CN114638284A - Power utilization behavior characterization method considering external influence factors - Google Patents
Power utilization behavior characterization method considering external influence factors Download PDFInfo
- Publication number
- CN114638284A CN114638284A CN202210145798.3A CN202210145798A CN114638284A CN 114638284 A CN114638284 A CN 114638284A CN 202210145798 A CN202210145798 A CN 202210145798A CN 114638284 A CN114638284 A CN 114638284A
- Authority
- CN
- China
- Prior art keywords
- user
- external influence
- influence factors
- power consumption
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012512 characterization method Methods 0.000 title claims abstract description 12
- 230000005611 electricity Effects 0.000 claims abstract description 88
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 230000006399 behavior Effects 0.000 claims description 71
- 238000000034 method Methods 0.000 claims description 21
- 238000012545 processing Methods 0.000 claims description 15
- 230000002159 abnormal effect Effects 0.000 claims description 13
- 238000004422 calculation algorithm Methods 0.000 claims description 10
- 230000008859 change Effects 0.000 claims description 7
- 230000000694 effects Effects 0.000 claims description 6
- 238000013139 quantization Methods 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 238000011425 standardization method Methods 0.000 claims description 2
- 239000011248 coating agent Substances 0.000 claims 1
- 238000000576 coating method Methods 0.000 claims 1
- 238000002360 preparation method Methods 0.000 claims 1
- 238000004458 analytical method Methods 0.000 abstract description 10
- 238000011160 research Methods 0.000 abstract description 5
- 238000004364 calculation method Methods 0.000 description 8
- 238000003064 k means clustering Methods 0.000 description 4
- 230000004044 response Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 230000005612 types of electricity Effects 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000036642 wellbeing Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Finance (AREA)
- Economics (AREA)
- Development Economics (AREA)
- Accounting & Taxation (AREA)
- Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- General Engineering & Computer Science (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Tourism & Hospitality (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a power utilization behavior characterization method considering external influence factors, which comprises the following steps: preprocessing the power load data; constructing a user electricity utilization characteristic system containing external influence factors; establishing and solving a user power consumption behavior clustering model considering external influence factors; depicting the user's electric image. The invention provides a power consumption behavior characterization method considering external influence factors based on the research on a user power consumption behavior analysis method in the prior art, so that the characterization of the user power consumption behavior is more precise. Compared with the traditional power utilization behavior analysis display mode, the display mode provided by the invention can visually reflect the multi-dimensional characteristics of the user, and provides powerful support for an electric power company to master the power utilization behavior of the user, mine the requirements of the user and improve the service level.
Description
Technical Field
The invention relates to the technical field of power consumption behavior analysis of power grid users, in particular to a power consumption behavior fine-characterization method considering external influence factors.
Background
With the popularization and promotion of the intelligent electric meter, mass power utilization data of a user side are collected in real time, and the data contain rich power utilization activity information of the user. The power utilization rule of the user is mined from the mass power utilization data, and basic data support can be provided for power enterprises to carry out accurate marketing, personalized power utilization service and flexible interaction with the user. The user power utilization behavior classification is an important link of user power utilization behavior analysis.
The electricity utilization behaviors of residential users are highly diverse and variable, only a few types of typical daily load curves cannot effectively reflect the actual electricity utilization behaviors of the users, and the daily load curves need to be finely divided to identify the dynamic electricity utilization modes of the users. One of the more common means in the existing user behavior classification is a clustering method based on electricity utilization characteristics, and the k-means clustering algorithm has the characteristics of simplicity and high efficiency, is suitable for processing a large data set, and is therefore often applied to the classification of electricity utilization behaviors. Document 1(Lc a, Vv a, Mh a, et al.development of electric continuity conditions profiles of residual building based on smart meter data clustering [ J ]. Energy and building, 2021, 252.), comparing three clustering methods of k-means clustering, fuzzy k-means clustering and aggregation hierarchical clustering, obtaining the best effect of the k-means clustering algorithm through simulation comparison, and analyzing the relationship between the user electricity consumption behavior and factors such as festival, season, geographical position, housing type and the like on the clustering result; however, these influence factors are not considered in the clustering process, and the classification result of the power utilization behavior of the user is not accurate enough. Document 2 (yanwei red, reiqingping, lanyu, and the like, power utilization behavior clustering analysis algorithm research for users based on adjustment potential indexes [ J ]. power construction, 2018, 39 (06): 96-104.) a user load transfer rate model is firstly constructed based on the principle of user psychology, then the adjustment potential characteristics of each user under the peak-valley electricity price are calculated by using the model, and finally, the users with different peak-valley characteristics are divided based on the adjustment potential characteristic clustering, so that a basis is provided for resource positioning of demand side response. Document 3(Wang Y, Chen Q, Kang C, et al. Cluster of electric stability Behavior aware Big Data Applications [ J ]. IEEE Transactions on Smart Grid, 2016, 7 (5): 2437-. The obtained classification result of the electricity utilization behaviors of the users can provide data support for the formulation of a management strategy on the demand side, but the influence of external factors such as temperature, date, economy and the like on the electricity consumption is not involved.
The prior art has carried out a great deal of research on the classification of power utilization behaviors of users, however, the prior research still has the following defects:
1. in the aspect of user electricity utilization behavior classification, most of the existing methods are based on the characteristic of a load curve, the influence of external multidimensional influence factors such as temperature, climate, household income, electricity price and the like on the user electricity utilization behavior is not combined, and the influence of the multidimensional influence factors on the electricity utilization behaviors of different types of users is not considered. The situation that users with greatly different power consumption degrees for the same type of influencing factors are divided into the same type of power consumption groups easily occurs in the dividing mode, the classification reliability is lowered, and the classification of the users is not fine enough.
2. In the aspect of the presentation mode of the power utilization behavior analysis result, the current depiction of the power utilization behavior of the user is mainly presented in a numerical value and curve mode on the basis of the clustering result, the power utilization behavior of the user cannot be visually and comprehensively presented, and the characteristic difference among different power utilization users cannot be well distinguished.
Disclosure of Invention
The invention aims to provide an electricity utilization behavior characterization method considering external influence factors, which can solve the problems that in the prior art, users with large differences in electricity utilization statuses for the same type of influence factors are divided into the same type of electricity utilization groups due to the fact that the external influence factors are not considered in analyzing the electricity utilization behaviors of the users, the classification reliability is reduced, the classification of the users is not fine enough, and the like.
The purpose of the invention is realized by the following technical scheme:
a power utilization behavior characterization method considering external influence factors comprises the following steps:
preprocessing the power load data;
constructing a user electricity utilization characteristic system containing external influence factors;
establishing and solving a user electricity consumption behavior clustering model considering external influence factors;
depicting the user power consumption portrait.
Further, the preprocessing of the power load data includes processing of abnormal data and data standardization processing.
Further, the constructing of the user electricity utilization characteristic system containing the external influence factors comprises:
constructing a user electricity utilization curve feature set;
constructing a characteristic set of external influence factors of the power utilization of the user;
and constructing a comprehensive characteristic vector of the power utilization.
Further, the constructing the external influence factor feature set for the power utilization of the user comprises:
calculating the quantitative characteristic R of the influence degree of the temperature factorn,1。；
Calculating the quantitative characteristic R of the influence degree of the humidity factorn,2；
Calculating quantitative characteristic R of influence degree of electricity price factorn,3；
Calculating income factor influence degree quantization characteristic Rn,4；
Obtaining the external influence factor characteristic set Rn＝[Rn,1,Rn,2,Rn,3,Rn,4]。
Further, the constructing of the electricity consumption comprehensive feature vector includes:
and processing the characteristic set of the power utilization curve of the user and the characteristic set of the external influence factors by adopting a zero-mean standardization method to obtain a power utilization comprehensive characteristic vector of the user.
Further, the establishing and solving of the user electricity consumption behavior clustering model considering the external influence factors comprises:
constructing an objective function of the clustering model;
and solving the objective function of the clustering model by adopting a k-means algorithm.
Further, the optimal clustering number K in the K-means algorithm is determined as follows:
respectively clustering by taking K as 2,3,4, … and 20;
respectively calculating clustering effectiveness indexes corresponding to the K values;
and selecting the K value with the best clustering effect as the clustering number of the model through the clustering index.
Further, the portraying the user power portrait includes:
determining a user type tag;
determining a user behavior feature label;
and drawing a user image according to the user type label and the user behavior characteristic label as the parameter of the radar map.
Further, the processing of the abnormal data comprises removing and correcting the abnormal data:
if the electric load data contains a missing value, but the missing amount is less than 10% of the whole, filling the missing part by adopting a whole averaging method;
if the loss amount of the power consumption load data exceeds 10 percent of the whole, the user data is rejected;
if the zero value contained in the electric load data exceeds 95% of the whole, rejecting the user data;
if the ith data point x in the electrical load dataiIf the value of (a) suddenly changes, the data point is rejected according to xiData points x before and after the pointi-1And xi+1And calculating interpolation for filling.
Further, the decision formula for determining whether the data point suddenly changes is as follows:
wherein: deltaiAnd (3) representing the degree of abrupt change of the ith data point, and when the value of the degree of abrupt change is greater than a specified threshold epsilon, considering that the load point is abnormal data, wherein the value range of the threshold epsilon is 0.5-0.8.
Based on the prior art, the invention develops research on a user power consumption behavior analysis method, and provides a power consumption behavior characterization method considering external influence factors, so that the characterization of the user power consumption behavior is more precise. Compared with the traditional power utilization behavior analysis display mode, the display mode provided by the invention can visually reflect the multi-dimensional characteristics of the user, and provides powerful support for an electric power company to master the power utilization behavior of the user, mine the requirements of the user and improve the service level.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a flow chart of a method for characterizing electricity usage behavior in consideration of external influences in accordance with the present invention;
FIG. 2 is a schematic diagram of a comprehensive clustering result of power utilization behaviors in consideration of external influence factors;
FIG. 3 is a multi-dimensional user power consumption behavior portrait obtained according to the power consumption behavior comprehensive clustering result.
Detailed Description
The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.
The embodiments of the present disclosure are described below with specific examples, and other advantages and effects of the present disclosure will be readily apparent to those skilled in the art from the disclosure in the specification. It is to be understood that the described embodiments are merely illustrative of some, and not restrictive, of the embodiments of the disclosure. The disclosure may be embodied or carried out in various other specific embodiments, and various modifications and changes may be made in the details within the description without departing from the spirit of the disclosure. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict. All other embodiments, which can be derived by a person skilled in the art from the embodiments disclosed herein without making any creative effort, shall fall within the protection scope of the present disclosure.
The invention relates to a power utilization behavior characterization method considering external influence factors, which comprises the following steps of:
step S1 is to preprocess the power load data.
Preprocessing the power load data includes two aspects: processing abnormal data and data standardization processing.
Further, the processing of the abnormal data comprises the elimination and the correction of the abnormal data.
For an original power load curve data set, data is processed according to the following conditions:
1) and if the electric load data contains a missing value and the missing amount is less than 10% of the whole, filling the missing part by adopting a whole averaging method.
2) And if the loss amount of the power consumption load data exceeds 10% of the whole, rejecting the user data.
3) And if the electric load data contains more than 95% of the whole zero value, rejecting the user data.
4) If the ith data point x in the electrical load dataiIf the value of (b) is suddenly changed, the data point is eliminated according to xiData points x before and after the pointi-1And xi+1And calculating interpolation for filling.
The decision formula for determining whether the data point suddenly changes is as follows:
(1) in the formula: deltaiAnd (3) representing the degree of abrupt change of the ith data point, and when the value of the degree of abrupt change is greater than a specified threshold value epsilon, considering the load point as abnormal data, wherein the value range of the epsilon is generally 0.5-0.8. After being removed according to xiData points x before and after the pointi-1And xi+1And calculating interpolation for filling.
Further, the data normalization process includes:
and standardizing the electrical load data after the abnormal data processing by adopting a normalization method.
The electricity load data set after data preprocessing comprises N electricity load sequences, and the sequences are marked as Q ═ X1,X2,…,XN]. Wherein, Xn(N-1, 2, …, N) represents electricity consumption sequence data of the nth user, and has m data points, denoted as Xn＝[xn,1,xn,2,…,xn,m]。
Through the data processing in step S1, the load data can be preliminarily screened and corrected, and errors in subsequent user classification can be reduced.
And step S2, constructing a user electricity utilization characteristic system containing external influence factors.
The step of constructing the user electricity utilization characteristic system comprises the steps of constructing a user electricity utilization curve characteristic set and constructing an external influence factor characteristic set, and how to construct the user electricity utilization characteristic system is explained from the two parts.
Further, step S2 specifically includes:
step S201, constructing a user power utilization curve feature set, including:
and inducing power utilization characteristic indexes reflecting the power utilization characteristics of the user.
The electricity utilization characteristics of the user are determined by combining the daily life law of residents, and 11 electricity utilization characteristic indexes reflecting the electricity utilization characteristics of the user are summarized from 5 time periods including the whole day, the peak period, the average period, the valley period, the noon and the like respectively: maximum daily load P1Daily minimum load P2Peak to valley difference of day P3Peak-to-valley ratio P4Daily average load P5Daily load factor P6Daily maximum load utilization factor P7Peak load factor P8Flat period load factor P9Load factor P in the valley period10Coefficient of meridian P11。
Load sequence X for user nn＝[xn,1,xn,2,…,xn,m]Calculating according to the electricity utilization characteristic indexes to obtain 11 electricity utilization curve characteristics, namely, an electricity utilization curve characteristic vector P forming the user nn＝[Pn,1,Pn,2,…,Pn,11]。
Step S202, constructing a user electricity consumption external influence factor feature set, comprising:
the invention takes four types of influence factors, namely electricity price, income, temperature and humidity, as external influence factors of the electricity utilization behavior of the user, and quantifies the influence degree of the influence factors on the electricity utilization behavior according to the difference of different influence factors.
Respectively calculating the power load sequence X of the user nn＝[xn,1,xn,2,…,xn,m]The calculation method comprises the following steps:
calculating the quantitative characteristics of the influence degree of the temperature factors:
when the influence degree of the temperature on the electricity utilization behavior is quantified, the electricity utilization data sequence X of the user nnThe temperature data sequence corresponding to the time point is denoted as Y ═ Y1,y2,…,ym]，XnThe pearson correlation coefficient for Y reflects the degree of correlation between the variables. The quantization formula is as follows:
in the formula:
are respectively a sequence XnY, mean value. The value range of | rho | is [0,1 ]]. The calculation result represents the quantitative characteristic of the influence degree of the temperature factor on the electricity utilization behavior of the user n and is marked as Rn,1。
Calculating the influence degree quantization characteristics of the humidity factors:
like the temperature factor, the pearson correlation coefficient is used as the quantification of the degree of influence of the humidity factor.
The electricity consumption data sequence of the user n is XnThe sequence of humidity data corresponding to time as variable Y1The calculation is carried out by the formula (2). The calculation result represents the quantitative characteristic of the influence degree of the humidity factor on the electricity consumption behavior of the user n and is recorded as Rn,2。
Calculating the quantitative characteristics of the influence degree of the electricity price factors:
for a certain user power sequence X ═ X1,x2,…,xm]Assuming that its corresponding electricity rate sequence D ═ D1,d2,…,dm]. Firstly, constructing an elastic coefficient matrix E of electricity consumption and electricity price, which is shown as the following formula:
in the formula, eij(i 1,2, … m; j 1,2, … m) is the electrovalence elastic modulus, which is defined as:
in the formula: Δ xiAnd Δ diRespectively representing the variation of the electricity demand and the electricity price at the ith time point.
Secondly, calculating an electricity price response rate index epsilon according to the constructed elastic coefficient matrix E, wherein the electricity price response rate index epsilon is calculated as follows:
the calculation result represents the quantitative characteristics of the influence degree of the electricity price on the electricity consumption behavior of the user, and for the user n, the calculation result is recorded as Rn,3。
Calculating the quantitative characteristics of the influence degree of the income factors:
the household income situation also has influence on the electricity consumption of the user, and the household income situation is divided into three types: average power consumption of the wealthy families, the well-off families and the poor families is gradually decreased. The affluent family, the well-being family and the poor family are respectively represented by 2, 1.5 and 1. For user n, notation Rn,4。
Step S203, constructing an electricity utilization comprehensive characteristic vector, which comprises the following steps:
the power utilization curve feature set P of the user n can be obtained through the calculationn＝[Pn,1,Pn,2,…,Pn,11]External influencing factor set of features Rn＝[Rn,1,Rn,2,Rn,3,Rn,4]. Considering the electrical characteristics set PnAnd the influence factor characteristic set RnAre different in dimension and influence factor characteristics RnThere may be outliers and extremes, for which a zero mean criterion is usedThe chemical method is used for processing.
Therefore, the comprehensive characteristic vector of the electricity utilization behavior of the user n can be obtained and recorded as: o isn＝(Pn,Rn). Further, the whole electricity consumption data set Q ═ X can be obtained1,X2,…,XN]Corresponding integrated characteristic set Z ═ O1,O2,…,ON]。
The user electricity utilization characteristic system constructed in the step S2 includes not only the characteristics of the electricity load curve but also the influence degree of various external influence factors, and compared with the conventional characteristic system only considering the characteristics of the electricity utilization curve, the user electricity utilization characteristic system includes more and more comprehensive user electricity utilization information, and provides sufficient data support for classification of subsequent user electricity utilization behaviors.
And step S3, establishing and solving a user electricity utilization behavior clustering model considering external influence factors.
Specifically, step S3 includes:
and S301, constructing an objective function of the clustering model.
The clustering objective function constructed by the invention is the minimum average error criterion function, and is as follows:
in the formula: k is the number of categories; n is the number of users; d (P)n,Pk) Representing a feature vector PnAnd the feature vector PkThe Euclidean distance of (c); w is a1And w2The power usage profile characteristics and the weight of external influencing factors are respectively. The invention equally considers the influence of the electricity utilization characteristics and the influence factor characteristics, and the weight of each eigenvector is calculated according to the ratio of the own dimension to the dimension of the comprehensive eigenvector. u. ofknIs a variable of 0-1, if the nth user belongs to the kth class, then u kn1, otherwise u kn0. To ensure that each data sample belongs to only one class, and each class is not a null set, the variable uknIt should satisfy:
and S302, solving the objective function of the clustering model of the formula (6) by adopting a k-means algorithm.
And solving the power utilization behavior analysis model considering the external influence factors by adopting a k-means algorithm. The input data is a comprehensive characteristic vector Z ═ O corresponding to each power consumption sequence1,O2,…,ON]。
However, in the solving process, the specific value of the optimal clustering number K in the K-means algorithm is difficult to determine. Therefore, the present invention determines the optimal K value according to the following procedure.
First, K is taken as 2,3,4, …, and 20, respectively, for clustering.
Secondly, calculating the clustering effectiveness indexes corresponding to the K values respectively. The invention comprehensively considers 3 types of effectiveness indexes, namely a contour Coefficient (SI), a Calinski-Harabasz Index (CHI) and a Davies-Bouldin Index (DBI).
And finally, selecting the K value with the best clustering effect as the clustering number of the model through the clustering index.
The electricity utilization data can be divided into K types through the comprehensive clustering model, and the clustering center of the K types of electricity utilization users is obtained.
Through the comprehensive clustering in step S3, different types of power users can be effectively distinguished, the clustering centers are as shown in fig. 2(a), and meanwhile, the user types corresponding to different clustering centers are influenced by external factors to different degrees, and the influence degrees are as shown in fig. 2 (b).
Step S4, fine-drawing the electrical image of the user.
The K typical load curve cluster centers obtained by step S3 correspond to K user types. In order to comprehensively and intuitively show the characteristic difference of different types of electricity utilization users, the electricity utilization behaviors of the users are visually presented in a radar map form.
Specifically, step S4 includes:
and S401, determining a user type label.
And respectively determining the user type labels of the K users by analyzing the characteristics of the clustering center curves of the K users. For example, if the clustering center curve of a certain type of user has a distinct peak in the early peak time period and the late peak time period, the user type of the user can be labeled as "early-late peak type" user.
And step S402, determining a user behavior feature label.
The behavior feature label comprises a power utilization curve feature and an influence factor feature. For each user type:
firstly, calculating the power utilization curve characteristic labels of the users. And (4) respectively calculating the clustering centers of the data according to the 11 electricity utilization characteristic indexes in the step S2, and taking the calculated values of all the indexes as parameters for drawing the characteristics of the 11 electricity utilization curves in the radar chart, as shown in the attached figure 3.
Next, the influential factor characteristic labels of the users of the type are calculated. Supposing that the electricity consumption data comprise h electricity load sequences, and the external influence factor characteristic vector corresponding to the jth electricity load sequence is Rj＝[Rj1,Rj2,Rj3,Rj4]Then, the calculation formula of the influence factor label of the type of user is:
in the formula: 1,2, 3,4, respectively, represent a temperature influence factor, a humidity influence factor, an electricity price influence factor, and a revenue influence factor. The calculated value is used as a parameter for drawing the characteristic labels of the 4 influencing factors in the radar map.
Step S403, drawing the user portrait.
And drawing a user image by taking the user type label and the behavior characteristic label obtained in the previous two steps as parameters of the radar map, as shown in the attached figure 3.
According to the clustering result of step S3, a multi-dimensional portrait of the user' S power consumption behavior is depicted, as shown in fig. 3. Step S4 visually and comprehensively depicts the electricity utilization characteristics and behavior characteristics of the users according to the clustering result of the electricity utilization users, and embodies the difference of the electricity utilization behavior characteristics of various users. Each item of power utilization curve feature selected in figure 3 reflects the power utilization level of the user; the temperature and the humidity reflect the sensitivity of the user to meteorological factors; the electricity price reflects the response degree of the user to the electricity price; the income reflects the average family economy of this class of users. Compared with the traditional power utilization behavior analysis display mode, the display mode provided by the invention can visually reflect the multi-dimensional characteristics of the user, and provides powerful support for an electric power company to master the power utilization behavior of the user, mine the requirements of the user and improve the service level.
The invention quantifies and analyzes the relationship between the user and factors such as temperature, humidity, electricity price, income level and the like, and extracts the correlation characteristics between the user and the influencing factors; and then, comprehensively considering the electricity utilization curve characteristics and the influence factor characteristics to classify the electricity utilization behaviors of the user, and depicting the behavior portrait of the user on a classification result.
The above description is for the purpose of illustrating embodiments of the invention and is not intended to limit the invention, and it will be apparent to those skilled in the art that any modification, equivalent replacement, or improvement made without departing from the spirit and principle of the invention shall fall within the protection scope of the invention.
Claims (10)
1. A power utilization behavior characterization method considering external influence factors is characterized by comprising the following steps:
preprocessing the power load data;
constructing a user electricity utilization characteristic system containing external influence factors;
establishing and solving a user power consumption behavior clustering model considering external influence factors;
depicting the user power consumption portrait.
2. The method according to claim 1, wherein the preprocessing of the power load data comprises processing abnormal data and processing data into a standardized form.
3. The method for characterizing power consumption behaviors in consideration of external influence factors according to claim 1, wherein the constructing the user power consumption feature system including the external influence factors comprises:
constructing a user electricity utilization curve feature set;
constructing a characteristic set of external influence factors of power utilization of a user;
and constructing a comprehensive characteristic vector of the power utilization.
4. The method for characterizing power consumption behaviors in consideration of external influence factors according to claim 3, wherein the step of constructing the feature set of the external influence factors for the power consumption of the user comprises the following steps:
calculating the quantitative characteristic R of the influence degree of the temperature factorn,1。；
Calculating the quantitative characteristic R of the influence degree of the humidity factorn,2；
Calculating quantitative characteristic R of influence degree of electricity price factorn,3；
Calculating income factor influence degree quantization characteristic Rn,4；
Obtaining the external influence factor characteristic set Rn＝[Rn,1,Rn,2,Rn,3,Rn,4]。
5. The method for characterizing power consumption behaviors in consideration of external influence factors according to claim 4, wherein the constructing of the power consumption comprehensive feature vector comprises:
and processing the characteristic set of the power utilization curve of the user and the characteristic set of the external influence factors by adopting a zero-mean standardization method to obtain a power utilization comprehensive characteristic vector of the user.
6. The method for characterizing power consumption behaviors in consideration of external influence factors according to claim 1, wherein the establishing and solving the clustering model of the power consumption behaviors of the user in consideration of the external influence factors comprises:
constructing an objective function of the clustering model;
and solving the objective function of the clustering model by adopting a k-means algorithm.
7. The method for characterizing power consumption behaviors in consideration of external influence factors according to claim 6, wherein the optimal clustering number K in the K-means algorithm is determined as follows:
respectively clustering by taking K as 2,3,4, … and 20;
respectively calculating clustering effectiveness indexes corresponding to the K values;
and selecting the K value with the best clustering effect as the clustering number of the model through the clustering index.
8. The method for characterizing power usage behaviors in consideration of external influence factors according to claim 1, wherein the characterizing the user power usage portraits includes:
determining a user type tag;
determining a user behavior feature label;
and drawing a user image according to the user type label and the user behavior characteristic label as the parameters of the radar map.
9. The method for characterizing power consumption behaviors in consideration of external influence factors according to claim 2, wherein the processing of abnormal data comprises removing and correcting the abnormal data:
if the electric load data contains a missing value, but the missing amount is less than 10% of the whole, filling the missing part by adopting a whole averaging method;
if the loss amount of the power consumption load data exceeds 10% of the whole, rejecting the user data;
if the zero value contained in the electric load data exceeds 95% of the whole, rejecting the user data;
if the ith data point x in the electrical load dataiIf the value of (b) is suddenly changed, the data point is eliminated according to xiData points x before and after the pointi-1And xi+1And calculating interpolation for filling.
10. The method according to claim 9, wherein the decision formula for determining whether the data points suddenly change is as follows:
wherein: delta. for the preparation of a coatingiAnd (3) showing the abrupt change degree of the ith data point, and when the value of the abrupt change degree is greater than a specified threshold value epsilon, considering that the load point is abnormal data, wherein the value range of the threshold value epsilon is 0.5-0.8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210145798.3A CN114638284A (en) | 2022-02-17 | 2022-02-17 | Power utilization behavior characterization method considering external influence factors |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210145798.3A CN114638284A (en) | 2022-02-17 | 2022-02-17 | Power utilization behavior characterization method considering external influence factors |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114638284A true CN114638284A (en) | 2022-06-17 |
Family
ID=81946695
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210145798.3A Pending CN114638284A (en) | 2022-02-17 | 2022-02-17 | Power utilization behavior characterization method considering external influence factors |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114638284A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI837819B (en) * | 2022-09-12 | 2024-04-01 | 財團法人資訊工業策進會 | Power consumption behaviors analyzing device and power consumption behaviors analyzing method |
-
2022
- 2022-02-17 CN CN202210145798.3A patent/CN114638284A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI837819B (en) * | 2022-09-12 | 2024-04-01 | 財團法人資訊工業策進會 | Power consumption behaviors analyzing device and power consumption behaviors analyzing method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Fu et al. | Clustering-based short-term load forecasting for residential electricity under the increasing-block pricing tariffs in China | |
| CN110097297A (en) | A kind of various dimensions stealing situation Intellisense method, system, equipment and medium | |
| CN111105104A (en) | Short-term power load prediction method based on similar day and RBF neural network | |
| CN104680261A (en) | Power load operation control method based on load curve clustering of major clients | |
| CN111008726B (en) | Class picture conversion method in power load prediction | |
| CN111667135B (en) | Load structure analysis method based on typical feature extraction | |
| Gajowniczek et al. | Electricity peak demand classification with artificial neural networks | |
| CN111626614A (en) | User classification method based on electric charge recovery | |
| CN111612228A (en) | User electricity consumption behavior analysis method based on electricity consumption information | |
| CN112001441A (en) | Power distribution network line loss anomaly detection method based on Kmeans-AHC hybrid clustering algorithm | |
| CN113283674A (en) | Baseline load prediction correction method based on user electricity utilization characteristics | |
| CN116304537B (en) | Electricity larceny user checking method based on intelligent measuring terminal | |
| CN108664653A (en) | A kind of Medical Consumption client's automatic classification method based on K-means | |
| CN111815026A (en) | Multi-energy system load prediction method based on feature clustering | |
| CN115792370A (en) | Energy utilization diagnosis method, device and equipment based on intelligent electric meter | |
| CN114638284A (en) | Power utilization behavior characterization method considering external influence factors | |
| CN106651093A (en) | Smart re-check management and control method used for low voltage set copy system | |
| CN115952429A (en) | Self-adaptive DBSCAN abnormal battery identification method based on Euclidean distance without prior weight | |
| CN113947444B (en) | Electricity selling package recommendation method considering multi-granularity hesitation fuzzy set and incomplete weight | |
| CN110782140B (en) | Multi-dimensional element evaluation method for electric charge recovery risk screening | |
| Yang et al. | An electricity data cluster analysis method based on SAGA-FCM algorithm | |
| CN115081893A (en) | User electricity consumption data analysis method and device, electronic equipment and readable storage medium | |
| CN114372835A (en) | Comprehensive energy service potential customer identification method, system and computer equipment | |
| CN115907228A (en) | Short-term power load prediction analysis method based on PSO-LSSVM | |
| CN114048200A (en) | User electricity consumption behavior analysis method considering missing data completion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |