current charging month are obtained, the number of times the user has paid the electric charge late fee before the current charging month is obtained, and the previous electric charge late fee generation time of the user is obtained.

S103: and (3) taking the electric charge amount of the previous charging month of the user, historical electric charge payment data, business expansion electric data and an industry development trend prediction value as the input of a pre-constructed electric charge recovery risk early warning model, predicting the probability value of the electric charge late payment generated in the current charging month in the electric charge recovery risk early warning model, and outputting the probability value.

In S103, the electricity amount of the previous charging month, the historical electricity payment data, and the business expansion transaction data of the user may be obtained by the power marketing internal system. Particularly, in the power marketing internal system, the relevant content is recorded in the mode of table 1.

Table 1:

each item in table 1 is recorded in units of months.

Wherein, the monthly electric charge payment frequency factor f of the user can be obtained according to the record of the payment electric charge_m。

According to the electric charge payment time t_pAnd the electricity charge issuing time t_rThe difference can obtain the user electric charge payment duration d_pAs shown in equation (1).

d_p＝t_p-t_r (1)

According to recorded business expansion transaction data, i.e. using business expansion capacity to complete capacity c_aVolume expansion and reduction completion capacity c_rBusiness expansion pause capacity c_sCapacity of Heyu-expanded recovery c_cThe user monthly capacity change Δ c can be obtained as shown in equation (2).

Δc＝c_a+c_r-c_s-c_c (2)

In the process of specifically implementing S103, the industry development trend prediction value obtained based on executing S101, the electricity fee amount of the previous charging month of the user obtained based on executing S102, the historical electricity fee payment data, and the business expansion transaction data are input as a pre-constructed electricity fee recovery risk early warning model, prediction processing is performed in the pre-constructed electricity fee recovery risk early warning model, and a probability value of electricity fee late payment generated in the current charging month is output.

Wherein, the output probability value comprises two conditions: one is a probability value within the risk pre-warning range, and the other is a probability value not within the risk pre-warning range.

It should be noted that the risk early warning range may be set in advance by a relevant responsible person according to an actual scene application requirement, and no specific numerical limitation is made herein.

If the probability value output by executing the step S103 is a probability value within the risk pre-warning range, the step S104 is continuously executed.

And if the probability value output by executing the step S103 is the probability value which is not in the risk early warning range, directly ending the process. Optionally, the process may not be finished, and prompt information for indicating that the electric charge recovery risk is low is generated to remind the relevant responsible person that the electric charge recovery risk of the current user is low, so that a reference basis is provided for the relevant responsible person to perform subsequent electric charge recovery work.

S104: and generating early warning information for indicating that the electric charge recycling risk is high.

In the specific implementation process of S104, early warning information indicating that the electric charge recovery risk is high is generated, so as to remind relevant responsible personnel that the electric charge recovery risk of the current user is high, and provide a reference for the relevant responsible personnel to perform subsequent electric charge recovery work.

Based on the electric charge recovery risk early warning method provided by the embodiment of the invention, the probability value of the electric charge late payment generated by the user in the current charging month is predicted by utilizing the pre-constructed industry development trend prediction model and the electric charge recovery risk early warning model, and the information such as the industry development trend prediction value of the industry where the user is located, the historical electric charge payment data and the industry expansion electric data of the user are comprehensively considered during prediction, so that the user with the electric charge recovery risk is objectively and comprehensively identified from multiple dimensions, and when the user is identified as the user with high electric charge recovery risk, relevant responsible personnel are reminded, so that the relevant responsible personnel can formulate a corresponding electric charge recovery scheme, and the loss of the operation of an electric power enterprise is avoided.

Based on the electric charge recycling risk early warning method provided by the embodiment of the invention, the construction process of the industry development trend prediction model which is constructed in advance in the step S101 is shown in FIG. 2. The method mainly comprises the following steps:

s201: and extracting the power selling trend characteristic data in the industry historical power selling data in a set period according to a season adjustment algorithm.

In S201, the season adjustment algorithm is a method of calculating a trend term for the original time series by using a moving average method, calculating a season term for the original time series based on a 3 × 3 moving average method, and subtracting a sum of the trend term and the season term from the original time series to obtain a random term.

The trend item reflects long-term trend variation of the time sequence, the season item reflects seasonal periodic variation of the time sequence in the same month in different years, and the random item reflects other irregular variation of the time sequence, such as weather of a non-seasonal item.

One of the season adjustment algorithms that is commonly used at present is the X13 season adjustment algorithm, which uses a centralized moving weighted average method to perform item-by-item decomposition, and completes each component sequence through multiple iterations and decomposition.

It should be noted that, the historical electricity sales data of the industry are influenced by factors such as economic development, seasonal changes and temperature. The industry historical electricity selling amount data can be decomposed into three subsequences of trend items of the industry historical electricity selling amount data, season items of the industry historical electricity selling amount data and random items of the industry historical electricity selling amount data by using an X13 season adjusting algorithm.

Specifically, the relationship between the trend item of the industry historical electricity sales data, the season item of the industry historical electricity sales data, the random item of the industry historical electricity sales data and the industry historical electricity sales data is shown in formula (3).

Q(i)＝Q^t(i)+Q^s(i)+Q^r(i) (3)

Wherein i represents monthly degree and can represent i month, Q (i) represents industry historical selling electricity quantity data, and Q^t(i) A trend term, Q, representing historical sales electricity data for the corresponding industry^s(i) Seasonal item, Q, representing historical electricity sales data for the corresponding industry^r(i) And random items representing historical electricity sales data of the corresponding industry.

Specifically, according to the difference of i, the industry historical electricity sales data can be expressed as: { Q (i) | i ∈ 1, 2.

The trend item of the industry historical electricity sales data can be expressed as: { Q^t(i)|i∈1，2，...，n}。

The seasonal item of industry historical electricity sales data can be expressed as: { Q^s(i)|i∈1，2，...，n}。

The random item of industry historical sales electricity data can be expressed as: { Q^r(i)|i∈1，2，...，n}。

Wherein the value of n is a positive integer greater than 1.

In the formula (3), the trend item of the industry historical electricity sales data represents the condition that the industry historical electricity sales data is influenced by economic development factors, namely electricity sales trend characteristic data in the industry historical electricity sales data.

The seasonal item of the industry historical electricity selling quantity data represents the condition that the industry historical electricity selling quantity data is influenced by seasonal variation factors, namely electricity selling quantity seasonal characteristic data in the industry historical electricity selling quantity data.

The random item of the industry historical electricity selling amount data represents the condition that the industry historical electricity selling amount data is influenced by other factors except for economic development factors and seasonal variation factors, namely the random electricity selling amount characteristic data in the industry historical electricity selling amount data.

In S201, the set period is in units of months. The set period can be preset by a human, for example, the set period can include all months before the current consumption month in the industry where the user is.

In the process of implementing S201 specifically, for each industry, the industry historical electricity sales data is decomposed by using a seasonal adjustment algorithm in a set period, so that electricity sales trend characteristic data corresponding to the industry historical electricity sales data can be obtained.

In a specific implementation, the seasonal adjustment algorithm used when the historical electricity consumption data is decomposed may be an X13 seasonal adjustment algorithm, or may be other seasonal adjustment algorithms, which is not limited herein.

S202: and acquiring a leading index and an instantaneity index related to the industry of the user.

In S202, the leading performance index and the instantaneity index are obtained by analyzing the correlation between the power selling trend characteristic data of each industry and the industry economic index.

The leading index is an index which can be used for analyzing the influence of the current influence factors on the trend development of the electricity selling amount of the user in the industry of the future charge month.

The instantaneity index is an index which can be used for analyzing the trend development of the current charging monthly electricity selling amount of the industry where the current influence factors influence users.

As shown in fig. 3, a schematic flow chart for obtaining a leading performance index and a timeliness index based on analyzing the correlation between the power selling trend characteristic data and the industry economic index of each industry mainly includes the following steps:

s301: and extracting the power selling trend characteristic data in the historical power selling data of each industry in a set period according to a seasonal adjustment algorithm.

In the specific implementation process of B1, for each industry, extracting the power selling trend characteristic data in the historical power selling data of the industry in a set period according to a season adjustment algorithm. For a specific extraction process, reference may be made to the related content in a1, and details are not repeated here.

S302: and acquiring the industrial economic indexes of various industries.

In S302, the trade economic indicator refers to an indicator related to trade economic development.

S303: and analyzing the correlation between the power selling trend characteristic data and the industry economic indexes corresponding to each industry to obtain the leading indexes and the instantaneity indexes of the respective industries.

In the specific implementation process of S303, based on the power selling trend characteristic data corresponding to each industry obtained by executing S303 and based on the industry economic indicators corresponding to each industry obtained by executing S302, for each industry, by analyzing the correlation between the power selling trend characteristic data corresponding to the industry and the industry economic indicators corresponding to the industry, a leading indicator and an instantaneity indicator of the industry are obtained.

That is, for each industry, according to the correlation analysis between the characteristic data of the electricity selling quantity trend in the historical electricity selling quantity data of the respective industry and the economic indicators of the industry, the instantaneity indicators and the leading indicators which need to be considered when the development trend of the electricity selling quantity of the corresponding industry is predicted can be obtained.

Further, the leading index is analyzed according to a Dynamic Time Warping (DTW) algorithm, and the number of months affecting different leading indexes can be obtained.

The DTW algorithm measures two time sequences of different lengths, X ═ X₁,…,x_i,…,x_m) And Y ═ Y₁,…,y_j,…,y_n) The method of similarity of (1).

The similarity between two time series is calculated by extending and shortening the time series. Dynamic Time Warping (DTW) is a typical optimization problem, and when solving the matching between two templates, the warping function corresponding to the minimum cumulative distance is expressed as the warping path distance between the points on the sequence and the point on the sequence, as shown in equation (4):

D(i,j)＝Dist(i,j)+min{D(i-1,j),D(i,j-1),D(i-1,j-1)} (4)

where Dist (i, j) is the Euclidean distance between point i on X and point j on Y.

As shown in table 2, the instantaneous indexes and the lead indexes are required to be considered when predicting the power sale development trend in each industry at present.

Table 2:

wherein the leading indicators at least comprise industry net amplification capacity, manufacturing PMI (Purchasing Managers' Index) and non-manufacturing PMI.

The instantaneity index at least includes the increase in industrial GDP (Gross Domestic Product) and increase in industrial added value.

The "√" in table 2 indicates that the index exists in the current industry.

Further, the method can be obtained by analyzing the leading performance index in the table 2 according to the dynamic time warping algorithm:

the monthly number of the leader indexes before the business net capacity expansion of each industry is influenced is 3 months, the monthly number of the PMI leader indexes of the manufacturing industry influencing the industrial industry is 3 months, and the monthly number of the PMI leader indexes of the non-manufacturing industry influencing the traffic industry, the information industry, the batch release industry, the lodging industry, the financial industry, the real estate industry, the lease industry and the public industry is 6 months.

The month number of PMI leading indexes affecting the manufacturing industry of the industrial industry is taken as 3 months for example. For example, the PMI leader indicator for the industry of 5 months in 2021 may affect the electricity sales data for the industry of 8 months in 2021.

It should be noted that other industries, other leading indicators and different monthly numbers may refer to the above examples, and are not necessarily described.

S203: and performing decorrelation calculation on the precursor index and the instantaneity index according to a principal component analysis algorithm to obtain a decorrelation index.

It should be noted that the Principal Component Analysis (PCA) algorithm refers to a statistical method that tries to recombine original variables into a new set of several independent synthetic variables, and can extract as many information of the original variables as possible from the new set of synthetic variables as needed, and is also a method used mathematically for reducing dimensions.

In the process of implementing S203 specifically, based on the leading performance index and the instantaneity index related to the industry where the user is located obtained by executing S202, decorrelation calculation is performed on the leading performance index and the instantaneity index by using a principal component analysis algorithm, so as to obtain a decorrelation index.

Specifically, a formula (5) is used for performing decorrelation on the leading performance index and the instantaneity index related to the industry where the user is located to obtain a decorrelation index P (i).

P(i)＝pca(E(i),G(i)),i＝1,2,...,n (5)

Where e (i) represents an instantaneity index, which may also be expressed as { e (i) | i ∈ 1, 2.

G (i) represents a leading index, which may be expressed as: g (i) { g (i) | i ∈ 1, 2.

It should be noted that if the industry does not have the instantaneity index, only the decorrelation calculation is performed on the leading index. And vice versa.

S204: and constructing an industry development trend prediction model by utilizing a Support Vector Machine (SVM) regression algorithm based on the power selling trend characteristic data and the decorrelation indexes.

In the process of specifically realizing a4, based on the power sale trend characteristic data obtained by executing S201 and based on the decorrelation index obtained by executing S203, an industry development trend prediction model is constructed through an SVM regression algorithm, and is specifically represented as the following formula:

wherein the content of the first and second substances,

expressing the industry development trend predicted value, Q, of the industry of the user in the ith month^t(i-1) shows the power selling trend characteristic data of the industry where the user is located in the i-1 th month, Q^t(i-2),...,Q^t(i-12) and so on, and will not be described herein.

It should be noted that, the SVM (support vector machine) regression algorithm: is the application of the support vector in the field of function regression. The sample points of the SVM regression are of only one type, and the optimal hyperplane sought is to minimize the "total deviation" of all sample points from the hyperplane. At this time, the sample points are all between two boundary lines, and the optimal regression hyperplane is equivalent to the maximum interval.

Given a data set D:

the essence of the regression problem is to find the function f (x) in order to infer the y value for any pattern x.

The SVM regression is determined by defining a hard-epsilon band hyperplane (as shown in fig. 4), i.e., all sample points in the data set D satisfy:

-ε≤y_i-f(x_i)≤ε,i＝1,2...N (7)

the regression problem was converted to find the optimal hard-epsilon band hyperplane, i.e.:

minε

f(x)

s.t.-ε≤y_i-f(x_i)≤ε,i＝1,2...N (8)

as shown in FIG. 5, solving the optimization problem can be converted into solving a set of normal points D⁺And negative class point set D^-The two-classification problem of (a), wherein,

in the embodiment of the invention, an industry development trend prediction model is constructed based on the power selling trend characteristic data corresponding to each industry and the lead performance index and the instantaneity index corresponding to each industry, so that the industry development trend prediction model is used for predicting the industry development trend prediction value of the industry where the user is located, and further the electricity charge recovery risk of the user is predicted in the follow-up process, and the relevant responsible personnel are prompted when the electricity charge recovery risk of the user is determined to be high, so that the relevant responsible personnel can make a corresponding electricity charge recovery scheme, and the loss of the electric power enterprise operation is avoided.

The process of constructing the industry development trend prediction model, based on the disclosure of FIG. 2, is illustrated here.

Assuming that the industry where the user is located is industry, as can be seen from table 1, there are 2 leading indicators and 2 instantaneity indicators among the influencing factors related to the industry.

The process of constructing the industrial development trend prediction model comprises the following steps:

firstly, extracting the power selling trend characteristic data in the industrial historical power selling data in a set period according to a season adjusting algorithm. Is recorded as: { Q₁ ^t(i)|i∈1,2,...,n}。

Next, 2 instantaneity indexes related to industry are determined from table 1 and are denoted as formula (11):

E₁(i)＝{E¹(i),E²(i)|i∈1,2,...,n} (11)

wherein E is¹(i) Showing the GDP acceleration of the second industry, E²(i) Indicating an increased value and speed of the industry.

The 2 industry-relevant leading indicators identified in table 1 are given by formula (12):

G₁(i)＝{G¹(i),G²(i)|i∈1,2,...,N} (12)

wherein G is¹(i) Indicating the net amplification capacity, G, corresponding to the industry²(i) And representing the manufacturing PMI corresponding to the industrial industry.

As can be seen from the above, since the monthly index of the pre-capacity for industrial amplification and the monthly index of PMI precursor in manufacturing industry are both 3 months, the index is expressed by the following equations (13) and (14):

G¹(i)＝{net(i-1),net(i-2),...,net(i-3)} (13)

G²(i)＝{PMI(i-1),PMI(i-2),...,PMI(i-3)} (14)

wherein net (i-1) represents the industry expansion capacity trend item of the industrial industry in the i-1 th month, and PMI (i-1) represents the manufacturing PMI of the industrial industry in the i-1 th month.

From the formulas (13) and (14), it can be understood that the influence of the industry net augmented capacity on the industry trend of the current consumption month is obtained based on the industry net augmented capacity trend term of the first three months of the current consumption month, and the influence of the industry PMI on the industry trend of the current consumption month is obtained based on the manufacturing PMI of the first three months of the current consumption month.

Then, according to the formula (5), decorrelation calculation is performed on the leading performance index represented by the formula (12) and the instantaneity index represented by the formula (11), so as to obtain a decorrelation index P₁(i) In that respect As shown in equation (15).

P₁(i)＝pca(E₁(i),G₁(i)),i＝1,2,...,n (15)

Finally, performing a SVM regression algorithm and a decorrelation index P₁(i) And establishing an industrial development trend prediction model. As shown in equation (16).

Wherein the content of the first and second substances,

expressing the industry development trend prediction value, Q, of the user in the ith month of the industrial industry₁ ^t(i-1) shows the power selling trend characteristic data of the user in the i-1 th month of the industrial industry, Q₁ ^t(i-2)，...，Q₁ ^t(i-12) and so on, and will not be described herein.

Based on the electric charge recovery risk early warning method provided by the embodiment of the invention and the industry development trend prediction model pre-constructed based on fig. 2, as shown in fig. 6, a flow diagram for constructing the electric charge recovery risk early warning model disclosed by the embodiment of the invention is shown. The method mainly comprises the following steps:

s601: and determining a sample charging month, and acquiring a sample industry development trend prediction value of the industry where the user is in the sample charging month based on a pre-constructed industry development trend prediction model.

In S601, a month with a known probability of generating an electric charge late fee can be selected as a sample consumption month by a person according to actual scene application needs.

In the process of specifically implementing S601, a sample industry development trend prediction value of the industry where the user is located in the sample charging month may be obtained according to the formula (6).

S602: and determining monthly electric charge payment sample frequency of the user based on the electric charge payment times of the user in the set period.

In S602, the monthly electric charge payment sample frequency of the user is obtained by monthly statistics of the electric charge payment times.

Specifically, in the process of implementing S602, the monthly subscription frequency is used as the frequency, and the number of times of electricity fee payment data generated when the user transacts the electricity fee payment service each time in the set period is counted to obtain the monthly electricity fee payment sample frequency f of the user_m(i-1)。

S603: and calculating the difference value of the user electric charge payment time and the electric charge issuing time of the previous charging month of the sample charging month, and taking the difference value as the user electric charge payment sample time length.

It should be noted that, for a user who is normal in production and operation and healthy in capital turnover, the interval between the general electricity fee issuance time and the time for completing the electricity fee payment is also short, which indicates that the electricity fee payment of the user is relatively timely, so that the time length of the electricity fee payment sample of the user can be determined according to the electricity fee payment time of the user and the electricity fee issuance time, and the subsequent situation for evaluating the electricity fee recovery risk of the user is facilitated.

In the process of specifically implementing S603, based on table 1, the user 'S electric charge payment time in the previous charging month of the sample charging month in the industry of the user and the electric charge issuance time in the previous charging month of the sample charging month in the industry of the user can be obtained, and the difference between the two is calculated to obtain the user' S electric charge payment sample duration d in the previous charging month of the sample charging month in the industry of the user_p(i-1)。

S604: and determining the monthly capacity sample variation of the user by using the user business expansion capacity completion capacity, the user business expansion capacity reduction completion capacity, the user business expansion pause capacity and the user business expansion recovery capacity of the previous charging month of the sample charging month.

It should be noted that when a user handles business such as increase, decrease, pause, and resume of business capacity expansion amount according to the operation requirement of the user, it indicates that the operation state of the user may change and the production plan may be adjusted, so that the monthly capacity sample change amount of the user is determined according to the business expansion capacity completion capacity, the business expansion capacity reduction completion capacity, the business expansion pause capacity, and the business expansion resume capacity of the user, which is convenient for subsequent use in evaluating the electricity charge recycling risk of the user.

If the user frequently handles the volume expansion reduction service, the user may have a poor operation condition and low economic benefit, and the user may have a relatively high risk of recovering the electric charge.

In the process of specifically implementing S604, the user industry expansion capacity completion capacity of the industry where the user is located in the previous charging month of the sample charging month, the user industry expansion capacity reduction completion capacity of the industry where the user is located in the previous charging month of the sample charging month, the user industry expansion suspension capacity of the industry where the user is located in the previous charging month of the sample charging month, and the user industry expansion recovery capacity of the industry where the user is located in the previous charging month of the sample charging month may be obtained from the power marketing internal system by using the formula (2), and the user monthly capacity sample variation Δ c (i-1) of the industry where the user is located in the previous charging month of the sample charging month is obtained.

S605: acquiring historical electricity charge late payment generation times C of a user before a sample charging month_fAnd the previous generation time t of electric charge late payment_l。

In the specific implementation process of S605, the electricity charge late fee generation time of all months before the sample charging month in the industry where the user is located is obtained based on table 1, and accordingly, the total number of times C of historical electricity charge late fee generation before the sample charging month by the user is counted_fAnd determining the time t of the last payment of the electric charge arrearage by the user_l。

S606: the electricity charge amount C (i-1) generated in the previous charging month of the sample charging month and the sample industry development trend predicted value

Monthly electric charge payment sample frequency f of user_m(i-1) user electric charge payment sample duration d_p(i-1), user monthly capacity sample variation delta C (i-1), and historical electricity charge late fund generation times C_fAnd the previous generation time t of electric charge late payment_lAs input, the probability P of generating the electric charge late payment by the sample charging month^l(i) And as output, constructing an electric charge recovery risk early warning model by using an SVM regression algorithm.

Considering that data required for constructing the electric charge recovery risk early warning model can be specifically divided into three types: and (4) outputting an industry development trend prediction model, performing data statistics and constructing data. For convenience of visual understanding, when data required for constructing the electric charge recycling risk early warning model is divided according to different types, the data can be shown in table 3:

table 3:

in the process of specifically implementing S606, the specifically obtained electric charge recycling risk early warning model is shown as formula (17):

in the embodiment of the invention, based on the sample industry development trend prediction value of the industry where the user is located in the sample charging month, the user electricity fee payment sample time length of the previous charging month of the sample charging month, the user monthly capacity sample variation, the electricity fee amount generated in the previous charging month of the sample charging month, the historical electricity fee late fee generation times of the user before the sample charging month and the previous electricity fee late fee generation time, an electricity fee recovery risk early warning model is constructed, so that the probability value of the electricity fee late fee generated in the sample charging month of the industry where the user is located is predicted by using the electricity fee recovery risk early warning model, the electricity fee recovery risk of the user can be determined according to the probability value, and corresponding prompts are sent to related responsible personnel according to the difference of the electricity fee recovery risk so as to formulate the corresponding electricity fee recovery scheme of the related responsible personnel, so as to avoid the loss of the electric power enterprise operation.

Based on the electric charge recovery risk early warning method disclosed by the embodiment of the invention, correspondingly, the embodiment of the invention also discloses an electric charge recovery risk early warning device.

Referring to fig. 7, a block diagram of a risk early warning device for recovering electric charges according to an embodiment of the present invention is shown.

This electric charge retrieves risk early warning device includes: the system comprises an industry development trend prediction model 701, an acquisition module 702, an electric charge recovery risk early warning model 703 and an early warning module 704.

The industry development trend prediction model is used for predicting an industry development trend prediction value of the industry where the user is located in the current charging month, and is constructed on the basis of industry historical electricity selling development trend characteristic data and industry economic indicators.

The acquisition module is used for acquiring the electric charge amount of the user in the previous charging month, historical electric charge payment data and business expansion electric charge transaction data.

And the electric charge recovery risk early warning model is used for predicting the historical electric charge payment data, the business expansion electric transaction data and the industry development trend prediction value based on the received electric charge amount of the user in the previous charging month, and outputting the probability value of the electric charge late payment generated in the current charging month.

And the early warning module is used for generating early warning information for indicating that the electric charge recovery risk is high if the probability value output by the electric charge recovery risk early warning model is within the risk early warning range.

Optionally, the apparatus further comprises: a first building block.

The first building block comprises:

the first extraction unit is used for extracting the power selling trend characteristic data in industry historical power selling data in a set period according to a season adjustment algorithm, wherein the set period takes monthly degrees as a unit.

The system comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring a leading performance index and an instantaneity index related to the industry where a user is located, the leading performance index and the instantaneity index are obtained by analyzing the correlation of power selling trend characteristic data and industry economic indexes of various industries, the leading performance index at least comprises industry amplification capacity, manufacturing PMI and non-manufacturing PMI, and the instantaneity index at least comprises industry GDP acceleration and industry added value acceleration.

And the decorrelation unit is used for performing decorrelation calculation on the antecedent indexes and the instantaneity indexes according to a principal component analysis algorithm to obtain decorrelation indexes.

And the first construction unit is used for constructing an industry development trend prediction model by utilizing an SVM regression algorithm based on the power selling amount trend characteristic data and the correlation index.

Optionally, the apparatus further comprises: and an analysis module.

The analysis module includes:

and the second extraction unit is used for extracting the power selling trend characteristic data in the historical power selling data of each industry in the set period according to the season adjustment algorithm.

And the second acquisition unit is used for acquiring the industrial economic indexes of various industries.

Optionally, the apparatus further comprises: a second building block.

The second building block comprises:

the third obtaining unit is used for determining a sample charging month and obtaining a sample industry development trend prediction value of the industry where the user is located in the sample charging month based on a pre-constructed industry development trend prediction model; determining monthly electric charge payment sample frequency of the user based on the electric charge payment times of the user in a set period; and acquiring the historical electricity charge late payment generation times of the user before the sample charging month and the previous electricity charge late payment generation time.

The calculating unit is used for calculating the difference value between the user electricity fee payment time and the electricity fee issuing time of the previous charging month of the sample charging month, and taking the difference value as the user electricity fee payment sample time length; and determining the monthly capacity sample variation of the user by using the user business expansion capacity completion capacity, the user business expansion capacity reduction completion capacity, the user business expansion pause capacity and the user business expansion recovery capacity of the previous charging month of the sample charging month.

And the second construction unit is used for taking the electric charge amount generated in the previous charging month of the sample charging month, the sample industry development trend predicted value, the monthly electric charge payment frequency of the user, the electric charge payment duration of the user, the monthly capacity sample variation of the user, the historical generation times of the electric charge late payment and the previous generation time of the electric charge late payment as input, taking the probability of the electric charge late payment generated in the sample charging month as output, and constructing an electric charge recovery risk early warning model by utilizing an SVM (support vector machine) regression algorithm.

Optionally, the early warning module is further configured to generate a prompt message for indicating that the electric charge recycling risk is low if the probability value output by the electric charge recycling risk early warning model is not within the risk early warning range.

For specific execution principles and implementation processes of each model and each module in the electric charge recycling risk early warning device disclosed in the embodiment of the present invention, reference may be made to corresponding contents in the electric charge recycling risk early warning method disclosed in the embodiment of the present invention, and details are not repeated here.

Based on the electric charge recovery risk early warning device provided by the embodiment of the invention, an industry development trend prediction model predicts an industry development trend prediction value of the industry where the user is located in the current charging month, and the industry development trend prediction model is constructed based on industry historical electricity sales development trend characteristic data and industry economic indicators; the acquisition module acquires the electric charge amount of the user in the previous charging month, historical electric charge payment data and business expansion transaction data; the electric charge recovery risk early warning model predicts the electric charge amount of the user in the previous charging month, historical electric charge payment data, business expansion electric transaction data and the industry development trend prediction value and outputs a probability value of electric charge late payment generated in the current charging month; and if the probability value is within the risk early warning range, the early warning module generates early warning information for indicating that the electric charge recovery risk is high. According to the scheme provided by the embodiment of the invention, the probability value of the electricity charge late payment generated by the user in the current charging month is predicted by utilizing the pre-constructed industry development trend prediction model and the electricity charge recovery risk early warning model, and the information such as the industry development trend prediction value of the industry where the user is located, the historical electricity charge payment data and the industry expansion electricity transaction data of the user are comprehensively considered during prediction, so that the user with the electricity charge recovery risk is objectively and comprehensively identified from multiple dimensions, and when the user is identified as a user with a high electricity charge recovery risk, relevant responsible personnel are reminded, so that the relevant responsible personnel can formulate a corresponding electricity charge recovery scheme, and the loss of the electric power enterprise operation is avoided.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, the system or system embodiments are substantially similar to the method embodiments and therefore are described in a relatively simple manner, and reference may be made to some of the descriptions of the method embodiments for related points. The above-described system and system embodiments are only illustrative, wherein the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

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 components and steps have been described above generally in terms of their functionality in order to clearly illustrate this 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 implementation. 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 invention.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. 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 invention. Thus, the present invention 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 electric charge recovery risk early warning method is characterized by comprising the following steps:

2. The method of claim 1, wherein the industry development trend prediction model is constructed by a process comprising:

3. The method of claim 2, wherein the pro-conductive index and the instantaneity index are obtained from a process of analyzing correlation between the power selling trend characteristic data of each industry and the industry economic index, and the process comprises:

acquiring industrial economic indexes of various industries;

4. The method according to claim 1, wherein the construction process of the electric charge recycling risk early warning model comprises the following steps:

5. The method of claim 1, further comprising:

6. An electric charge recovery risk early warning device, characterized in that the device includes:

7. The apparatus of claim 6, further comprising: a first building block;

the first building block comprises:

8. The apparatus of claim 6, further comprising: an analysis module;

the analysis module includes:

9. The apparatus of claim 6, further comprising: a second building block;

the second building block comprises:

10. The device of claim 6, wherein the early warning module is further configured to generate a prompt message indicating that the risk of recovering the electric charge is low if the probability value is not within the risk early warning range.