CN109615555A - User's exception electricity consumption Activity recognition method and system based on Ensemble Learning Algorithms - Google Patents
User's exception electricity consumption Activity recognition method and system based on Ensemble Learning Algorithms Download PDFInfo
- Publication number
- CN109615555A CN109615555A CN201810859842.0A CN201810859842A CN109615555A CN 109615555 A CN109615555 A CN 109615555A CN 201810859842 A CN201810859842 A CN 201810859842A CN 109615555 A CN109615555 A CN 109615555A
- Authority
- CN
- China
- Prior art keywords
- user
- electric power
- electricity consumption
- data
- power 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.)
- Granted
Links
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
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The present invention discloses user's exception electricity consumption Activity recognition method and system based on Ensemble Learning Algorithms, and wherein recognition methods includes Step 1: obtaining all electric power datas of power grid；Step 2: the electric power data that step 1 obtains is screened, the electric power data for not meeting data protocol call format is rejected；Step 3: the electric power data of step 2 qualification is labeled, notation methods include region, date, the amount of consumption and age of user；Step 4: building learning model, the electric power data of above-mentioned mark is input in learning model according to notation methods；Step 5: the learning model orients doubtful abnormal electricity consumption user according to corresponding learning algorithm.The present invention is according to the region for including in electric power data, the date, the information such as the amount of consumption and age of user, using learning algorithm, quickly finds multiplexing electric abnormality, and accuracy rate is high, and relevant department can effectively be helped to reduce investigation range, and saves human and material resources resource for Utilities Electric Co..
Description
Technical field
The invention belongs to field of data recognition, it is related to user's exception electricity consumption Activity recognition method based on Ensemble Learning Algorithms
And system improves the working efficiency of power utility check on the basis of making full use of available data.
Background technique
With power customer quantity rapid growth, stealing electricity phenomenon is also got worse, and stealing not only compromises power supply company
Economic interests, while also hidden danger is brought to Electrical Safety.Electricity filching means develop to equipment by original plain mode at present
Intelligent, means specialization, the high-tech stealing of behavior hiddenization, implement scale, conventional inspecting method are difficult to obtain evidence.
Summary of the invention
In order to solve the problems, such as that stealing electricity phenomenon is got worse, the present invention is analyzed and processed using the data of statistics, it is intended to
User's exception electricity consumption Activity recognition method and system based on Ensemble Learning Algorithms are provided, can quickly identify doubtful stealing user,
Foundation is provided for power utility check work.
User's exception electricity consumption Activity recognition method based on Ensemble Learning Algorithms, comprising the following steps:
Step 1: obtaining all electric power datas of power grid；
Step 2: the electric power data that step 1 obtains is screened, the electric power number for not meeting data protocol call format is rejected
According to；
Step 3: the electric power data of step 2 qualification is labeled, notation methods include region, date, the amount of consumption and use
The family age；
Step 4: building learning model, the electric power data of above-mentioned mark is input in learning model according to notation methods；
Step 5: the learning model orients doubtful abnormal electricity consumption user according to corresponding learning algorithm.
In a preferred embodiment of the invention, age of user is set as the first weighted value in the notation methods,
Sub-region is set as the second weighted value, and consumption is set as third weighted value, and the amount of consumption is set as the 4th weighted value；Described
The weighted value of one weighted value is maximum, and the sum of the first weighted value, the second weighted value, third weighted value and the 4th weighted value are 1.
In a preferred embodiment of the invention, first weighted value is 0.3-0.6.
In a preferred embodiment of the invention, the electricity consumption behavior of age of user is screened, and by the result of screening into
Row segmentation, and calculate the degree of association of age of user and abnormal electricity consumption user.
In a preferred embodiment of the invention, learning model is constructed by the way of supervised learning, i.e., by step
Four qualified electric power datas are known as training data, and every group of training data has a specific result and establish a learning process,
The immanent structure of study electric power data obtains result reasonably to organize organization data.
It in a preferred embodiment of the invention, further include step 6: the doubtful abnormal electricity consumption user information step of output；
Doubtful abnormal electricity consumption user information will be oriented to integrate, and be periodically delivered in management system, and finally determine abnormal use
Electric user.
In a preferred embodiment of the invention, the regional historical oriented where doubtful abnormal electricity consumption user is recalled
Electric power data and route total losses data, if history electric power data and route total losses data generate deviation simultaneously,
Assert doubtful abnormal electricity consumption user for abnormal electricity consumption user.
User's exception electricity consumption Activity recognition system based on Ensemble Learning Algorithms, comprising:
Electric power data obtains module, and the electric power data obtains the electric power data of all users of module collection；
Screening module screens the electric power data of acquisition, rejects the electric power data for not meeting data protocol call format；
Electric power data is input in the processing module by processing module according to notation methods, and the processing module is oriented doubtful
Like abnormal electricity consumption user.
In a preferred embodiment of the invention, the notation methods of the electric power data include region, the date, the amount of consumption with
And age of user.
In a preferred embodiment of the invention, further include comparison module, the comparison module recall oriented it is doubtful
Regional historical electric power data and route total losses data where abnormal electricity consumption user, if history electric power data and line simultaneously
Road total losses data generate deviation, then assert doubtful abnormal electricity consumption user for abnormal electricity consumption user.
By above technical scheme, the technical effects of the invention are that:
The present invention is calculated according to the region for including in electric power data, date, the information such as the amount of consumption and age of user using study
Method quickly finds multiplexing electric abnormality, and accuracy rate is high, can effectively help relevant department to reduce investigation range, and save for Utilities Electric Co.
Human and material resources.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is the principle of the present invention block diagram.
Fig. 2 is the structural block diagram of system of the invention.
Specific embodiment
In order to be easy to understand the technical means, the creative features, the aims and the efficiencies achieved by the present invention, tie below
Conjunction is specifically illustrating, and the present invention is further explained.
Embodiment 1:
Referring to Fig.1, user's exception electricity consumption Activity recognition method based on Ensemble Learning Algorithms, comprising the following steps:
Step 1: obtaining all electric power datas of power grid；
Step 2: the electric power data that step 1 obtains is screened, the electric power number for not meeting data protocol call format is rejected
According to；
Step 3: the electric power data of step 2 qualification is labeled, notation methods include region, date, the amount of consumption and use
The family age；
Step 4: building learning model, the electric power data of above-mentioned mark is input in learning model according to notation methods；
Step 5: the learning model orients doubtful abnormal electricity consumption user according to corresponding learning algorithm.
It further include step 6: the doubtful abnormal electricity consumption user information step of output；Doubtful abnormal electricity consumption user letter will be oriented
Breath is integrated, and is periodically delivered in management system, and finally determines abnormal electricity consumption user.
The present invention is analyzed according to existing electric power data, integration, the doubtful abnormal electricity consumption user of energy quick lock in, then
Abnormal electricity consumption user is further determined according to doubtful abnormal electricity consumption user, i.e., preferably selects out doubtful exception from mass data
Electricity consumption user provides accuracy so that subsequent processing data are less.
It is above-mentioned to recall the area oriented where doubtful abnormal electricity consumption user in order to further enhance accuracy and speed
Domain history electric power data and route total losses data, if history electric power data and route total losses data generate partially simultaneously
Difference then assert doubtful abnormal electricity consumption user for abnormal electricity consumption user.
Further, age of user is set as the first weighted value in the notation methods, and sub-region is set as the second power
Weight values, and consumption are set as third weighted value, and the amount of consumption is set as the 4th weighted value；The weighted value of first weighted value is most
Greatly, the sum of the first weighted value, the second weighted value, third weighted value and the 4th weighted value are 1；I.e. the present invention has screened 4 power
Weight parameter is determined, and first weighted value (age of user) is 0.3-0.6.
Age of user is preferably selected in the present invention as primary weighted value, the electricity consumption data of in general young user is universal
The electricity consumption data of extra year frequent customer；
Specifically, the electricity consumption behavior of age of user is screened, and the result of screening is split, and calculate age of user with it is different
The degree of association of common electricity user.And carrying out whole identification in combination with the amount of consumption, i.e. the young user amount of consumption is more, accordingly
Electric power data should be also high；Year, frequent customer's amount of consumption was less, and corresponding electric power data is equally less.
Embodiment 2:
The present embodiment 2 compared to implement 1 in be described in detail how to construct learning model:
Learning model is constructed by the way of supervised learning, i.e., the electric power data of step 4 qualification is known as training data, often
Group training data has a specific result and establishes a learning process, learns the immanent structure of electric power data so as to reasonably
Group organization data obtains result.
In learning model, by constantly optimizing, training to entire algorithm, it is proposed that a whole set of inherent study side
Formula constructs the result of electric power data according to the particularity of electric power data.
Embodiment 3:
Compared to the recognition methods that embodiment 1, embodiment 2 provide, specific identifying system is given in the present embodiment:
Parameter Map 2, user's exception electricity consumption Activity recognition system based on Ensemble Learning Algorithms, comprising:
Electric power data obtains module, and the electric power data obtains the electric power data of all users of module collection；
Screening module screens the electric power data of acquisition, rejects the electric power data for not meeting data protocol call format；
Electric power data is input in the processing module by processing module according to notation methods, and the processing module is oriented doubtful
Like abnormal electricity consumption user；The notation methods of the electric power data include region, date, the amount of consumption and age of user.
It further include comparison module, the comparison module recalls the regional historical oriented where doubtful abnormal electricity consumption user
Electric power data and route total losses data, if history electric power data and route total losses data generate deviation simultaneously,
Assert doubtful abnormal electricity consumption user for abnormal electricity consumption user.
The present invention utilizes study according to the region for including in electric power data, date, the information such as the amount of consumption and age of user
Algorithm quickly finds multiplexing electric abnormality, and accuracy rate is high, can effectively help relevant department to reduce investigation range, and save for Utilities Electric Co.
About human and material resources.
Claims (10)
1. user's exception electricity consumption Activity recognition method based on Ensemble Learning Algorithms, which comprises the following steps:
Step 1: obtaining all electric power datas of power grid；
Step 2: the electric power data that step 1 obtains is screened, the electric power number for not meeting data protocol call format is rejected
According to；
Step 3: the electric power data of step 2 qualification is labeled, notation methods include region, date, the amount of consumption and use
The family age；
Step 4: building learning model, the electric power data of above-mentioned mark is input in learning model according to notation methods；
Step 5: the learning model orients doubtful abnormal electricity consumption user according to corresponding learning algorithm.
2. user's exception electricity consumption Activity recognition method according to claim 1 based on Ensemble Learning Algorithms, feature exist
In age of user is set as the first weighted value in the notation methods, and sub-region is set as the second weighted value, and consumption is set
It is set to third weighted value, the amount of consumption is set as the 4th weighted value；The weighted value of first weighted value is maximum, the first weighted value,
The sum of second weighted value, third weighted value and the 4th weighted value are 1.
3. user's exception electricity consumption Activity recognition method according to claim 2 based on Ensemble Learning Algorithms, feature exist
In first weighted value is 0.3-0.6.
4. user's exception electricity consumption Activity recognition method according to claim 1 based on Ensemble Learning Algorithms, feature exist
In being screened to the electricity consumption behavior of age of user, and the result of screening is split, and calculate age of user and abnormal electricity consumption is used
The degree of association at family.
5. user's exception electricity consumption Activity recognition method according to claim 1 based on Ensemble Learning Algorithms, feature exist
In, learning model is constructed by the way of supervised learning, i.e., the electric power data of step 4 qualification is known as training data, every group
Training data has a specific result and establishes a learning process, learns the immanent structure of electric power data so as to reasonably group
Organization data obtains result.
6. user's exception electricity consumption Activity recognition method according to claim 1 based on Ensemble Learning Algorithms, feature exist
In further including step 6: the doubtful abnormal electricity consumption user information step of output；Doubtful abnormal electricity consumption user information will be oriented to carry out
Integration, and be periodically delivered in management system, and finally determine abnormal electricity consumption user.
7. user's exception electricity consumption Activity recognition method according to claim 6 based on Ensemble Learning Algorithms, feature exist
In the regional historical electric power data and route total losses data where having oriented doubtful abnormal electricity consumption user being recalled, if together
When history electric power data and route total losses data generate deviation, then assert that doubtful abnormal electricity consumption user uses for abnormal electricity consumption
Family.
8. user's exception electricity consumption Activity recognition system based on Ensemble Learning Algorithms characterized by comprising
Electric power data obtains module, and the electric power data obtains the electric power data of all users of module collection；
Screening module screens the electric power data of acquisition, rejects the electric power data for not meeting data protocol call format；
Electric power data is input in the processing module by processing module according to notation methods, and the processing module is oriented doubtful
Like abnormal electricity consumption user.
9. user's exception electricity consumption Activity recognition system based on Ensemble Learning Algorithms according to claim 8, which is characterized in that
The notation methods of the electric power data include region, date, the amount of consumption and age of user.
10. user's exception electricity consumption Activity recognition system based on Ensemble Learning Algorithms, feature exist according to claim 9
In further including comparison module, the comparison module recalls the regional historical electric power oriented where doubtful abnormal electricity consumption user
Data and route total losses data are assert if history electric power data and route total losses data generate deviation simultaneously
Doubtful exception electricity consumption user is abnormal electricity consumption user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810859842.0A CN109615555B (en) | 2018-08-01 | 2018-08-01 | User abnormal electricity behavior identification method and system based on integrated learning algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810859842.0A CN109615555B (en) | 2018-08-01 | 2018-08-01 | User abnormal electricity behavior identification method and system based on integrated learning algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109615555A true CN109615555A (en) | 2019-04-12 |
| CN109615555B CN109615555B (en) | 2023-05-05 |
Family
ID=66002633
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810859842.0A Active CN109615555B (en) | 2018-08-01 | 2018-08-01 | User abnormal electricity behavior identification method and system based on integrated learning algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109615555B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103488867A (en) * | 2013-07-16 | 2014-01-01 | 深圳市航天泰瑞捷电子有限公司 | Method for automatically screening abnormal electricity consumption user |
| JP2014079138A (en) * | 2012-10-12 | 2014-05-01 | Toshiba Corp | Monitor system and monitor apparatus for distribution system |
| CN104636816A (en) * | 2013-11-07 | 2015-05-20 | 财团法人资讯工业策进会 | Device and method for establishing power utilization model |
| JP2015171210A (en) * | 2014-03-06 | 2015-09-28 | 三菱電機株式会社 | Support information provision system for home electric appliance |
| CN106355518A (en) * | 2016-11-29 | 2017-01-25 | 国网山东省电力公司电力科学研究院 | Electricity fee payment customer screening method and system |
| CN106651424A (en) * | 2016-09-28 | 2017-05-10 | 国网山东省电力公司电力科学研究院 | Electric power user figure establishment and analysis method based on big data technology |
| CN106780115A (en) * | 2016-11-30 | 2017-05-31 | 国网上海市电力公司 | Abnormal electricity consumption monitoring and alignment system and method |
| CN107330459A (en) * | 2017-06-28 | 2017-11-07 | 联想(北京)有限公司 | A kind of data processing method, device and electronic equipment |
| CN107862427A (en) * | 2017-09-20 | 2018-03-30 | 成都秦川物联网科技股份有限公司 | Water meter energy-conservation reminding method and Internet of things system based on compound Internet of Things |
| CN108198408A (en) * | 2017-12-08 | 2018-06-22 | 囯网河北省电力有限公司电力科学研究院 | A kind of adaptive oppose electricity-stealing monitoring method and system based on power information acquisition system |
-
2018
- 2018-08-01 CN CN201810859842.0A patent/CN109615555B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2014079138A (en) * | 2012-10-12 | 2014-05-01 | Toshiba Corp | Monitor system and monitor apparatus for distribution system |
| CN103488867A (en) * | 2013-07-16 | 2014-01-01 | 深圳市航天泰瑞捷电子有限公司 | Method for automatically screening abnormal electricity consumption user |
| CN104636816A (en) * | 2013-11-07 | 2015-05-20 | 财团法人资讯工业策进会 | Device and method for establishing power utilization model |
| JP2015171210A (en) * | 2014-03-06 | 2015-09-28 | 三菱電機株式会社 | Support information provision system for home electric appliance |
| CN106651424A (en) * | 2016-09-28 | 2017-05-10 | 国网山东省电力公司电力科学研究院 | Electric power user figure establishment and analysis method based on big data technology |
| CN106355518A (en) * | 2016-11-29 | 2017-01-25 | 国网山东省电力公司电力科学研究院 | Electricity fee payment customer screening method and system |
| CN106780115A (en) * | 2016-11-30 | 2017-05-31 | 国网上海市电力公司 | Abnormal electricity consumption monitoring and alignment system and method |
| CN107330459A (en) * | 2017-06-28 | 2017-11-07 | 联想(北京)有限公司 | A kind of data processing method, device and electronic equipment |
| CN107862427A (en) * | 2017-09-20 | 2018-03-30 | 成都秦川物联网科技股份有限公司 | Water meter energy-conservation reminding method and Internet of things system based on compound Internet of Things |
| CN108198408A (en) * | 2017-12-08 | 2018-06-22 | 囯网河北省电力有限公司电力科学研究院 | A kind of adaptive oppose electricity-stealing monitoring method and system based on power information acquisition system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109615555B (en) | 2023-05-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106022592B (en) | Electricity consumption behavior abnormity detection and public security risk early warning method and device | |
| CN110097297A (en) | A kind of various dimensions stealing situation Intellisense method, system, equipment and medium | |
| CN103226736B (en) | Based on the long-medium term power load forecasting method of cluster analysis and grey target theory | |
| CN107679634A (en) | A kind of method that power supply trouble based on data visualization reports analysis and prediction for repairment | |
| CN104573947A (en) | Comprehensive evaluation method for low-voltage transformer areas of regional intelligent distribution network | |
| CN105374209B (en) | A kind of urban area road network running status characteristics information extraction method | |
| CN105184455A (en) | High dimension visualized analysis method facing urban electric power data analysis | |
| CN111861211B (en) | System with double-layer anti-electricity-stealing model | |
| CN109035067A (en) | Building energy consumption processing method and processing device based on RF and ARMA algorithm | |
| CN104408667A (en) | Comprehensive assessment method and system of power quality | |
| CN106650959A (en) | Power distribution network repair ability assessment method based on improved grey clustering | |
| CN103310388A (en) | Method for calculating composite index of grid operation based on information source entropy | |
| CN112286063A (en) | Non-intrusive measurement-based regional energy consumption monitoring system and method | |
| CN105868887A (en) | Building comprehensive energy efficiency analysis method based on subentry measure | |
| CN108154258A (en) | Forecasting Methodology, device, storage medium and the processor of air source heat pump load | |
| CN111179108A (en) | Method and device for predicting power consumption | |
| CN108256724B (en) | Power distribution network open capacity planning method based on dynamic industry coefficient | |
| CN104092215B (en) | Distribution transformer capacity control method and system | |
| CN106056233A (en) | Power load prediction method | |
| CN108734359A (en) | A kind of wind power prediction data preprocessing method | |
| CN108596227A (en) | A kind of leading influence factor method for digging of user power utilization behavior | |
| CN111738876A (en) | Electric power electricity-saving management system based on Internet of things | |
| CN109670998A (en) | Based on the multistage identification of accurate subsidy and system under the big data environment of campus | |
| CN109902743A (en) | A kind of Wind turbines output power predicting method | |
| CN112052985B (en) | Middle-short-term low-voltage prediction method based on lightgbm |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |