CN114970660A - Power load clustering method - Google Patents
Power load clustering method Download PDFInfo
- Publication number
- CN114970660A CN114970660A CN202210189375.1A CN202210189375A CN114970660A CN 114970660 A CN114970660 A CN 114970660A CN 202210189375 A CN202210189375 A CN 202210189375A CN 114970660 A CN114970660 A CN 114970660A
- Authority
- CN
- China
- Prior art keywords
- clustering
- vector
- cluster
- vectors
- power load
- 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 37
- 239000013598 vector Substances 0.000 claims abstract description 151
- 238000012545 processing Methods 0.000 claims abstract description 13
- 238000005259 measurement Methods 0.000 claims abstract description 12
- 239000011159 matrix material Substances 0.000 claims abstract description 10
- 238000007781 pre-processing Methods 0.000 claims abstract description 8
- 238000004422 calculation algorithm Methods 0.000 claims description 16
- 238000010606 normalization Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000000513 principal component analysis Methods 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 description 5
- 230000009191 jumping Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000003064 k means clustering Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 229910052731 fluorine Inorganic materials 0.000 description 1
- 125000001153 fluoro group Chemical group F* 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- 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
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Public Health (AREA)
- Probability & Statistics with Applications (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to a power load clustering method, which comprises the following steps: acquiring power load data to be clustered, and preprocessing the acquired power load data; carrying out initialization processing; calculating to obtain an initial clustering center vector matrix according to the preprocessed data set; distributing the rest load vectors to the cluster where the corresponding clustering center is located according to the improved similarity measurement standard to obtain a plurality of clustering clusters; calculating the intra-cluster vector average value of each cluster to be used as the current cluster center vector of the cluster; and judging whether the current clustering center vector meets the set clustering standard or not according to the current clustering center vector and the corresponding previous clustering center vector, if so, outputting the current clustering result and finishing clustering, otherwise, further judging whether the set maximum iteration frequency is reached, if so, outputting the current clustering result and finishing clustering, and otherwise, returning to continue to carry out iterative computation. Compared with the prior art, the method can effectively improve the clustering precision of the power load curves.
Description
Technical Field
The invention relates to the technical field of power load analysis, in particular to a power load clustering method.
Background
The power load cluster analysis is an important basis for work such as demand side management, load modeling, power system planning and the like, and has important significance for analysis, operation and planning of a power system. With the continuous improvement of the informatization degree of the power system, the power distribution side continuously produces and records mass data, and meanwhile, the marketing and distribution integration of the power company realizes the integration of various different business systems, so that a foundation is laid for more effectively mining load data. The load common characteristic can be extracted through the accurate power load clustering, the power utilization mode can be extracted through the load clustering analysis of a user layer, and the power utilization rule of a user can be deeply mastered; the load clustering analysis of the substation layer can reflect the running state of the system to a great extent, and can effectively solve the problems of time-varying property and regional dispersity of the load.
With the development of various intelligent algorithms, researchers at home and abroad have studied power load clustering algorithms, and currently, popular clustering algorithms include a K-means clustering algorithm, wavelet analysis, a fuzzy C-means clustering algorithm (FCM), an integrated clustering algorithm, a self-organizing feature mapping neural network (SOM), an Extreme Learning Machine (ELM), a cloud model and the like. However, the power load is data with characteristics of nonlinearity and time sequence, and the clustering method is difficult to accurately cluster the power load data and cannot ensure the clustering accuracy of the power load curve.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provide a power load clustering method, which improves the accuracy of power load curve clustering by improving a K-means clustering algorithm.
The purpose of the invention can be realized by the following technical scheme: a power load clustering method comprises the following steps:
s1, acquiring power load data to be clustered, and preprocessing the acquired power load data to obtain a preprocessed data set;
s2, initialization processing;
s3, calculating to obtain an initial clustering center vector matrix according to the preprocessed data set, wherein the initial clustering center vector matrix comprises a plurality of initial clustering center vectors;
s4, distributing the rest load vectors to the cluster where the corresponding clustering center is located according to the improved similarity measurement standard to obtain a plurality of clustering clusters;
s5, calculating the intra-cluster vector average value of each cluster to be used as the current cluster center vector of the cluster;
s6, judging whether the current clustering center vector and the corresponding previous clustering center vector meet the set clustering standard, if so, executing a step S8, otherwise, executing a step S7;
s7, further judging whether the set maximum iteration number is reached, if so, executing a step S8, otherwise, returning to the step S4;
and S8, outputting the current clustering result and ending the clustering process.
Further, the preprocessing process of the power load data obtained in step S1 includes denoising, dimensionality reduction, and normalization processing.
Further, the step S1 is specifically to perform denoising processing by using a DBSCAN algorithm; performing dimensionality reduction by using a KPCA (kernel principal component analysis) algorithm; and carrying out normalization processing by using a Max algorithm.
Further, the initialization processing in step S2 specifically includes setting the number of clusters, setting the maximum number of iterations, and setting the minimum distance between the cluster center vectors before and after updating.
Further, the step S3 specifically includes the following steps:
s31, if the set clustering number is k, selecting k vectors corresponding to k power load curves from the preprocessed data set;
s32, calculating the combined distance d between every two vectors in the k vectors c If there are n power load curves, they coexistCombining distances and then selecting a combined distance d c And taking k vectors corresponding to the maximum value as initial clustering center vectors to construct and obtain an initial clustering center vector matrix.
Further, the combined distance d between every two vectors in the step S32 c The method specifically comprises the following steps:
wherein, X i 、X j For two vectors, p (X), out of k vectors selected from the preprocessed data set i ,X j ) Is a vector X i And vector X j Pearson's correlation coefficient, x is Is the ith vector X i The s-th element, x js Is the jth vector X j N is the dimension of the vector, p (X) i ,X j ) A larger value indicates X i And X j The stronger the positive correlation between them.
Further, the step S4 specifically includes the following steps:
s41, calculating the distance between the other vectors and the current clustering center vector respectively, and screening out the minimum distance value from the distance calculation result;
s42, distributing the rest vectors to the cluster where the cluster center vector corresponding to the minimum distance value is located;
and S43, correspondingly distributing all the rest vectors to the cluster where the corresponding cluster center is located according to the processes of the step S41 and the step S42 to obtain a plurality of cluster clusters.
Further, the distances between the remaining vectors and the cluster center vector in step S41 are specifically:
where γ is the current iteration number, x is Is the ith vector X i The s-th element of (u) js As the jth cluster center vector U j Is the s-th dimension of (a), n is the vector dimension,for the gamma time of iteration X i And U j Pearson correlation coefficient therebetween.
Further, the specific process of step S6 is as follows:
and calculating the distance between the current clustering center vector and the corresponding previous clustering center vector, and judging whether the distance value meets the set clustering standard, if so, executing the step S8, otherwise, executing the step S7.
Further, the set clustering criterion is specifically that the distance between the current clustering center vector and the corresponding previous clustering center vector is smaller than the set minimum distance between the clustering center vectors before and after updating.
Compared with the prior art, the method has the advantages that the initial clustering center vector is selected according to the principle of the maximum combination distance, so that the power load curve with the maximum difference can be divided into different clusters at the beginning; and then, by combining with the improved similarity measurement standard, the rest power load curves can be further distributed to the most similar clusters, so that the similarity of the load curves in the clusters and the difference of the power load curves among the clusters can be increased in the power load curve clustering process, and the power load curve clustering precision can be effectively improved.
The invention considers the characteristics of nonlinearity and time sequence of the power load curve, improves the traditional Euclidean distance formula by combining with the Pearson correlation coefficient to be used as a new similarity measurement standard, distributes the rest load vectors to the cluster where the most similar clustering center is located, and can obtain a clustering result with higher precision and better quality through the updating and comparison of subsequent cluster cohesion center vectors.
Drawings
FIG. 1 is a schematic flow diagram of the process of the present invention;
fig. 2 is a schematic diagram of an application process of the embodiment.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments.
Examples
As shown in fig. 1, a method for clustering power loads includes the following steps:
s1, acquiring power load data to be clustered, and preprocessing the acquired power load data to obtain a preprocessed data set, wherein the preprocessing specifically includes denoising, dimensionality reduction and normalization processing on the acquired power load data, and in the embodiment, denoising is performed by using a DBSCAN algorithm; performing dimension reduction processing by using a KPCA algorithm; carrying out normalization processing by using a Max algorithm;
s2, initialization processing: setting a clustering number, setting a maximum iteration number and setting a minimum distance between clustering center vectors before and after updating;
s3, calculating according to the preprocessed data set to obtain an initial clustering center vector matrix, wherein the initial clustering center vector matrix comprises a plurality of initial clustering center vectors, and specifically:
s31, if the set clustering number is k, selecting k vectors corresponding to k power load curves from the preprocessed data set;
s32, calculating the combined distance d between every two vectors in the k vectors c If there are n power load curves, they coexistCombining distances and then selecting a combined distance d c K vectors corresponding to the maximum value are used as initial clustering center vectors to construct and obtain an initial clustering center vector matrix, wherein the combined distance d between every two vectors c Comprises the following steps:
in the formula, X i 、X j To go from a predetermined positionTwo vectors, rho (X), of the k vectors selected in the processed dataset i ,X j ) Is a vector X i And vector X j Pearson's correlation coefficient, x is Is the ith vector X i The s-th element, x js Is the jth vector X j N is the dimension of the vector, p (X) i ,X j ) A larger number indicates X i And X j The stronger the positive correlation between them;
s4, distributing the rest load vectors to the cluster where the corresponding cluster center is located according to the improved similarity measurement standard to obtain a plurality of cluster clusters, specifically:
s41, calculating the distance between the remaining vectors and the current cluster center vector:
wherein gamma is the current iteration number, x is Is the ith vector X i The s-th element of (u) js As the jth cluster center vector U j Is the s-th dimension of (a), n is the vector dimension,for the gamma time of iteration X i And U j Pearson correlation coefficient therebetween;
then screening out a value with the minimum distance from the distance calculation result;
s42, distributing the rest vectors to the cluster where the cluster center vector corresponding to the minimum distance value is located;
s43, correspondingly distributing all the rest vectors to the cluster where the corresponding clustering center is located according to the processes of the step S41 and the step S42 to obtain a plurality of clustering clusters;
s5, calculating the intra-cluster vector average value of each cluster to be used as the current cluster center vector of the cluster;
s6, judging whether the current clustering center vector and the corresponding previous clustering center vector meet the set clustering standard, if so, executing a step S8, otherwise, executing a step S7;
specifically, the distance between the current clustering center vector and the corresponding previous clustering center vector is calculated, and then whether the distance value meets the set clustering standard is judged, wherein the set clustering standard specifically is that the distance between the current clustering center vector and the corresponding previous clustering center vector is smaller than the set minimum distance between the clustering center vectors before and after updating;
s7, further judging whether the set maximum iteration number is reached, if so, executing a step S8, otherwise, returning to the step S4;
and S8, outputting the current clustering result and ending the clustering process.
The technical scheme aims at improving the distance formula and realizing the optimization of the initial clustering center: 1) the distance formula is improved by combining the Pearson correlation coefficient to serve as a new similarity measurement standard; 2) k curves are selected from the original power load curves as initial clustering center curves, and the principle is that the sum of distances (combined distance) between every two k curves is maximum.
The embodiment applies the above technical solution, as shown in fig. 2:
I. firstly, denoising, dimensionality reduction and normalization preprocessing are respectively carried out on original load data by using DBSCAN, KPCA and Max.
II. In the preprocessed data set, k vectors corresponding to k power load curves are selected, and the sum (combined distance) d of the distances between every two vectors is calculated c If there are n curves, they coexistCombining distances and then selecting a combined distance d c And taking the k vectors corresponding to the maximum as initial clustering center vectors.
And III, distributing the rest load vectors to the most similar clustering centers according to the improved similarity measurement standard.
IV, after the vectors are distributed, calculating the average value of the vectors in the cluster as a new centroid vector, judging whether the distance between the new centroid vector and the original centroid vector is smaller than a set value, and if so, jumping to the step VI. If not, jumping to the V step.
And V, judging whether the iteration times reach a preset value, if so, jumping to the step VI, and if not, jumping to the step III.
And VI, finishing the iterative process and outputting a clustering result.
When the initial clustering center vector is calculated, the distance formula is improved to serve as a similarity measurement standard, and the principle of the maximum combined distance is selected to optimize the initial clustering center, so that the power load curve clustering precision is improved. The following formula is utilized:
in the formula: x i ,X j For two vectors, p (X), out of k vectors selected from the original data set i ,X j ) Is a vector X i And vector X j The pearson correlation coefficient of (a) is expressed as follows:
in the formula: x is a radical of a fluorine atom is Represents the ith vector X i The s-th element, x js Represents the jth vector X j The s-th dimension element of (1), n represents the dimension of the vector; ρ (X) i ,X j ) The larger the value, the stronger its positive correlation. So when the positive correlation between two vectors is small, the corresponding distance becomes large, and this can allocate the vectors with small correlation to different clusters as much as possible.
After the centroid is determined, when other loads are distributed, the distance formula after improvement is used as a similarity measurement standard:
in the formula, ρ (X) i γ ,U j γ ) Is a vector X i γ And U j γ Pearson's correlation coefficient, X i γ 、U j γ Respectively the ith rest vector and the jth centroid vector in the gamma iteration, wherein n is the vector dimension, gamma represents the iteration number, and x is Represents the ith vector X i Element of dimension s, u js Representing the jth centroid vector U j The (d) th-dimensional element of (1),i.e. representing the pearson correlation coefficient between the ith remaining vector and the jth centroid vector at the γ -th iteration.
In summary, in the technical scheme, in order to increase the similarity of the load curves in the clusters and the difference of the power load curves between the clusters in the power load curve clustering process, the accuracy of the power load curve clustering can be improved. Therefore, an improved clustering method is provided: considering the characteristics of nonlinearity and time sequence of a power load curve, combining with a Pearson correlation coefficient to improve an Euclidean distance formula, and taking the Euclidean distance formula as a similarity measurement standard; and selecting an initial clustering center by a principle of solving the maximum combination distance. Therefore, the accuracy of power load curve clustering is improved.
By selecting an initial centroid vector according to the principle of the maximum combined distance, the power load curve with the maximum difference can be divided into different clusters at first; and then, a distance formula improved by combining with the Pearson correlation coefficient is used as a new similarity measurement standard, and other power load curves are further distributed to the most similar clusters, so that a clustering result with higher precision and better quality is obtained.
Claims (10)
1. A power load clustering method is characterized by comprising the following steps:
s1, acquiring power load data to be clustered, and preprocessing the acquired power load data to obtain a preprocessed data set;
s2, initialization processing;
s3, calculating to obtain an initial clustering center vector matrix according to the preprocessed data set, wherein the initial clustering center vector matrix comprises a plurality of initial clustering center vectors;
s4, distributing the rest load vectors to the cluster where the corresponding clustering center is located according to the improved similarity measurement standard to obtain a plurality of clustering clusters;
s5, calculating the intra-cluster vector average value of each cluster to be used as the current cluster center vector of the cluster;
s6, judging whether the current clustering center vector and the corresponding previous clustering center vector meet the set clustering standard, if so, executing a step S8, otherwise, executing a step S7;
s7, further judging whether the set maximum iteration number is reached, if so, executing a step S8, otherwise, returning to the step S4;
and S8, outputting the current clustering result and ending the clustering process.
2. The method for clustering power loads according to claim 1, wherein the preprocessing of the acquired power load data in step S1 includes denoising, dimensionality reduction and normalization.
3. The power load clustering method according to claim 2, wherein the step S1 is specifically to perform denoising processing by using a DBSCAN algorithm; performing dimensionality reduction by using a KPCA (kernel principal component analysis) algorithm; and carrying out normalization processing by using a Max algorithm.
4. The method according to claim 1, wherein the initialization process in step S2 specifically includes setting a cluster number, setting a maximum iteration number, and setting a minimum distance between cluster center vectors before and after updating.
5. The method for clustering power loads according to claim 4, wherein the step S3 specifically comprises the following steps:
s31, if the set clustering number is k, selecting k vectors corresponding to k power load curves from the preprocessed data set;
s32, calculating the combined distance d between every two vectors in the k vectors c If there are n power load curves, they coexistCombining distances and then selecting a combined distance d c And taking k vectors corresponding to the maximum value as initial clustering center vectors to construct and obtain an initial clustering center vector matrix.
6. The method for clustering power loads according to claim 5, wherein the combined distance d between every two vectors in the step S32 c The method specifically comprises the following steps:
wherein, X i 、X j For two vectors, p (X), out of k vectors selected from the preprocessed data set i ,X j ) Is a vector X i And vector X j Pearson's correlation coefficient, x is Is the ith vector X i The s-th dimension element of (1), x js Is the jth vector X j N is the dimension of the vector, p (X) i ,X j ) A larger number indicates X i And X j The stronger the positive correlation between them.
7. The method for clustering power loads according to claim 6, wherein the step S4 specifically comprises the following steps:
s41, calculating the distance between the other vectors and the current clustering center vector respectively, and screening out the minimum distance value from the distance calculation result;
s42, distributing the rest vectors to the cluster where the cluster center vector corresponding to the minimum distance value is located;
and S43, correspondingly distributing all the rest vectors to the cluster where the corresponding cluster center is located according to the processes of the step S41 and the step S42 to obtain a plurality of cluster clusters.
8. The method for clustering power loads according to claim 7, wherein the distances between the remaining vectors and the cluster center vector in the step S41 are specifically:
9. The method for clustering power loads according to claim 4, wherein the specific process of the step S6 is as follows:
and calculating the distance between the current clustering center vector and the corresponding previous clustering center vector, judging whether the distance value meets the set clustering standard, if so, executing the step S8, otherwise, executing the step S7.
10. The method according to claim 9, wherein the set clustering criterion is that a distance between a current cluster center vector and a corresponding previous cluster center vector is smaller than a set minimum distance between cluster center vectors before and after updating.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210189375.1A CN114970660A (en) | 2022-02-28 | 2022-02-28 | Power load clustering method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210189375.1A CN114970660A (en) | 2022-02-28 | 2022-02-28 | Power load clustering method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114970660A true CN114970660A (en) | 2022-08-30 |
Family
ID=82976228
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210189375.1A Pending CN114970660A (en) | 2022-02-28 | 2022-02-28 | Power load clustering method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114970660A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115310879A (en) * | 2022-10-11 | 2022-11-08 | 浙江浙石油综合能源销售有限公司 | Multi-fueling-station power consumption control method based on semi-supervised clustering algorithm |
-
2022
- 2022-02-28 CN CN202210189375.1A patent/CN114970660A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115310879A (en) * | 2022-10-11 | 2022-11-08 | 浙江浙石油综合能源销售有限公司 | Multi-fueling-station power consumption control method based on semi-supervised clustering algorithm |
CN115310879B (en) * | 2022-10-11 | 2022-12-16 | 浙江浙石油综合能源销售有限公司 | Multi-fueling-station power consumption control method based on semi-supervised clustering algorithm |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111199016B (en) | Daily load curve clustering method for improving K-means based on DTW | |
CN106600059B (en) | Intelligent power grid short-term load prediction method based on improved RBF neural network | |
CN104951425A (en) | Cloud service performance adaptive action type selection method based on deep learning | |
CN106067034B (en) | Power distribution network load curve clustering method based on high-dimensional matrix characteristic root | |
CN108805213B (en) | Power load curve double-layer spectral clustering method considering wavelet entropy dimensionality reduction | |
CN111507470A (en) | Abnormal account identification method and device | |
CN108428055B (en) | Load clustering method considering load longitudinal characteristics | |
CN110796159A (en) | Power data classification method and system based on k-means algorithm | |
CN116187835A (en) | Data-driven-based method and system for estimating theoretical line loss interval of transformer area | |
CN114970660A (en) | Power load clustering method | |
CN115879602A (en) | Ultra-short-term photovoltaic output prediction method based on transient weather | |
CN108647714A (en) | Acquisition methods, terminal device and the medium of negative label weight | |
CN117035837B (en) | Method for predicting electricity purchasing demand of power consumer and customizing retail contract | |
CN107093005A (en) | The method that tax handling service hall's automatic classification is realized based on big data mining algorithm | |
CN112149052A (en) | Daily load curve clustering method based on PLR-DTW | |
CN111126467B (en) | Remote sensing image space spectrum clustering method based on multi-target sine and cosine algorithm | |
CN112561179A (en) | Stock tendency prediction method and device, computer equipment and storage medium | |
CN117113126A (en) | Industry electricity utilization characteristic analysis method based on improved clustering algorithm | |
CN111798152A (en) | Intelligent store management method and device | |
CN111353525A (en) | Modeling and missing value filling method for unbalanced incomplete data set | |
Chen et al. | A GPU-accelerated approximate algorithm for incremental learning of Gaussian mixture model | |
CN114091904A (en) | Enterprise migration park recruitment analysis method based on artificial intelligence algorithm | |
CN112613562A (en) | Data analysis system and method based on multi-center cloud computing | |
CN111784381A (en) | Privacy protection and SOM network-based power customer segmentation method and system | |
CN111310842A (en) | Density self-adaptive rapid clustering method |
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 |