CN109726740A - A kind of trade power consumption behavior analysis method based on clustering - Google Patents
A kind of trade power consumption behavior analysis method based on clustering Download PDFInfo
- Publication number
- CN109726740A CN109726740A CN201811479796.8A CN201811479796A CN109726740A CN 109726740 A CN109726740 A CN 109726740A CN 201811479796 A CN201811479796 A CN 201811479796A CN 109726740 A CN109726740 A CN 109726740A
- Authority
- CN
- China
- Prior art keywords
- clustering
- analysis
- data
- consumption behavior
- electricity consumption
- 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
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P80/00—Climate change mitigation technologies for sector-wide applications
- Y02P80/10—Efficient use of energy, e.g. using compressed air or pressurized fluid as energy carrier
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/80—Management or planning
- Y02P90/82—Energy audits or management systems therefor
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to power marketing fields, and in particular to a kind of electricity consumption behavior analysis method.Trade power consumption behavior analysis method of one of the present invention based on clustering includes the following steps: data acquisition, data analysis and research and development of software.Beneficial effects of the present invention: (1) electricity consumption behavior cluster is carried out by branch trade, identifies the electricity consumption behavioural characteristic of power customer in different industries；(2) it is based on cluster result, by formulating science, reasonable, individual character electricity consumption boot policy, network load is reduced, guarantees power grid security, stable operation；(3) by research and development of software, different industries power customer electricity consumption behavior monitoring, identification, analysis are realized.
Description
Technical field
The present invention relates to power marketing fields, and in particular to a kind of electricity consumption behavior analysis method.
Background technique
The fluctuation of power grid power load is big, and power supply and electricity consumption mismatch will cause a large amount of waste, existing electricity consumption behavioural analysis
Method only carries out on-line monitoring and simple statistical analysis to the power load of power customer, does not know that in different industries and is deposited
Part throttle characteristics, can not make science, rationally, the electricity consumption boot policy of individual character.
Summary of the invention
In order to solve the shortcomings of the prior art, the present invention provides a kind of easy to use, accurate industries of analysis to use
Electric behavior analysis method.
Trade power consumption behavior analysis method of one of the present invention based on clustering, includes the following steps:
(1) data acquisition
Internal data obtains: due to historical accumulation, each power consumer in electricity consumption acquisition system generates daily at 96 points and bears
Lotus data cause power load data volume very huge, consider the speed and stability of access, take the mode of open interface
Internal data acquisition is carried out, and is stored with database, while desensitization process and trade classification processing are carried out to data, is ensured
The safety and availability of data；
(2) data are analyzed
Clustering is also referred to as cluster analysis, cluster analysis, refers to that the set by physics or abstract object is grouped by similar
The analytic process of multiple classes of object composition, clustering is a kind of quantitative approach, and from the point of view of data analysis, it is to more
A sample carries out the Multielement statistical analysis method of quantitative analysis, and clustering includes K-means cluster and Hierarchical Clustering；
1. K-means clustering procedure
K-means algorithm is the evaluation index very typically based on the clustering algorithm of distance, using distance as similitude,
Think that the distance of two objects is closer, similarity is bigger, the algorithm think cluster by being formed apart from close object,
Therefore compact and independent cluster is obtained as final goal, using Euclidean distance as similarity measure, it is K-means algorithm
Corresponding a certain initial cluster center vector V optimal classification is sought, so that evaluation index J is minimum, algorithm uses error sum of squares criterion
Function is as clustering criteria function；
2. hierarchical clustering method
Hierarchical clustering method, a kind of method of clustering, way be each sample as a kind of when starting, then
Hithermost sample gathers first for group, then the group having polymerize is remerged by its between class distance, constantly continues, finally
All subclasses are all aggregated to a major class；
Will use between class distance in the process, corresponding to different between class distances, the i.e. not Tongfang of generation system cluster
Method, there are commonly the methods of average between minimum distance method, maximum distance method and group；
Based on both the above clustering procedure, electricity consumption behavior cluster is carried out to the big customer of certain industry, is divided into bimodal pattern, opens in the daytime
Drum, night go into operation type, continuous production type and its alloytype.
Preferably, further including research and development of software；
Based on big data mining analysis technology, relies on R language, Matlab and Python authoring tool to complete clustering algorithm and compile
It writes, the development for carrying out analysis software is designed with computer information technology, bottom data storage processing work is carried out with database
Make, merging for algorithm and page rear end is realized using Java+R, Java+Matlab, Java+Python form, develops collection number
According to acquisition, storage, analysis application and visualize in the customer electricity behavioural characteristic discriminance analysis software of one, according to exploitation
Research and development of products and test job are completed in plan, write analysis software manuals.
Beneficial effects of the present invention:
(1) electricity consumption behavior cluster is carried out by branch trade, identifies the electricity consumption behavioural characteristic of power customer in different industries；
(2) it is based on cluster result, by formulating science, reasonable, individual character electricity consumption boot policy, network load is reduced, protects
Demonstrate,prove power grid security, stable operation；
(3) by research and development of software, different industries power customer electricity consumption behavior monitoring, identification, analysis are realized.
Detailed description of the invention
Fig. 1 is a kind of trade power consumption behavior analysis method flow diagram based on clustering.
Fig. 2 is bimodal pattern schematic diagram data.
Fig. 3 is the type schematic diagram data that goes into operation in the daytime.
Fig. 4 is the type schematic diagram data that goes into operation at night.
Fig. 5 is continuous production type schematic diagram data.
Fig. 6 is its alloytype schematic diagram data.
Fig. 7 is research and development of software interface.
Specific embodiment
The present invention is described in further detail below in conjunction with the accompanying drawings, but this should not be interpreted as to above-mentioned theme of the invention
Range be only limitted to above-described embodiment.
As shown in figs. 1-7, a kind of trade power consumption behavior analysis method based on clustering, includes the following steps:
(1) data acquisition
Internal data obtains: due to historical accumulation, each power consumer in electricity consumption acquisition system generates daily at 96 points and bears
Lotus data cause power load data volume very huge, consider the speed and stability of access, take the mode of open interface
Internal data acquisition is carried out, and is stored with database, while desensitization process and trade classification processing are carried out to data, is ensured
The safety and availability of data.
(2) data are analyzed
Clustering is also referred to as cluster analysis, cluster analysis, refers to that the set by physics or abstract object is grouped by similar
The analytic process of multiple classes of object composition, clustering is a kind of quantitative approach, and from the point of view of data analysis, it is to more
A sample carries out the Multielement statistical analysis method of quantitative analysis.Clustering includes K-means cluster, Hierarchical Clustering etc.；
1. K-means clustering procedure
K-means algorithm is the evaluation index very typically based on the clustering algorithm of distance, using distance as similitude,
Think that the distance of two objects is closer, similarity is bigger.The algorithm think cluster by forming apart from close object,
Therefore handle obtains compact and independent cluster as final goal.Using Euclidean distance as similarity measure, it is K-means algorithm
Corresponding a certain initial cluster center vector V optimal classification is sought, so that evaluation index J is minimum.Algorithm uses error sum of squares criterion
Function is as clustering criteria function；
2. hierarchical clustering method
Hierarchical clustering method (hierarchical cluster method) one translates " hierarchical clustering method ".The one of clustering
Kind method.Its way is when starting using each sample as one kind, then hithermost sample (i.e. apart from the smallest group's product)
Gather first for group, then the group having polymerize is remerged by its between class distance, constantly continues, finally all subclasses
It is aggregated to a major class；
In the process, between class distance is used, corresponding to different between class distances, the i.e. not Tongfang of generation system cluster
Method, there are commonly the methods of average between minimum distance method, maximum distance method, group；
Based on both the above clustering procedure, electricity consumption behavior cluster is carried out to the big customer of certain industry, is divided into bimodal pattern, opens in the daytime
Drum, night go into operation type, continuous production type and its alloytype:
Based on the poly- five seed type user groups out of clustering, all types of user and total load curve are calculated according to the following formula
Similarity, analyze influence of all types of user to total load curve, influence of the smaller expression of distance d to total load curve is bigger
N is total sampling number, and such as sampling in 15 minutes is primary, and sampling sum n=96, i indicate sampling node number within 24 hours one day,
K indicates user group type number, xjiThe total load of all categories user when being sampled for i-th, xkiK class when being sampled for i-th
The load of user；
It is similar to total load curve by calculating all types of user based on the poly- five seed type user groups out of clustering
Degree analyzes influence of all types of user to total load curve, and the influence apart from smaller expression to total load curve is bigger.According to all kinds of
The similarity of user group curve and total load curve, power supply company can analyze influence of the user to total load curve, according to phase
Corresponding sales tactics is formulated like degree.
| Classification | K-means cluster |
| 1st class | 1.572 5E+04 |
| 2nd class | 1.564 4E+04 |
| 3rd class | 1.559 7E+04 |
| 4th class | 1.580 6E+04 |
| 5th class | 1.433 3E+04 |
Similarity is calculated to all types of user groups and total load curve respectively, upper figure can be seen that the user group of the 5th type
Influence to total load curve is the largest.1st class and the 4th class and total load curve distance are maximum；
Based on classification results, power supply company can adjust corresponding confession according to all types of user and the similarity of total load curve
Electric service strategy.It can be adjusted by additional electric service to total load song for the user type being affected to total load
The influence of line, incentive mechanism can be used by influencing lesser user type to total load, thus the fortune for promoting the stabilization of power grids orderly
Row, plays maximum power supply capacity.
(3) research and development of software
Based on big data mining analysis technology, the algorithm for relying on R language, Matlab, Python etc. currently to flow to writes work
Tool is completed clustering algorithm and is write, and the development for carrying out analysis software is designed with computer information technology, carries out bottom with database
Layer data storage processing work, realizes algorithm and page rear end using forms such as Java+R, Java+Matlab, Java+Python
Fusion, develop integrate data acquisition, storage, analysis application and visualize customer electricity behavioural characteristic identify
Software is analyzed, research and development of products and test job are completed according to development plan, write analysis software manuals.
Claims (2)
1. a kind of trade power consumption behavior analysis method based on clustering, includes the following steps:
(1) data acquisition
Internal data obtains: due to historical accumulation, each power consumer in electricity consumption acquisition system generates 96 point load numbers daily
According to, cause power load data volume very huge, consider access speed and stability, take the mode of open interface to carry out
Internal data obtains, and is stored with database, while carrying out desensitization process and trade classification processing to data, ensures data
Safety and availability；
(2) data are analyzed
Clustering is also referred to as cluster analysis, cluster analysis, refers to that the set by physics or abstract object is grouped by similar object
The analytic process of multiple classes of composition, clustering is a kind of quantitative approach, and from the point of view of data analysis, it is to multiple samples
The Multielement statistical analysis method of this progress quantitative analysis, clustering include K-means cluster and Hierarchical Clustering；
1. K-means clustering procedure
K-means algorithm is the evaluation index very typically based on the clustering algorithm of distance, using distance as similitude, that is, is recognized
Distance for two objects is closer, and similarity is bigger；
2. hierarchical clustering method
Hierarchical clustering method, a kind of method of clustering, way be each sample as a kind of when starting, then near
Close sample gathers first for group, then the group having polymerize is remerged by its between class distance, constantly continues, finally one
It cuts subclass and is all aggregated to a major class；
Based on both the above clustering procedure, electricity consumption behavior cluster is carried out to the big customer of certain industry, is divided into bimodal pattern, goes into operation in the daytime
Type, night go into operation type, continuous production type and its alloytype.
2. a kind of trade power consumption behavior analysis method based on clustering according to claim 1, which is characterized in that also
Including research and development of software；
Based on big data mining analysis technology, relies on R language, Matlab and Python authoring tool to complete clustering algorithm and writes,
The development for carrying out analysis software is designed with computer information technology, and bottom data storage processing work is carried out with database,
Merging for algorithm and page rear end is realized using Java+R, Java+Matlab, Java+Python form, is developed and is collected data and adopt
The customer electricity behavioural characteristic discriminance analysis software integrate, store, analyzing application and visualize, according to development plan
Research and development of products and test job are completed, analysis software manuals are write.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811479796.8A CN109726740A (en) | 2018-12-05 | 2018-12-05 | A kind of trade power consumption behavior analysis method based on clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811479796.8A CN109726740A (en) | 2018-12-05 | 2018-12-05 | A kind of trade power consumption behavior analysis method based on clustering |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109726740A true CN109726740A (en) | 2019-05-07 |
Family
ID=66295772
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811479796.8A Pending CN109726740A (en) | 2018-12-05 | 2018-12-05 | A kind of trade power consumption behavior analysis method based on clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109726740A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111539845A (en) * | 2020-04-21 | 2020-08-14 | 国网四川省电力公司电力科学研究院 | Enterprise environment-friendly management and control response studying and judging method based on power consumption mode membership grade |
| CN111881190A (en) * | 2020-08-05 | 2020-11-03 | 厦门力含信息技术服务有限公司 | Key data mining system based on customer portrait |
| CN112417640A (en) * | 2020-09-15 | 2021-02-26 | 国网浙江省电力有限公司湖州供电公司 | Method for evaluating openable capacity of feeder line containing energy storage |
| CN115577258A (en) * | 2022-09-08 | 2023-01-06 | 中国电信股份有限公司 | Vibration signal recognition model training method, motor fault detection method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105488628A (en) * | 2015-11-30 | 2016-04-13 | 国网天津市电力公司 | Electric power big data visualization oriented data mining method |
| CN105654196A (en) * | 2015-12-29 | 2016-06-08 | 中国电力科学研究院 | Adaptive load prediction selection method based on electric power big data |
| CN106296315A (en) * | 2016-11-03 | 2017-01-04 | 广东电网有限责任公司佛山供电局 | Context aware systems based on user power utilization data |
| CN107507038A (en) * | 2017-09-01 | 2017-12-22 | 美林数据技术股份有限公司 | A kind of electricity charge sensitive users analysis method based on stacking and bagging algorithms |
| CN107977737A (en) * | 2017-11-19 | 2018-05-01 | 国网浙江省电力公司信息通信分公司 | Distribution transformer load Forecasting Methodology based on mxnet frame depth neutral nets |
-
2018
- 2018-12-05 CN CN201811479796.8A patent/CN109726740A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105488628A (en) * | 2015-11-30 | 2016-04-13 | 国网天津市电力公司 | Electric power big data visualization oriented data mining method |
| CN105654196A (en) * | 2015-12-29 | 2016-06-08 | 中国电力科学研究院 | Adaptive load prediction selection method based on electric power big data |
| CN106296315A (en) * | 2016-11-03 | 2017-01-04 | 广东电网有限责任公司佛山供电局 | Context aware systems based on user power utilization data |
| CN107507038A (en) * | 2017-09-01 | 2017-12-22 | 美林数据技术股份有限公司 | A kind of electricity charge sensitive users analysis method based on stacking and bagging algorithms |
| CN107977737A (en) * | 2017-11-19 | 2018-05-01 | 国网浙江省电力公司信息通信分公司 | Distribution transformer load Forecasting Methodology based on mxnet frame depth neutral nets |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111539845A (en) * | 2020-04-21 | 2020-08-14 | 国网四川省电力公司电力科学研究院 | Enterprise environment-friendly management and control response studying and judging method based on power consumption mode membership grade |
| CN111881190A (en) * | 2020-08-05 | 2020-11-03 | 厦门力含信息技术服务有限公司 | Key data mining system based on customer portrait |
| CN112417640A (en) * | 2020-09-15 | 2021-02-26 | 国网浙江省电力有限公司湖州供电公司 | Method for evaluating openable capacity of feeder line containing energy storage |
| CN112417640B (en) * | 2020-09-15 | 2024-03-29 | 国网浙江省电力有限公司湖州供电公司 | Energy-storage-containing feeder openable capacity assessment method |
| CN115577258A (en) * | 2022-09-08 | 2023-01-06 | 中国电信股份有限公司 | Vibration signal recognition model training method, motor fault detection method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tsai et al. | Combining multiple feature selection methods for stock prediction: Union, intersection, and multi-intersection approaches | |
| CN109726740A (en) | A kind of trade power consumption behavior analysis method based on clustering | |
| WO2021088499A1 (en) | False invoice issuing identification method and system based on dynamic network representation | |
| CN111324642A (en) | Model algorithm type selection and evaluation method for power grid big data analysis | |
| Park et al. | Explainability of machine learning models for bankruptcy prediction | |
| CN105184455A (en) | High dimension visualized analysis method facing urban electric power data analysis | |
| CN108280541A (en) | Customer service strategies formulating method, device based on random forest and decision tree | |
| CN110069467A (en) | System peak load based on Pearson's coefficient and MapReduce parallel computation clusters extraction method | |
| Lu et al. | A weekly load data mining approach based on hidden Markov model | |
| CN116934486B (en) | Decision evaluation method and system based on deep learning | |
| CN108364191A (en) | Top-tier customer Optimum Identification Method and device based on random forest and logistic regression | |
| CN107808337A (en) | Factor Clustering and device, equipment and storage medium | |
| Janková | A Bibliometric Analysis of Artificial Intelligence Technique in Financial Market. | |
| Das et al. | Stock Market Prediction with Big Data Through Hybridization of Data Mining and Optimized Neural Network Techniques. | |
| Guo et al. | Research on risk features and prediction of China’s crude oil futures market based on machine learning | |
| Waraga et al. | Investigating water consumption patterns through time series clustering | |
| CN110458581A (en) | Merchant business turnover abnormal recognition methods and device | |
| Tiong et al. | Forex prediction engine: Framework, modelling techniques and implementations | |
| Ma | The Research of Stock Predictive Model based on the Combination of CART and DBSCAN | |
| CN115358797A (en) | Comprehensive energy user energy behavior analysis method and system based on cluster analysis method and storage medium | |
| Quan | Stock prediction by searching similar candlestick charts | |
| CN113077108A (en) | Data prediction system for power material configuration requirements | |
| Jatain | Performance Optimization of an Enterprise using Data-Driven Strategy | |
| Felden et al. | Web farming and data warehousing for energy tradefloors | |
| Maung et al. | Comparison of Data Mining Classification Algorithms, C5. 0 and CART for Car Evaluation and Credit Card Information Datasets |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190507 |
|
| RJ01 | Rejection of invention patent application after publication |