CN114066239A - User power consumption abnormity detection method and device - Google Patents
User power consumption abnormity detection method and device Download PDFInfo
- Publication number
- CN114066239A CN114066239A CN202111354910.6A CN202111354910A CN114066239A CN 114066239 A CN114066239 A CN 114066239A CN 202111354910 A CN202111354910 A CN 202111354910A CN 114066239 A CN114066239 A CN 114066239A
- Authority
- CN
- China
- Prior art keywords
- data
- power consumption
- abnormal
- user
- clustering
- 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
- 238000001514 detection method Methods 0.000 title claims description 12
- 230000002159 abnormal effect Effects 0.000 claims abstract description 81
- 238000000034 method Methods 0.000 claims abstract description 38
- 230000005611 electricity Effects 0.000 claims abstract description 34
- 238000007621 cluster analysis Methods 0.000 claims abstract description 12
- 238000012545 processing Methods 0.000 claims description 46
- 238000004422 calculation algorithm Methods 0.000 claims description 23
- 238000000605 extraction Methods 0.000 claims description 22
- 238000010606 normalization Methods 0.000 claims description 19
- 238000007781 pre-processing Methods 0.000 claims description 15
- 238000012937 correction Methods 0.000 claims description 12
- 239000013598 vector Substances 0.000 claims description 12
- 238000003066 decision tree Methods 0.000 claims description 6
- 238000012217 deletion Methods 0.000 claims description 5
- 230000037430 deletion Effects 0.000 claims description 5
- 238000012549 training Methods 0.000 abstract description 16
- 230000008569 process Effects 0.000 abstract description 12
- 238000010801 machine learning Methods 0.000 abstract description 7
- 230000006399 behavior Effects 0.000 description 10
- 238000003491 array Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000005856 abnormality Effects 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 239000002360 explosive Substances 0.000 description 2
- 230000001174 ascending effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 230000007717 exclusion Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- 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)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Strategic Management (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Artificial Intelligence (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Health & Medical Sciences (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Water Supply & Treatment (AREA)
- Probability & Statistics with Applications (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a method and a device for detecting abnormal electricity consumption of a user. According to the method, the cluster analysis and Bayesian decision model are utilized to identify the users with abnormal power consumption, models such as machine learning are not needed, the processes of training the models are reduced, and the accuracy of prediction can be guaranteed within an acceptable range without depending on excessive data quantity.
Description
Technical Field
The invention belongs to the technical field of power consumption analysis, and particularly relates to a method and a device for detecting abnormal power consumption of a user.
Background
With the rapid development of economy, the power consumption also shows explosive growth, and higher requirements are also put forward on abnormal power consumption detection. The abnormal electricity utilization behavior directly influences the electricity charge metering of a power supply company, and the development of a smart power grid is also hindered. Moreover, the electricity stealing behavior taking the change of the metering device as a main means not only damages the electric power facilities, but also easily causes fire disasters and threatens the safe and stable operation of the power grid. Therefore, the power consumer is bound to detect abnormal electricity consumption.
In the prior art, generally, whether the power consumption of a user is abnormal or not is predicted by using a deep learning algorithm or a machine learning algorithm model, and in the algorithm models, the prediction algorithm model is required to train data, a large amount of training data is required, and a quite complex training process is performed to realize the convergence of the model, so that the prediction accuracy required by the user is achieved; on a single user, a small amount of historical data is generally used for training, so that training data of the model is seriously insufficient, effective training cannot be performed, and the model is difficult to converge to meet the requirements of the user.
Disclosure of Invention
In view of the above, the present invention aims to solve the problems that when the current deep learning algorithm or machine learning algorithm model is used for predicting power consumption, training data is seriously insufficient, effective training cannot be performed, and the model is difficult to converge to meet the user requirement.
In order to solve the technical problems, the invention provides the following technical scheme:
in a first aspect, the present invention provides a method for detecting abnormal electricity consumption of a user, including the following steps:
collecting power consumption data and preprocessing the data;
performing feature extraction on the preprocessed power consumption data;
clustering the extracted data features, and obtaining users with abnormal power consumption according to clustering results;
and carrying out power consumption abnormity decision on the users with abnormal power consumption by utilizing a Bayesian decision model.
Further, clustering the extracted data features, and obtaining the users with abnormal power consumption according to the clustering result specifically comprises:
clustering the data characteristics by adopting a clustering method based on Euclidean distance;
defining a clustering center of power consumption data under a normal power consumption behavior;
and taking the user corresponding to the power consumption data which is deviated relative to the clustering center in the clustering result as the abnormal power consumption user.
Further, the feature extraction of the preprocessed power consumption data specifically comprises:
and performing feature extraction according to the time, the temperature of the day, whether the weekend is present, whether the holiday is present and the current power consumption data to construct a data feature vector of the power consumption data.
Further, the collecting of the power consumption data specifically includes:
and acquiring the power consumption data of the user according to a preset time point by the arranged data acquisition equipment.
Further, the preprocessing of the power consumption data specifically includes:
marking the collected power consumption data;
deleting the marked acquired data by using a decision tree mode or a sorting and merging algorithm of a rough set theory;
carrying out exception correction processing on the acquired data subjected to deletion processing by adopting a three-sigma rule;
performing data completion processing on the acquired data subjected to the abnormal correction processing by using an interpolation algorithm;
carrying out normalization processing on the acquired data subjected to data completion processing by adopting a maximum and minimum normalization method;
and dividing the acquired data after the normalization processing into a plurality of data sets.
In a second aspect, the present invention provides an apparatus for detecting abnormal electricity consumption of a user, comprising:
the data acquisition module is used for acquiring power consumption data and preprocessing the power consumption data;
the characteristic extraction module is used for extracting the characteristics of the preprocessed power consumption data;
the cluster analysis module is used for clustering the extracted data characteristics and obtaining users with abnormal power consumption according to a clustering result;
and the abnormal decision module is used for carrying out power consumption abnormal decision on the power consumption abnormal user by utilizing the Bayesian decision model.
Further, the cluster analysis module is specifically configured to:
clustering the data characteristics by adopting a clustering method based on Euclidean distance;
defining a clustering center of power consumption data under a normal power consumption behavior;
and taking the user corresponding to the power consumption data which is deviated relative to the clustering center in the clustering result as the abnormal power consumption user.
Further, the feature extraction module is specifically configured to:
and performing feature extraction according to the time, the temperature of the day, whether the weekend is present, whether the holiday is present and the current power consumption data to construct a data feature vector of the power consumption data.
Further, the data acquisition module comprises: a power consumption acquisition module;
the power consumption acquisition module is used for acquiring power consumption data of the user according to a preset time point through the set data acquisition equipment.
Further, the data acquisition module includes a data preprocessing module, and the data preprocessing module is specifically configured to:
marking the collected power consumption data;
deleting the marked acquired data by using a decision tree mode or a sorting and merging algorithm of a rough set theory;
carrying out exception correction processing on the acquired data subjected to deletion processing by adopting a three-sigma rule;
performing data completion processing on the acquired data subjected to the abnormal correction processing by using an interpolation algorithm;
carrying out normalization processing on the acquired data subjected to data completion processing by adopting a maximum and minimum normalization method;
and dividing the acquired data after the normalization processing into a plurality of data sets.
In summary, the invention provides a method and a device for detecting abnormal electricity consumption of a user, wherein data characteristics of electricity consumption of the user are extracted after the data of the electricity consumption of the user are collected and processed, clustering analysis is carried out on the basis of the extracted data characteristics, users with abnormal electricity consumption are obtained preliminarily, and decision processing is carried out on all users with abnormal electricity consumption by utilizing a Bayesian decision model, so that the abnormal users are finally determined. According to the method, the cluster analysis and Bayesian decision model are utilized to identify the users with abnormal power consumption, models such as machine learning are not needed, the processes of training the models are reduced, and the accuracy of prediction can be guaranteed within an acceptable range without depending on excessive data quantity.
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, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without inventive exercise.
Fig. 1 is a schematic flow chart of a method for detecting abnormal power consumption of a user according to an embodiment of the present invention.
Detailed Description
In order to make the objects, features and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the embodiments described below 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.
With the rapid development of economy, the power consumption also shows explosive growth, and higher requirements are also put forward on abnormal power consumption detection. The abnormal electricity utilization behavior directly influences the electricity charge metering of a power supply company, and the development of a smart power grid is also hindered. Moreover, the electricity stealing behavior taking the change of the metering device as a main means not only damages the electric power facilities, but also easily causes fire disasters and threatens the safe and stable operation of the power grid. Therefore, the power consumer is bound to detect abnormal electricity consumption.
In the prior art, generally, whether the power consumption of a user is abnormal or not is predicted by using a deep learning algorithm or a machine learning algorithm model, and in the algorithm models, the prediction algorithm model is required to train data, a large amount of training data is required, and a quite complex training process is performed to realize the convergence of the model, so that the prediction accuracy required by the user is achieved; on a single user, a small amount of historical data is generally used for training, so that training data of the model is seriously insufficient, effective training cannot be performed, and the model is difficult to converge to meet the requirements of the user.
Based on this, the embodiment of the invention provides a method and a device for detecting abnormal electricity consumption of a user.
The following is a detailed description of an embodiment of a method for detecting abnormal electricity consumption of a user according to the present invention.
Referring to fig. 1, the present embodiment provides a method for detecting abnormal power consumption of a user, including:
s100: and collecting power consumption data and preprocessing the data.
It should be noted that, through the set data acquisition device (smart meter), the used power consumption data (power consumption) is acquired according to the preset time points (for example, data acquisition is performed at each time point in sequence, that is, data acquisition is performed at all points of 0, 1, 2, 3, 4, etc. for 24 hours each day, and 24 data are acquired each day), and then the acquired data is obtained.
S200: and performing feature extraction on the preprocessed power consumption data.
The collected data is labeled according to the collected date, the collected time period and whether the collected date corresponds to a holiday or a weekend or not; after the labeling is completed, data cleaning, missing completion, data normalization processing and the like need to be performed on the data.
Deleting the invalid and repeated data of the marked acquired data by using a decision tree mode of a rough set theory or a sorting and merging algorithm to obtain the deleted marked acquired data; then carrying out abnormal correction processing on the data, and judging abnormal values by adopting a three-sigma rule to realize abnormal data correction, wherein the abnormal values are values of which the deviation from the average value exceeds 3 times of standard deviation in a group of measurement values; the three sigma criterion is also called Lauda criterion, which is that firstly, a group of detection data is supposed to only contain random errors, the detection data is calculated to obtain standard deviation, an interval is determined according to a certain probability, the error exceeding the interval is considered not to belong to the random errors but to be coarse errors, and the data containing the errors are removed; in the completion missing processing process, an interpolation algorithm is adopted to realize completion missing data, and concretely, Lagrange interpolation, Newton interpolation, Hermite interpolation, segmented interpolation, spline interpolation and the like can be used for realizing the completion missing data.
The specific implementation process of the sorting and merging algorithm is as follows: totally divided into two steps, and treated with the same treatment; when merging and sorting, firstly, dividing an array to be sorted into two parts, namely Left and Right, and if the two parts are sorted, namely, the child arrays Left and Right are both ordered arrays; combining the two arrays, wherein the combining mode is that firstly, an array with the same capacity as the original array is created to store data during combination, then the arrays in Left and Right are compared, if Left [0] < Right [0], the Left [0] is put into the 0 index of the new array, then Left [1] and Right [0] are compared, and all the arrays in Left and Right can be copied into the new array according to a certain sequence by analogy of ascending or descending, and at the moment, the array sorting is finished; merging and sorting an array is subjected to limited-time division, then the array is subjected to merging with the same times, and the whole array is sorted.
The complemented data is subjected to normalization processing, a maximum and minimum normalization method is adopted for data normalization processing in the application, and the processed data is convenient for subsequent calculation and feature extraction.
Storing the normalized data and dividing the normalized data into data sets, wherein each data set contains normalized data for a plurality of consecutive days; generally, the time can be 7 days, and the time can be set according to relevant conditions.
Data feature extraction is required after pre-processing the acquired data.
Specifically, feature extraction processing is performed on each data set, in this embodiment, feature extraction is performed according to time, temperature of the day, whether weekend or not, whether holiday or not, power consumption data at that time, and a feature vector of each relevant data is constructed, that is, the feature vector includes a time period, temperature of the day, whether weekend or not, whether holiday or not, and power consumption data at that time; wherein, whether weekends and holidays are represented by 0 or 1 is represented by 1, and whether holidays are represented by 0; a set of feature vectors may be formed.
S300: and clustering the extracted data features, and obtaining users with abnormal power consumption according to clustering results.
It should be noted that the purpose of clustering is to say that the data with higher similarity in the data are grouped, the data with large differences are in different groups, the clustering process is based on a feature vector formed by a series of features associated with the data features, and the similarity degree between the data is described through the similarity degree or distance quantity; thus making XidA feature vector for an ith time period for a day d of the user; in the clustering process, the dimension of each clustering prototype or clustering center is the same as the dimension of the data characteristic vector, namely the set of the clustering centers is V, and in the clustering based on the Euclidean distance, the distance center is realIt is actually the geometric center of all data from that center, i.e., the average; in cluster anomaly detection, the data in the cluster center set V can be used for representing a normal state, so that the distance between the data and the corresponding data in the V needs to be calculated, and the distance is used for representing the deviation degree of the electricity utilization behavior of a user relative to the normal electricity utilization behavior at a certain moment; i.e. for a certain point in time i, y (X)idV) is equal to the minimum distance of the data to the cluster center, i.e.:
y(Xij|V)=min d(Xij,v)
wherein, y (X)ijV) represents that the electricity consumption of the user at the ith moment of the jth day may have abnormal conditions; v represents a cluster center set; xijA feature vector representing the ith time of day j; v denotes a cluster center.
S400: and carrying out power consumption abnormity decision on the users with abnormal power consumption by utilizing a Bayesian decision model.
It should be noted that, when the clustering result may have an abnormality, the bayesian decision model may be used to perform a decision process for the abnormality of the power consumption of the user. The overall steps are that the posterior probability is firstly calculated, then the risk is calculated, and finally the decision processing is carried out. The specific process is as follows:
dividing the decision processing result into two conditions, wherein one condition is that the power utilization of the user is abnormal, the other condition is that the power utilization of the user is normal, H is 1 to represent that the power utilization of the user is normal, and H is 0 to represent that the power utilization of the user is abnormal; then, knowing the bayesian prior probability g ═ P (H ═ 1), P (H ═ 0) ═ 1-g due to mutual exclusion, i.e. according to bayesian theory, the posterior probability is: p ═ P (H ═ 1| S)ij)=(P(Sij|H=1)g)/(P(Sij|H=0)(1-g)+(Sij1) g), wherein SijAs a decision score, it is compared with XijX with correlation, i.e. possible abnormalityijAnd X under the condition of normal power utilizationijThe difference in (c) is the decision score.
The risk function defining the correct decision (i.e., G ═ 0) and the wrong decision (i.e., G ═ 1) at the time of decision recognition is Enm(ii) a Wherein n, m is 0 or 1, when n ═ m, EnmWhen n is not equal to m, Enm1. The bayesian conditional risk is then: f ═ F { Enm}=E00P(G=0,H=0)+E01P(G=0,H=1)+E10P(G=1,H=0)+E11P(G=1,H=1)。
And identifying an authentication decision behavior according to the minimum risk Bayesian decision rule:
if P (S)ij|H=1)/P(Sij|H=0)>((1-g)/g)*((E10-E00)/(E01-E11) G ═ 1); in other cases, G is 0.
In order to simplify the threshold, a risk function of 0 to 1 is determined, and when the risk function is correctly judged, the risk is 0, and when the risk function is wrongly judged, the risk is 1; let g ═ P (H ═ 1) ═ 1/2; i.e. the recognition passing probability and the recognition failing probability are considered to be equal, in this case, the decision function is: p (S)ij|H=1)/P(Sij|H=0)>1, G ═ 1; in other cases, G is 0. Namely, when the Bayesian decision model is used for carrying out decision processing on user abnormity, the decision score S is observed when the user abnormity is respectively substituted into the user power utilization abnormityijThe probability of (A) and the observed decision score when the electricity consumption of the user is abnormal are SijIs greater than 1, the identification is considered to be at risk. Otherwise, there is no risk.
The embodiment provides a method for detecting abnormal power consumption of a user, which includes the steps of collecting and processing power consumption data of the user, extracting data characteristics of the power consumption, carrying out cluster analysis based on the extracted data characteristics to obtain users with abnormal power consumption preliminarily, and then carrying out decision processing on all users with abnormal power consumption by using a Bayesian decision model to finally determine abnormal users. According to the method, the cluster analysis and Bayesian decision model are utilized to identify the users with abnormal power consumption, models such as machine learning are not needed, the processes of training the models are reduced, and the accuracy of prediction can be guaranteed within an acceptable range without depending on excessive data quantity.
The above is a detailed description of an embodiment of the method for detecting abnormality in user power consumption according to the present invention, and the following is a detailed description of an embodiment of the apparatus for detecting abnormality in user power consumption according to the present invention.
The present embodiment provides an abnormal amount of electricity detection apparatus for a user, including: the device comprises a data acquisition module, a feature extraction module, a cluster analysis module and an abnormity decision module.
In this embodiment, the data acquisition module is configured to acquire power consumption data and perform preprocessing.
It should be noted that the data acquisition module specifically includes a power consumption acquisition module and a data preprocessing module.
The power consumption acquisition module is used for acquiring power consumption data of a user according to a preset time point through the set data acquisition equipment.
The data preprocessing module is specifically configured to: marking the collected power consumption data; deleting the marked acquired data by using a decision tree mode or a sorting and merging algorithm of a rough set theory; carrying out exception correction processing on the acquired data subjected to deletion processing by adopting a three-sigma rule; performing data completion processing on the acquired data subjected to the abnormal correction processing by using an interpolation algorithm; carrying out normalization processing on the acquired data subjected to data completion processing by adopting a maximum and minimum normalization method; and dividing the acquired data after the normalization processing into a plurality of data sets.
In this embodiment, the feature extraction module is configured to perform feature extraction on the preprocessed power consumption data.
It should be noted that the feature extraction module is specifically configured to: and performing feature extraction according to the time, the temperature of the day, whether the weekend is present, whether the holiday is present and the current power consumption data to construct a data feature vector of the power consumption data.
In this embodiment, the cluster analysis module is configured to cluster the extracted data features, and obtain a user with abnormal power consumption according to a clustering result.
When it needs to be explained, the cluster analysis module is specifically configured to: clustering the data characteristics by adopting a clustering method based on Euclidean distance; defining a clustering center of power consumption data under a normal power consumption behavior; and taking the user corresponding to the power consumption data which is deviated relative to the clustering center in the clustering result as the abnormal power consumption user.
In this embodiment, the abnormal decision module is configured to perform a power consumption abnormal decision on a power consumption abnormal user by using a bayesian decision model.
It should be noted that the apparatus for detecting abnormal electricity consumption of a user provided in this embodiment is used to implement the method for detecting abnormal electricity consumption of a user corresponding to the foregoing embodiment, and the specific settings of the modules of the apparatus are based on implementing the method, which is not described herein again.
The embodiment provides a user power consumption abnormity detection device, which is characterized in that data characteristics of power consumption of a user are extracted after the data of the power consumption of the user are collected and processed, clustering analysis is carried out on the basis of the extracted data characteristics, users with abnormal power consumption are obtained preliminarily, and decision processing is carried out on all users with abnormal power consumption by utilizing a Bayesian decision model, so that abnormal users are finally determined. According to the method, the cluster analysis and Bayesian decision model are utilized to identify the users with abnormal power consumption, models such as machine learning are not needed, the processes of training the models are reduced, and the accuracy of prediction can be guaranteed within an acceptable range without depending on excessive data quantity.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for detecting abnormal electricity consumption of a user is characterized by comprising the following steps:
collecting power consumption data and preprocessing the data;
performing feature extraction on the preprocessed power consumption data;
clustering the extracted data features, and obtaining users with abnormal power consumption according to clustering results;
and carrying out power consumption abnormity decision on the users with abnormal power consumption by utilizing a Bayesian decision model.
2. The method for detecting abnormal power consumption of a user according to claim 1, wherein the clustering the extracted data features and obtaining the abnormal power consumption user according to the clustering result specifically comprises:
clustering the data characteristics by adopting a clustering method based on Euclidean distance;
defining a clustering center of power consumption data under a normal power consumption behavior;
and taking the user corresponding to the power consumption data which is deviated relative to the clustering center in the clustering result as the abnormal power consumption user.
3. The method for detecting abnormal electricity consumption of a user according to claim 1, wherein the feature extraction of the preprocessed electricity consumption data is specifically as follows:
and performing feature extraction according to time, temperature of the current day, whether weekends exist, whether holidays exist and current power utilization data, and constructing a data feature vector of the power utilization data.
4. The method for detecting abnormal electricity consumption of a user according to claim 1, wherein the collecting electricity consumption data specifically comprises:
and acquiring the power consumption data of the user according to a preset time point by the arranged data acquisition equipment.
5. The method for detecting abnormal electricity consumption of a user according to claim 1, wherein the preprocessing of the electricity consumption data specifically comprises:
marking the collected power consumption data;
deleting the marked acquired data by using a decision tree mode or a sorting and merging algorithm of a rough set theory;
carrying out exception correction processing on the acquired data subjected to deletion processing by adopting a three-sigma rule;
performing data completion processing on the acquired data subjected to the abnormal correction processing by using an interpolation algorithm;
performing normalization processing on the acquired data subjected to data completion processing by adopting a maximum and minimum normalization method;
and dividing the acquired data after normalization into a plurality of data sets.
6. An abnormal amount of electricity detection apparatus for a user, comprising:
the data acquisition module is used for acquiring power consumption data and preprocessing the power consumption data;
the characteristic extraction module is used for extracting the characteristics of the preprocessed power consumption data;
the cluster analysis module is used for clustering the extracted data characteristics and obtaining users with abnormal power consumption according to a clustering result;
and the abnormal decision module is used for carrying out power consumption abnormal decision on the power consumption abnormal user by utilizing a Bayesian decision model.
7. The apparatus for detecting abnormal electricity consumption of a user according to claim 6, wherein the cluster analysis module is specifically configured to:
clustering the data characteristics by adopting a clustering method based on Euclidean distance;
defining a clustering center of power consumption data under a normal power consumption behavior;
and taking the user corresponding to the power consumption data which is deviated relative to the clustering center in the clustering result as the abnormal power consumption user.
8. The apparatus according to claim 6, wherein the feature extraction module is specifically configured to:
and performing feature extraction according to time, temperature of the current day, whether weekends exist, whether holidays exist and current power utilization data, and constructing a data feature vector of the power utilization data.
9. The abnormal power consumption detection device of claim 6, wherein the data acquisition module comprises: a power consumption acquisition module;
the power consumption acquisition module is used for acquiring power consumption data of the user according to a preset time point through the set data acquisition equipment.
10. The abnormal user power consumption detection device according to claim 6, wherein the data acquisition module comprises a data preprocessing module, and the data preprocessing module is specifically configured to:
marking the collected power consumption data;
deleting the marked acquired data by using a decision tree mode or a sorting and merging algorithm of a rough set theory;
carrying out exception correction processing on the acquired data subjected to deletion processing by adopting a three-sigma rule;
performing data completion processing on the acquired data subjected to the abnormal correction processing by using an interpolation algorithm;
performing normalization processing on the acquired data subjected to data completion processing by adopting a maximum and minimum normalization method;
and dividing the acquired data after normalization into a plurality of data sets.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111354910.6A CN114066239A (en) | 2021-11-16 | 2021-11-16 | User power consumption abnormity detection method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111354910.6A CN114066239A (en) | 2021-11-16 | 2021-11-16 | User power consumption abnormity detection method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114066239A true CN114066239A (en) | 2022-02-18 |
Family
ID=80272624
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111354910.6A Pending CN114066239A (en) | 2021-11-16 | 2021-11-16 | User power consumption abnormity detection method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114066239A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115951295A (en) * | 2022-11-11 | 2023-04-11 | 国网山东省电力公司营销服务中心(计量中心) | Automatic identification method and system for daily clear power abnormity |
-
2021
- 2021-11-16 CN CN202111354910.6A patent/CN114066239A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115951295A (en) * | 2022-11-11 | 2023-04-11 | 国网山东省电力公司营销服务中心(计量中心) | Automatic identification method and system for daily clear power abnormity |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111401460B (en) | Abnormal electric quantity data identification method based on limit value learning | |
CN112699913A (en) | Transformer area household variable relation abnormity diagnosis method and device | |
CN111444615B (en) | Photovoltaic array fault diagnosis method based on K nearest neighbor and IV curve | |
CN109740648A (en) | Electric load disorder data recognition method, apparatus and computer equipment | |
CN110288039B (en) | Electricity stealing detection method based on user electricity load characteristics | |
CN107230013B (en) | Method for identifying abnormal power consumption and time positioning of distribution network users under unsupervised learning | |
CN110795690A (en) | Wind power plant operation abnormal data detection method | |
CN111126820A (en) | Electricity stealing prevention method and system | |
CN113177357A (en) | Transient stability assessment method for power system | |
CN111625399A (en) | Method and system for recovering metering data | |
CN112070180B (en) | Power grid equipment state judging method and device based on information physical bilateral data | |
CN115021679A (en) | Photovoltaic equipment fault detection method based on multi-dimensional outlier detection | |
CN114611738A (en) | Load prediction method based on user electricity consumption behavior analysis | |
CN111612149A (en) | Main network line state detection method, system and medium based on decision tree | |
CN107402859A (en) | Software function verification system and verification method thereof | |
CN114066239A (en) | User power consumption abnormity detection method and device | |
CN114460481A (en) | Energy storage battery thermal runaway early warning method based on Bi-LSTM and attention mechanism | |
CN107274025B (en) | System and method for realizing intelligent identification and management of power consumption mode | |
CN117495422A (en) | Cost management system and method based on power communication network construction | |
CN117113086A (en) | Energy storage unit load prediction method, system, electronic equipment and medium | |
CN116933201A (en) | Method and system for identifying illegal electricity utilization behavior of low-voltage charging pile | |
CN116720095A (en) | Electrical characteristic signal clustering method for optimizing fuzzy C-means based on genetic algorithm | |
CN116433049A (en) | Power consumption abnormality detection method based on fuzzy rough entropy | |
CN114330440B (en) | Distributed power supply load abnormality identification method and system based on simulation learning discrimination | |
CN115936926A (en) | SMOTE-GBDT-based unbalanced electricity stealing data classification method and device, computer equipment and storage medium |
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 |