A User Classification Method and System Based on Electricity Consumption Sensitivity

Abstract

The present disclosure belongs to the field of power supply technology, and provides a user classification method and system based on electricity consumption sensitivity, including the following steps: obtaining user electricity load data; obtaining user daily load characteristic values according to the user electricity sensitivity factor data in the obtained user electricity load data; performing a first classification on the obtained user daily load characteristic values; calculating the electricity consumption sensitivity after the first classification; and based on the obtained first classification result and electricity consumption sensitivity, performing a second classification of the electricity consumption sensitivity to achieve labeled classification of electricity users.

Description

Technical Field

[0001] The present disclosure belongs to the technical field of power supply, and particularly relates to a user classification method and system based on electricity consumption sensitivity. Background Art

[0002] The statements in this part merely provide background technical information related to the present disclosure, and do not necessarily constitute prior art.

[0003] With the continuous development of economy and technology, electrical equipment and electricity consumption demands have gradually become diversified. At the same time, the development of smart distribution networks and the progress of electricity consumption information collection systems have led to a rapid growth in the scale of collected electricity consumption data, increasing the difficulty of distribution calculation and regulation processing. User classification technology reduces the amount of calculation for supply-side prediction management while analyzing user characteristics, and can help the power grid understand the personalized and differentiated service demands of users, which is of great significance for formulating load dispatching plans and operation control.

[0004] As understood by the inventors, many different technologies have been used for load analysis currently, and the clustering analysis method plays an important role in load analysis. Domestic and foreign scholars have conducted certain research on power grid user classification. However, the existing research mostly classifies users based on the curve shape or characteristics of typical load days. Such user classification methods cannot meet the increasingly diversified user types and user demands. Summary of the Invention

[0005] To solve the above problems, the present disclosure proposes a user classification method and system based on electricity consumption sensitivity, which solves the electricity management problems caused by the rapid growth in the quantity scale of electricity consumption data and the insufficient refinement of user classification. On the basis of classifying users according to their electricity consumption characteristics, the degree of change in user electricity consumption characteristics when a certain factor changes is defined as electricity consumption sensitivity. Users are classified by tagging based on their electricity consumption sensitivity, refining the user classification method. When conducting electricity management, analyzing the sensitivity can reduce the amount of data processing and assist in rapid calculation and decision-making.

[0006] According to some embodiments, the first solution of the present disclosure provides a user classification method based on electricity consumption sensitivity, adopting the following technical solution:

[0007] A user classification method based on electricity consumption sensitivity includes the following steps:

[0008] Obtain user electricity load data;

[0009] Obtain the user daily load characteristic value according to the user electricity consumption sensitive factor data in the obtained user electricity load data;

[0010] Perform a primary classification on the obtained user daily load characteristic value;

[0011] Calculate the electricity consumption sensitivity after the first classification;

[0012] Based on the obtained first classification result and electricity consumption sensitivity, perform a second classification of the electricity consumption sensitivity to achieve the labeled classification of electricity users.

[0013] As a further technical limitation, preprocess the obtained user electricity load data to obtain standardized sample data.

[0014] Furthermore, the preprocessing at least includes eliminating distorted data, complementing missing data, and smoothing outlier data.

[0015] As a further technical limitation, the user electricity sensitivity factor data at least includes holidays, temperature, wind speed, and rain and snow.

[0016] As a further technical limitation, use the k-means clustering method for the first classification to group users with similar daily load characteristic values of users into one category.

[0017] As a further technical limitation, the electricity consumption sensitivity is the cumulative sum after multiplying all sensitivity values by their characteristic sensitivity weights; each sensitivity value corresponds to the daily load characteristic value of the user's electricity consumption.

[0018] Furthermore, use a combined weighting method based on the analytic hierarchy process and entropy weight method to calculate the characteristic sensitivity weights.

[0019] According to some embodiments, the second solution of the present disclosure provides a user classification system based on electricity consumption sensitivity, adopting the following technical solution:

[0020] A user classification system based on electricity consumption sensitivity, comprising:

[0021] An acquisition module, configured to acquire user electricity load data, and obtain daily load characteristic values of users according to the user electricity sensitivity factor data in the acquired user electricity load data;

[0022] A first classification module, configured to perform a first classification on the obtained daily load characteristic values of users;

[0023] A calculation module, configured to calculate the electricity consumption sensitivity after the first classification;

[0024] A second classification module, configured to perform a second classification of the electricity consumption sensitivity based on the obtained first classification result and electricity consumption sensitivity to achieve the labeled classification of electricity users.

[0025] According to some embodiments, the third solution of the present disclosure provides a computer-readable storage medium, adopting the following technical solution:

[0026] A computer-readable storage medium has a program stored thereon, and when the program is executed by a processor, the steps in the user classification method based on electricity consumption sensitivity described in the first aspect of the present disclosure are implemented.

[0027] According to some embodiments, a fourth solution of the present disclosure provides an electronic device, adopting the following technical solution:

[0028] An electronic device includes a memory, a processor, and a program stored on the memory and executable on the processor. When the processor executes the program, the steps in the user classification method based on electricity consumption sensitivity described in the first aspect of the present disclosure are implemented.

[0029] Compared with the prior art, the beneficial effects of the present disclosure are as follows:

[0030] (1) Based on classifying users according to their electricity consumption characteristics, the present disclosure proposes to classify users by tagging based on their electricity consumption sensitivity, refining the user classification rules.

[0031] (2) According to the degree of influence of various factors on the user load characteristics, the present disclosure proposes a calculation method for user electricity consumption sensitivity, obtains new user classification indicators, assists in analyzing user electricity consumption characteristics, helps understand the load characteristic rules and user personalized electricity consumption needs, assists the power grid in providing targeted services, and improves the power grid load forecasting and power supply management capabilities.

[0032] (3) The present disclosure proposes to classify users by tagging according to the electricity consumption sensitivity value, which can provide data support when considering changes in load characteristics caused by external factor changes, reduce the load forecasting calculation amount, and provide a reference for the power grid company to formulate a compliance dispatching plan and differentiated services. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The accompanying drawings forming a part of this disclosure are used to provide a further understanding of the present disclosure. The schematic embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation of the present disclosure.

[0034] Figure 1 is a flowchart of the user classification method based on electricity consumption sensitivity in Embodiment 1 of the present disclosure;

[0035] Figure 2 is a schematic diagram of the specific steps of the user classification method based on electricity consumption sensitivity in Embodiment 1 of the present disclosure;

[0036] Figure 3 is a structural block diagram of the user classification system based on electricity consumption sensitivity in Embodiment 2 of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0037] The present disclosure will be further described below in conjunction with the accompanying drawings and embodiments.

[0038] It should be noted that the following detailed description is illustrative and aims to provide further explanation of the present disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present disclosure pertains.

[0039] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present disclosure. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0040] In the case of no conflict, the embodiments in the present disclosure and the features in the embodiments may be combined with each other.

[0041] Embodiment 1

[0042] Embodiment 1 of the present disclosure introduces a user classification method based on electricity consumption sensitivity.

[0043] As Figure 1 shown, a user classification method based on electricity consumption sensitivity includes the following steps:

[0044] Obtain user electricity load data;

[0045] According to the user electricity sensitivity factor data in the obtained user electricity load data, obtain the user daily load characteristic value;

[0046] Perform a primary classification on the obtained user daily load characteristic value;

[0047] Calculate the electricity consumption sensitivity after the primary classification;

[0048] Based on the obtained primary classification result and electricity consumption sensitivity, perform a secondary classification of the electricity consumption sensitivity to achieve labeled classification of electricity users.

[0049] To solve the electricity management problems caused by the rapid growth of the quantity scale of electricity consumption data and the insufficient refinement degree of user classification, according to the sensitivity degree of users to various factors, this embodiment proposes a user classification method based on user electricity consumption sensitivity, which is used to analyze load characteristics, refine user classification rules, and reduce the calculation data capacity of load management prediction.

[0050] As Figure 2 shown, the user classification method based on electricity consumption sensitivity specifically includes the following steps:

[0051] Step S01: Preprocess the user electricity load data and standardize the sample data;

[0052] Step S02: Organize the user's electricity consumption sensitive factor data, find the typical days of factor changes and the corresponding conventional typical days;

[0053] Step S03: Extract and calculate the user's average daily load characteristic values, and conduct a primary classification of the sample data according to the load characteristics;

[0054] Step S04: Calculate the relevant characteristic values of the change typical days and the corresponding conventional typical days, and obtain the corresponding characteristic sensitivity values through comparison calculation;

[0055] Step S05: Determine the characteristic sensitivity weights of different categories of users, and obtain the final electricity consumption sensitivity;

[0056] Step S06: Conduct a secondary classification according to the electricity consumption sensitivity, include the user's electricity consumption sensitivity values in the corresponding evaluation ranges, and conduct a labeled classification of the users.

[0057] As one or more implementation manners, in step S01, missing values and outliers will have an adverse impact on the relevant calculations of the samples. Removing distorted data, filling in missing data, and smoothing outliers can reduce such adverse impacts. Data standardization can remove the unit limitations of the data and convert it into a dimensionless pure numerical value; the specific process of step S01 is as follows:

[0058] Step S101: Delete the data with a high missing rate;

[0059] Step S102: Fill in the missing quantitative data using the Lagrange interpolation method, and convert the missing qualitative data into dummy variables;

[0060] For n samples (x1, y1), (x2, y2), …, (x n , y n ) without missing values, construct a smooth curve, and substitute the abscissa x corresponding to the sample with missing values to obtain an approximate value of the missing value:

[0061]

[0062] Step S103: Use the load data of the previous and next 1 hour as the reference data, and smooth the outliers using the average value method;

[0063] It should be noted that step S103 can reduce the impact caused by distorted data;

[0064] Step S104: Data standardization.

[0065] The standardization method adopts range normalization; data standardization facilitates the comparison of indicators with different units or magnitudes, that is

[0066]

[0067] As one or more embodiments, in step S02, factor data related to electricity consumption characteristics is sorted out from the available data, such as common user electricity consumption sensitive factors like holidays, temperature, wind speed, rain and snow. Such factors are related to weather and date, and the data is relatively easy to obtain. Since the sensitive factors are common, there is more relevant data, and the reliability of the calculated classification results is stronger. At the same time, it is easier to find data with combined changes of multiple factors.

[0068] Find the typical days of factor changes and the corresponding normal typical days. Factor changes include common factor changes such as sudden temperature rise, sudden temperature drop, strong wind, holidays, rain and snow weather, etc. The typical day of factor change is the date when the corresponding factor change occurs and other factors are at a relatively normal level. The normal typical day is the date when other factors are basically the same but the corresponding factor change does not occur. The normal typical day corresponding to the typical day of factor change should be close in time to avoid the influence of redundant factors on the results.

[0069] As one or more embodiments, in step S03, the user's electricity consumption data is read, the average value of the daily load in the data is taken, and the user's daily load characteristic value is calculated. Common daily load characteristic values include daily average load, daily load rate, peak-hour power consumption rate, valley electricity coefficient, peak electricity consumption period, etc. Calculating using the average daily load characteristic is simple and can reduce the influence of atypical electricity consumption to a certain extent.

[0070] The k-means clustering method is used for the first classification. The k-means method is relatively simple to implement and has good clustering effect. The purpose of the first clustering is to classify users with similar relevant load characteristics together, reducing the influence of the special load curve form of some users on the determination of weights in the calculation of user sensitivity. For example, for a load type with a single index approaching the limit, if the weight determination is unreasonable, it is difficult to correctly obtain the user's electricity consumption sensitivity.

[0071] It should be noted that in this embodiment, the k-means clustering method is used for the first classification. In other embodiments, other methods can also be used to achieve the first classification.

[0072] As one or more embodiments, in step S04, the relevant daily load characteristic values of the selected typical day of change and the corresponding normal typical day are calculated.

[0073] The specific process of sensitivity calculation is as follows:

[0074] Step S401: The sensitivity calculation method of user n for the quantitative index of feature i under the influence of factor k:

[0075]

[0076]

[0077]

[0078] Wherein, i represents the user serial number, j represents the load sample serial number, m represents the total number of corresponding samples, R is a set constant, represents the eigenvalue of the quantifiable load characteristic k1 of the user load sample under the influence of factor l, represents the corresponding eigenvalue of the conventional typical day, Δl represents the quantified value of the change of factor l after standardization, and the factor represents the characteristic change rate, represents the average change rate of the quantifiable load characteristic k1 of the user i load sample under the influence of factor l, represents the sensitivity of the quantifiable load characteristic k1 of the user i to factor l. In this calculation method, the sensitivity is related to the scale of the characteristic quantity change rate, and the value range of the sensitivity can be limited within [0, 1].

[0079] Step S402: Fuzzily quantify the qualitative index, and the value range is [0, 1]. The calculation method of the sensitivity of user n to the qualitative characteristic k2 index under the influence of factor l:

[0080]

[0081] Wherein, is the fuzzy quantification value of different qualitative characteristic k2 indexes of user i under the influence of factor l, is the number of samples of the typical day corresponding to the characteristic value change, is the total number of samples of the typical day of the characteristic change.

[0082] As one or more implementation manners, in step S05, a combined weighting method based on the analytic hierarchy process and the entropy weight method is used to determine the sensitivity weight of each characteristic.

[0083] The analytic hierarchy process is a method of comparing the importance degree of each pair of decision-making indexes in each layer according to expert suggestions to finally obtain the weight, which reflects the intention of the decision maker.

[0084] The entropy weight method is a method of determining the weight according to the data dispersion degree. The entropy weight method can avoid subjectivity when assigning weights. Take the average value of the two weights as the final weight α, and list the characteristics selected to participate in the sensitivity calculation as k1, k2, k3,..., k n ,, the combined weighting method determines the weights as α1, α2, α3,..., α n , and calculate the final electricity consumption sensitivity of the user:

[0085]

[0086] Wherein, The final electricity consumption sensitivity of user i under the influence of factor l.

[0087] It should be noted that in this embodiment, the combined weighting method is used to determine the weights of the feature sensitivities. In other embodiments, other methods can also be used to obtain the weights of the feature sensitivities.

[0088] As one or more embodiments, in step S06, the larger the value of the final electricity consumption sensitivity, the stronger the sensitivity. When the corresponding factor changes, the electricity consumption characteristics of the user may change more significantly. The sensitivity is divided into insensitive, sensitive, extremely sensitive, etc., and different numerical ranges are set, with the total range in the [0, 1] area.

[0089] The user sensitivity list is shown in Table 1. When factor changes occur and user prediction and analysis are required, the degree of response of each type of user to different factors can be obtained from the table. When the calculation amount needs to be reduced, only the users labeled as sensitive can be selected for prediction and management. At the same time, this user classification can provide certain reference in load forecasting and power supply planning.

[0090] Table 1 User Sensitivity List

[0091]

[0092] In this embodiment, according to the user classification method based on user electricity consumption sensitivity, the user electricity consumption sensitivity is defined as the degree of change in user electricity consumption characteristics when a certain factor changes; a new user classification standard is proposed, the user classification method is refined, and the sensitivity is analyzed during electricity consumption management, which can reduce the amount of data processing and assist in fast calculation and decision-making.

[0093] Embodiment Two

[0094] Embodiment Two of the present disclosure introduces a user classification system based on electricity consumption sensitivity.

[0095] As Figure 3 shown, a user classification system based on electricity consumption sensitivity includes:

[0096] An acquisition module, configured to acquire user electricity load data, and obtain user daily load characteristic values according to the user electricity sensitive factor data in the acquired user electricity load data;

[0097] A primary classification module, configured to perform a primary classification on the obtained user daily load characteristic values;

[0098] A calculation module, configured to calculate the electricity consumption sensitivity after the primary classification;

[0099] The secondary classification module is configured to perform secondary classification of power consumption sensitivity based on the obtained primary classification result and power consumption sensitivity, so as to achieve labeled classification of power consumption users.

[0100] The detailed steps are the same as those of the user classification method based on power consumption sensitivity provided in the first embodiment, and will not be elaborated here.

[0101] Embodiment Three

[0102] Embodiment Three of the present disclosure provides a computer-readable storage medium.

[0103] A computer-readable storage medium, on which a program is stored, and when the program is executed by a processor, it implements the steps in the user classification method based on power consumption sensitivity as described in Embodiment One of the present disclosure.

[0104] The detailed steps are the same as those of the user classification method based on power consumption sensitivity provided in the first embodiment, and will not be elaborated here.

[0105] Embodiment Four

[0106] Embodiment Four of the present disclosure provides an electronic device.

[0107] An electronic device includes a memory, a processor, and a program stored on the memory and executable on the processor. When the processor executes the program, it implements the steps in the user classification method based on power consumption sensitivity as described in Embodiment One of the present disclosure.

[0108] The detailed steps are the same as those of the user classification method based on power consumption sensitivity provided in the first embodiment, and will not be elaborated here.

[0109] The above are only the preferred embodiments of the present disclosure and are not used to limit the present disclosure. For those skilled in the art, the present disclosure can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A user classification method based on electricity consumption sensitivity, characterized in that, It includes the following steps: Obtain the user's electricity load data; Based on the user electricity sensitive factor data in the obtained user electricity load data, obtain the user's daily load characteristic value; The user electricity sensitive factor data at least includes holidays, temperature, wind speed, and rain and snow; Perform a first classification on the obtained user's daily load characteristic value; Calculate the electricity sensitivity after the first classification; The electricity sensitivity is the cumulative sum of the products of all sensitivity values and their characteristic sensitivity weights; each sensitivity value corresponds to the user's daily load characteristic value of electricity consumption; The specific process of sensitivity calculation is: In the formula, i represents the user serial number, j represents the load sample serial number, m represents the total number of corresponding samples, R is a set constant, represents the quantifiable load characteristic l of the user load sample under the influence of factor k1 eigenvalue, represents the eigenvalue of the corresponding conventional typical day, represents the quantified value of the change of factor l after standardization, and factor represents the characteristic change rate, represents the user i load sample under the influence of factor l quantifiable load characteristic k1 average change rate, represents the user i quantifiable load characteristic k1 sensitivity to factor l ; Fuzzily quantify the qualitative indicators, and the calculation method of the user's sensitivity to the qualitative characteristic indicators under the influence of factors: In the formula, is the fuzzy quantification value of the k2 index of different qualitative characteristics of user i under the influence of factor l, is the number of samples on the typical day when the corresponding characteristic value changes, is the total number of samples on the typical day when the characteristic changes; Adopt a combined weighting method based on the analytic hierarchy process and the entropy weight method to calculate the characteristic sensitivity weight; The user's final electricity sensitivity is: In the formula, is the final electricity consumption sensitivity of user i under the influence of factor l; Based on the obtained first classification result and electricity sensitivity, perform a second classification of the electricity sensitivity to achieve the labeled classification of electricity users.

2. The user classification method based on electricity consumption sensitivity according to claim 1, wherein Preprocess the obtained user electricity load data to obtain standardized sample data.

3. A user classification method based on electricity consumption sensitivity as described in claim 2, characterized in that, The preprocessing at least includes eliminating distorted data, complementing missing data, and smoothing outliers.

4. A user classification method based on electricity consumption sensitivity as described in claim 1, characterized in that Adopt k- means A clustering method is used for the first classification to group users with similar daily load characteristic values into one category.

5. A user classification system based on electricity consumption sensitivity, characterized in that including: An acquisition module, configured to obtain the user's electricity load data, and based on the user electricity sensitive factor data in the obtained user electricity load data, obtain the user's daily load characteristic value; The user electricity sensitive factor data at least includes holidays, temperature, wind speed, and rain and snow; A first classification module, configured to perform a first classification on the obtained user's daily load characteristic value; A calculation module, configured to calculate the electricity sensitivity after the first classification; The electricity sensitivity is the cumulative sum of the products of all sensitivity values and their characteristic sensitivity weights; each sensitivity value corresponds to the user's daily load characteristic value of electricity consumption; The specific process of sensitivity calculation is: In the formula, i represents the user serial number, j represents the load sample serial number, m represents the total number of corresponding samples, R is a set constant, represents the quantifiable load characteristics l of the user load sample under the influence of factor k1 feature value, represents the corresponding conventional typical day feature value, represents the quantified value of the change of factor l after standardization, and factor represents the feature change rate, represents the user i load sample under the influence of factor l quantifiable load characteristics k1 average change rate, represents the user i quantifiable load characteristics k1 sensitivity to factor l ; Fuzzily quantify the qualitative indicators, and the calculation method of the user's sensitivity to the qualitative characteristic indicators under the influence of factors: In the formula, is the fuzzy quantification value of the k2 index of different qualitative characteristics of user i under the influence of factor l, is the number of samples of the typical day corresponding to the change of the characteristic value, is the total number of samples of the typical day of the characteristic change; Adopt a combined weighting method based on the analytic hierarchy process and the entropy weight method to calculate the characteristic sensitivity weight; The user's final electricity sensitivity is: In the formula, is the final electricity consumption sensitivity of user i under the influence of factor l; A second classification module, configured to perform a second classification of the electricity sensitivity based on the obtained first classification result and electricity sensitivity to achieve the labeled classification of electricity users.

6. A computer-readable storage medium having a program stored thereon, characterized in that, When the program is executed by a processor, it implements the steps in the user classification method based on electricity sensitivity described in any one of claims 1-4.

7. An electronic device, comprising a memory, a processor, and a program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the user classification method based on electricity sensitivity described in any one of claims 1-4.

Images