CN113935410A - Electric power customer portrait method based on cross-correlation density clustering - Google Patents
Electric power customer portrait method based on cross-correlation density clustering Download PDFInfo
- Publication number
- CN113935410A CN113935410A CN202111192403.7A CN202111192403A CN113935410A CN 113935410 A CN113935410 A CN 113935410A CN 202111192403 A CN202111192403 A CN 202111192403A CN 113935410 A CN113935410 A CN 113935410A
- Authority
- CN
- China
- Prior art keywords
- samples
- clustering
- iou
- cross
- sample
- 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
- 238000000034 method Methods 0.000 title claims abstract description 16
- 230000005611 electricity Effects 0.000 claims abstract description 19
- 239000011159 matrix material Substances 0.000 claims abstract description 13
- 238000012545 processing Methods 0.000 claims abstract description 9
- 238000001514 detection method Methods 0.000 claims abstract description 4
- 238000010606 normalization Methods 0.000 claims abstract description 4
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000001932 seasonal effect Effects 0.000 claims description 3
- 230000035945 sensitivity Effects 0.000 claims description 3
- 238000003384 imaging method Methods 0.000 claims 2
- 230000006399 behavior Effects 0.000 abstract 2
- 238000005516 engineering process Methods 0.000 description 3
- 238000005259 measurement Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001739 density measurement Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- 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
-
- 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
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Resources & Organizations (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of customer management in the power industry, in particular to a power customer portrait method based on intersection specific density clustering. Collecting a sample set, and carrying out missing value processing, consistency check, noise detection and normalization processing on the sample set by using a sample; constructing a power consumption behavior tag of a power customer; assigning the finished labels and then clustering the electricity utilization behaviors: calculating the mean value of all samples in each cluster according to the primary clustering result; and calculating the similarity among different clusters by using the obtained mean value to obtain a cluster similarity matrix, and then merging the potential similar clusters according to the similarity until the final cluster number is obtained. The invention relates to an electric power customer portrait method based on intersection specific density clustering, which can provide differentiated services for different power consumption groups.
Description
Technical Field
The invention relates to the technical field of customer management in the power industry, in particular to a power customer portrait method based on intersection specific density clustering.
Background
At present, due to factors such as the improvement of the electrification level, the popularization of social informatization and the like, power supply enterprises begin to show a plurality of defects, and the contradiction between electric power marketing and high-quality service is increasingly excited. Customers are beginning to have increased demands on power quality, and even though power supply enterprises are always improving customer satisfaction, the number of complaints is still large every year. Furthermore, power marketing systems have stocked large amounts of power data, but these data are not being utilized efficiently. Nowadays, big data technology has become a trend, and many industries have established marketing systems related to big data, established enterprise client images, and helped enterprises to perform accurate marketing and information recommendation, and have been applied to industries such as finance, telecommunication, traffic, medical education and the like. Many industries at home and abroad carry out customer portrait in a marketing system, realize accurate marketing and personalized recommendation and the like, and promote multi-channel development. With the deepening of power system innovation, the power marketing department also sets forth power customer portraits, and the aim is to prepare in advance after the power selling side is released in terms of humanistic care and customer satisfaction, so that the application research of the power marketing informationized customer portraits has profound significance. The power grid enterprise has mass power data for many years, and the data resource is sufficiently reserved. The power data is fused by means of big data technologies such as multi-data fusion, data analysis mining, visualization, data storage and processing, and the like, so that the customer portrait of the power consumer becomes possible in the technical level. The application research of the electric power marketing information-based customer portrait has execution conditions in the aspects of requirements, resources and technologies, and is a guarantee for better serving users and improving the satisfaction degree of electric power users.
In the existing density clustering algorithm, a cut-off distance needs to be defined according to the requirement of the data set scale, however, an objective measurement standard is not provided to measure whether the data set is large-scale or small-scale, and different density measurement criteria and cut-off distances can cause large influence on a clustering result. Meanwhile, the allocation strategies of different categories are prone to generate a continuous error of category allocation, that is, once a certain sample is allocated incorrectly, a subsequent series of sample allocation errors can be caused.
Therefore, a power customer portrait method based on cross-comparison density clustering is provided.
Disclosure of Invention
The invention aims to provide a power customer portrait method based on cross-comparison density clustering, which aims to solve the problem that the conventional density clustering algorithm cannot define a truncation distance according to the size of a data set, does not have an objective measurement standard to measure whether the data set is large-scale or small-scale, and the distribution strategies of different categories are easy to generate category distribution errors.
A power customer portrait method based on cross-comparison density clustering comprises the following steps:
s1: collecting a sample set, and carrying out missing value processing, consistency check, noise detection and normalization processing on the sample set by using the sample;
s2: constructing a power consumption behavior tag of a power customer;
s3: assigning values to the labels completed in step S2 and then clustering the electricity usage behavior:
s4: calculating the mean value of all samples in each cluster according to the primary clustering result obtained in the step S3;
s5: and calculating the similarity among different clusters by using the average value obtained in the step S5 to obtain a cluster similarity matrix, and then merging the potential similar clusters according to the similarity until the final cluster number is obtained.
Further, in step S1, the sample set is X ═ X (X)1,x2,...,xm) Where m is the number of samples.
Further, in step S2, the electricity consumption behavior labels of the electricity consumers include more than capacity, electricity type, seasonal characteristics, temperature sensitivity, load stability, capacity utilization, holiday electricity characteristics, peak characteristics, and electricity consumption increase and decrease.
Further, in step S3, the electrical behavior clustering step includes:
s3.2: computing arbitrary samples x from the distance matrixiK number of neighboring samples Ni(k);
S3.3: calculating a sample xiOf neighboring samples Ni(k) And each sample remainedxpOf neighboring samples Np(k) The cross-over-cross-over ratio of (c),
s3.4: clustering intersection ratio IOU;
wherein d isijIs xiAnd xjDistance between, Ni(k) Is a neighbor sample, Np(k) For each sample x remainingpThe IOU is the cross-over ratio.
Further, in step S3.4, the clustering method of the intersection ratio IOU includes:
s3.4.1: when IOU is greater than threshold, then xiAnd its neighbor samples Ni (k) and xpAnd its neighbor samples np (k) are grouped into a cluster and these samples are removed from X and no longer participate in the calculation of IOU;
s3.4.2: when x ispE Ni (k) and IOU is less than the threshold, then xpAnd the samples in Np (k) that intersect with Ni (k) still belong to xiOne class, np (k) other samples are other classes and still participate in IOU calculations;
s3.4.3: when in useAnd IOU is less than the threshold, the samples in Np (k) intersected with Ni (k) still belong to xiClass I, xpAnd np (k) where the other samples are of other classes and still participate in the IOU computation.
in summary, due to the adoption of the technical scheme, the beneficial technical effects of the invention are as follows:
the power customer portrait method based on cross-comparison density clustering can intuitively and quickly master power consumption behavior characteristics of different power customer groups, provides support for power companies to carry out power customer load prediction, and meanwhile, the power companies can also provide power consumption suggestions for power customers, specify good power price policies and provide differentiated services for different power customer groups.
Drawings
FIG. one is a flow chart of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
A power customer portrait method based on cross-comparison density clustering comprises the following steps:
s1: collecting a sample set, and carrying out missing value processing, consistency check, noise detection and normalization processing on the sample set by using the sample;
s2: constructing a power consumption behavior tag of a power customer;
s3: assigning values to the labels completed in step S2 and then clustering the electricity usage behavior:
s4: calculating the mean value of all samples in each cluster according to the primary clustering result obtained in the step S3;
s5: and calculating the similarity among different clusters by using the average value obtained in the step S5 to obtain a cluster similarity matrix, and then merging the potential similar clusters according to the similarity until the final cluster number is obtained.
In step S1, the sample set is X ═ X (X)1,x2,...,xm) Where m is the number of samples.
In step S2, the electricity consumption behavior labels of the electricity consumers include not only capacity, electricity consumption type, seasonal characteristics, temperature sensitivity, load stability, capacity utilization rate, holiday electricity characteristics, peak characteristics, and electricity consumption increase and decrease conditions.
In step S3, the electrical behavior clustering step includes:
s3.2: computing arbitrary samples x from the distance matrixiK number of neighboring samples Ni(k);
S3.3: calculating a sample xiOf neighboring samples Ni(k) And each sample x remainspOf neighboring samples Np(k) The cross-over-cross-over ratio of (c),
s3.4: clustering intersection ratio IOU;
wherein d isijIs xiAnd xjDistance between, Ni(k) Is a neighbor sample, Np(k) For each sample x remainingpThe IOU is the cross-over ratio.
In step S3.4, the clustering method of the intersection-to-parallel ratio IOU includes:
s3.4.1: when IOU is greater than threshold, then xiAnd its neighbor samples Ni (k) and xpAnd its neighbor samples np (k) are grouped into a cluster and these samples are removed from X and no longer participate in the calculation of IOU;
s3.4.2: when x ispE Ni (k) and IOU is less than the threshold, then xpAnd the samples in Np (k) that intersect with Ni (k) still belong to xiOne class, np (k) other samples are other classes and still participate in IOU calculations;
s3.4.3: when in useAnd IOU is less than the threshold, the samples in Np (k) intersected with Ni (k) still belong to xiClass I, xpAnd np (k) where the other samples are of other classes and still participate in the IOU computation.
the above description is not intended to limit the present invention, but rather, the present invention is intended to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the present invention.
Claims (6)
1. A power customer portrait method based on cross-comparison density clustering comprises the following steps:
s1, collecting a sample set, and carrying out missing value processing, consistency check, noise detection and normalization processing on the sample set by the sample;
s2: constructing a power consumption behavior tag of a power customer;
s3: assigning values to the labels completed in step S2 and then clustering the electricity usage behavior:
s4: calculating the mean value of all samples in each cluster according to the primary clustering result obtained in the step S3;
s5: and calculating the similarity among different clusters by using the average value obtained in the step S5 to obtain a cluster similarity matrix, and then merging the potential similar clusters according to the similarity until the final cluster number is obtained.
2. The method as claimed in claim 1, wherein in step S1, the sample set is X ═ X (X ═ X)1,x2,...,xm) Where m is the number of samples.
3. The method as claimed in claim 1, wherein in step S2, the electricity consumption behavior labels of the electricity consumers include more than capacity, electricity type, seasonal characteristics, temperature sensitivity, load stability, capacity utilization rate, holiday electricity characteristics, peak characteristics, electricity consumption increase and decrease.
4. The electric power customer imaging method based on cross-comparison density clustering according to claim 1, characterized in that: in step S3, the electrical behavior clustering step includes:
s3.2: computing arbitrary samples x from the distance matrixiK number of neighboring samples Ni(k);
S3.3: calculating a sample xiOf neighboring samples Ni(k) And each sample x remainspOf neighboring samples Np(k) The cross-over-cross-over ratio of (c),
s3.4: clustering the cross-over-parallel ratio IOU;
wherein d isijIs xiAnd xjDistance between, Ni(k) Is a neighbor sample, Np(k) For each sample x remainingpThe IOU is the cross-over ratio.
5. The method for clustering intersection-to-parallel ratios IOUs according to claim 4, comprising the steps of:
s3.4.1: when IOU is greater than threshold, then xiAnd its neighbor samples Ni (k) and xpAnd its neighbor samples np (k) are grouped into a cluster and these samples are removed from X and no longer participate in the calculation of IOU;
s3.4.2: when x ispE Ni (k) and IOU is less than the threshold, then xpAnd the samples in Np (k) that intersect with Ni (k) still belong to xiOne class, np (k) other samples are other classes and still participate in IOU calculations;
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111192403.7A CN113935410A (en) | 2021-10-13 | 2021-10-13 | Electric power customer portrait method based on cross-correlation density clustering |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111192403.7A CN113935410A (en) | 2021-10-13 | 2021-10-13 | Electric power customer portrait method based on cross-correlation density clustering |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113935410A true CN113935410A (en) | 2022-01-14 |
Family
ID=79278628
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111192403.7A Pending CN113935410A (en) | 2021-10-13 | 2021-10-13 | Electric power customer portrait method based on cross-correlation density clustering |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113935410A (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107358368A (en) * | 2017-07-21 | 2017-11-17 | 国网四川省电力公司眉山供电公司 | A kind of robust k means clustering methods towards power consumer subdivision |
CN111046787A (en) * | 2019-12-10 | 2020-04-21 | 华侨大学 | Pedestrian detection method based on improved YOLO v3 model |
CN111177208A (en) * | 2019-10-18 | 2020-05-19 | 姚长征 | Power consumption abnormity detection method based on big data analysis |
CN112241958A (en) * | 2019-07-18 | 2021-01-19 | 普天信息技术有限公司 | Target segmentation modeling method and device for three-dimensional point cloud data |
US20210056847A1 (en) * | 2019-08-19 | 2021-02-25 | Here Global B.V. | Method, apparatus, and computer program product for identifying street parking based on aerial imagery |
-
2021
- 2021-10-13 CN CN202111192403.7A patent/CN113935410A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107358368A (en) * | 2017-07-21 | 2017-11-17 | 国网四川省电力公司眉山供电公司 | A kind of robust k means clustering methods towards power consumer subdivision |
CN112241958A (en) * | 2019-07-18 | 2021-01-19 | 普天信息技术有限公司 | Target segmentation modeling method and device for three-dimensional point cloud data |
US20210056847A1 (en) * | 2019-08-19 | 2021-02-25 | Here Global B.V. | Method, apparatus, and computer program product for identifying street parking based on aerial imagery |
CN111177208A (en) * | 2019-10-18 | 2020-05-19 | 姚长征 | Power consumption abnormity detection method based on big data analysis |
CN111046787A (en) * | 2019-12-10 | 2020-04-21 | 华侨大学 | Pedestrian detection method based on improved YOLO v3 model |
Non-Patent Citations (1)
Title |
---|
吉训生;滕彬;: "基于深度神经网络的扶梯异常行为检测", 激光与光电子学进展, no. 06 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106570778B (en) | A kind of method that data integration based on big data is calculated with line loss analyzing | |
Flath et al. | Cluster analysis of smart metering data: An implementation in practice | |
CN111178624B (en) | New product demand prediction method | |
CN111144468A (en) | Power consumer information labeling method and device, electronic equipment and storage medium | |
Rathnayaka et al. | Formation of virtual community groups to manage prosumers in smart grids | |
CN108388955A (en) | Customer service strategies formulating method, device based on random forest and logistic regression | |
US20160132913A1 (en) | Multivariate Canonical Data Model for Tagging Customer Base of Energy Utility Enterprise | |
Voulis et al. | Understanding spatio-temporal electricity demand at different urban scales: A data-driven approach | |
Khan et al. | Segmentation of factories on electricity consumption behaviors using load profile data | |
Wang et al. | Smart Meter Data Analytics | |
CN110427418A (en) | A kind of customer analysis grouping method based on client's energy value index system | |
CN115130811A (en) | Method and device for establishing power user portrait and electronic equipment | |
CN110188255A (en) | Power consumer Behavior mining method and system based on the shared fusion of business datum | |
CN110334739A (en) | A kind of two-part component load forecasting method and device | |
CN113935410A (en) | Electric power customer portrait method based on cross-correlation density clustering | |
CN112163781A (en) | Park electricity utilization group life cycle evaluation method based on multi-dimensional index clustering | |
CN114372835B (en) | Comprehensive energy service potential customer identification method, system and computer equipment | |
CN103246987A (en) | Sales information management system and method | |
CN116226293A (en) | Method and system for generating and managing power customer portrait | |
CN111127186A (en) | Application method of customer credit rating evaluation system based on big data technology | |
CN115858126A (en) | Distributed computing power resource allocation method, device and medium | |
CN115907389A (en) | Urban area scale demand response potential power consumer positioning method, system and medium based on user portrait | |
CN110992087A (en) | Method and device for differentiating customer electricity consumption behavior areas | |
Sun et al. | Cloud-based data analysis of user side in smart grid | |
Wahyudi et al. | Prior and Future Research on Quality-Driven Product Service Systems: A Literature Review |
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 |