CN103646354A - Effective index FCM and RBF neural network-based substation load characteristic categorization method - Google Patents
Effective index FCM and RBF neural network-based substation load characteristic categorization method Download PDFInfo
- Publication number
- CN103646354A CN103646354A CN201310612449.9A CN201310612449A CN103646354A CN 103646354 A CN103646354 A CN 103646354A CN 201310612449 A CN201310612449 A CN 201310612449A CN 103646354 A CN103646354 A CN 103646354A
- Authority
- CN
- China
- Prior art keywords
- classification
- neural network
- load
- substation
- 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
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses an effective index FCM and RBF neural network-based substation load characteristic categorization method. The method comprises the following steps that: load constituent ratios of a substation are adopted as characteristic vectors of load characteristic categorization of the substation; clustering analysis is performed on data samples of the load constituent ratios of the substation through using a fuzzy clustering analysis method so as to obtain data categorization results under different numbers of clusters, and an optimal number of clusters is determined through three kinds of clustering effect evaluation indexes, and a fuzzy subordination degree matrix and the clustering center of each category of under the optimal number of clusters are obtained; one group of samples are selected in each clustering category according to a principle of minimum distance, and category numbers corresponding to each group of samples are set, such that a training sample set is formed; a substation load characteristic secondary categorization model is established through adopting an RBF neural network, and the formed training sample set is utilized to train the neural network, and the trained neural network is further utilized to realize the load characteristic categorization of the substation. The effective index FCM and RBF neural network-based substation load characteristic categorization method of the invention has the advantages of simple operation and high accuracy.
Description
Technical field
The present invention relates to load modeling technical field, be specifically related to the substation load characteristics classification method based on efficiency index FCM and RBF neural network.
Background technology
Along with improving constantly of socioeconomic development and scientific and technological level, the scale of electrical network constantly expands, NETWORK STRUCTURE PRESERVING POWER SYSTEM is day by day complicated, power system security, stable, reliable operation have been proposed to more and more higher requirement, therefore set up and accurately reflect that the Real-time Load model of whole network load seems very important.
Because electric load spatially shows the dispersiveness of region, show in time random time variation, so in order accurately to reflect part throttle characteristics, need to set up the integrated load model of large amount of complex.But if the integrated load model number that same electrical network adopts is too huge, form is too complicated, at engineering field, is just difficult to have practical value.Therefore the load classification of transformer station is that load model moves towards one of practical important means, has realized the reasonable compromise of accuracy and the practicality of model.Load classification based on transformer station, the aspects such as safety and reliability of being optimized, formulated dispatching of power netwoks plan, operation planning reliability assessment, raising equipment the later stage of transformer station are significant.
At present, the conventional method of the classification of transformer station's part throttle characteristics has based on Statistics Method, Grey Correlation Cluster method, fuzzy C-means clustering and neural network etc.By fuzzy clustering, the degree of uncertainty that sample belongs to each classification can be obtained, more real world can be objectively responded; Neural network can approach any Nonlinear Mapping with arbitrary accuracy, can be for describing the classification problem of transformer station's part throttle characteristics.
Summary of the invention
The object of the invention is to: a kind of substation load characteristics classification method based on efficiency index FCM and RBF neural network is provided, fuzzy clustering method and neural net method are combined, the classification of realization to transformer station's part throttle characteristics, effectively improve the accuracy of load modeling, improve the accuracy rate of substation load characteristics classification, guarantee the operation of power system safety and stability, improve the safety and reliability of equipment.
Technical solution of the present invention is that this substation load characteristics classification method based on efficiency index FCM and RBF neural network comprises the steps:
1) choose the proper vector of the load classification of transformer station, with the load structure of transformer station, be compared to the proper vector of substation load characteristics classification;
2) proper vector of substation load characteristics classification is carried out to a cluster analysis, utilize method of fuzzy cluster analysis to try to achieve the Data classification result under different clusters number, it is a class that the sample with similar part throttle characteristics is gathered;
3) according to three kinds of Cluster Assessment target functions, try to achieve the division factor V of all Data classifications
pc, divide closely related V
peand Xie-Beni Validity Index V
xb, the numerical value of three kinds of validity check indexs under comprehensive more different clusters number, determines best clusters number, obtains fuzzy membership matrix and all kinds of cluster centre under best clusters number;
4) according to the cluster result under best clusters number and inter-object distance minimum principle, calculate the distance of the cluster centre of all samples in each cluster classification and this classification, choose all kinds of in cluster centre apart from a minimum h sample as one group of sample, by setting the classification number corresponding with each group sample, form training sample set;
5) adopt RBF neural network transformer station part throttle characteristics secondary classification model, described in the training sample set pair that utilization forms, neural network is trained, the load structure ratio that input data are transformer station, output data are the classification number under this sample, the neural network that recycling trains is classified than sample to all transformer stations load structure, realizes the load classification to transformer station.
The present invention compared with prior art, has the following advantages: the method adopts 3 kinds of Cluster Assessment target functions to obtain best cluster number, obtains the cluster result under best clusters number; Respectively in each class according to the training sample of selecting one group of sample as RBF neural network apart from minimum principle, by setting the classification number corresponding with each group sample, formation training sample set; Adopt the secondary classification model of RBF neural network to transformer station's part throttle characteristics, utilize the training sample set forming to realize the training to RBF neural network, by the neural network training, realize all transformer stations load structure is classified than sample, the science and rationality and the accuracy that have effectively improved the load classification of transformer station, significantly improved economic benefit and social benefit.
Accompanying drawing explanation
Fig. 1 is FB(flow block) of the present invention.
Embodiment
Embodiment: classify according to following steps
1) choose the proper vector of the load classification of transformer station, because the composition of load structure is the essential characteristic of transformer station's load, load structure composition be not both the basic reason that causes synthetic load property difference, so the load structure of choosing transformer station is compared to the proper vector of substation load characteristics classification, embody science and the rationality of substation load characteristics classification;
2) proper vector of substation load characteristics classification is carried out to a cluster analysis, utilize method of fuzzy cluster analysis to try to achieve the Data classification result under different clusters number, it is a class that the sample with similar part throttle characteristics is gathered;
3) according to three kinds of Cluster Assessment target functions, try to achieve the division factor V of all Data classifications
pc, divide closely related V
peand Xie-Beni Validity Index V
xb, the numerical value of three kinds of validity check indexs under comprehensive more different clusters number, determines best clusters number c, obtains fuzzy membership matrix and all kinds of cluster centre under best clusters number;
(a) division factor V
pcby Bezdek, proposed, this is the function of the 1st tolerance fuzzy clustering validity, and its expression is:
Wherein: n is number of samples, c is clusters number,
for degree of membership matrix
in element,
represent j sample
degree of membership in i class, U is degree of membership matrix; Division factor is a standard of weighing cluster result fog-level; Overstepping the bounds of propriety V when bright of cluster result
pcvalue just larger, V when cluster result is fuzzyyer
pcvalue just more close to 1/c, i.e. V
pccluster result corresponding to maximum desired value;
(b) divide closely related V
pebe Bezdek according to the closely related formula of Shannon information, the division of the fuzzy division of proposition is closely related, its expression is:
Wherein: n is number of samples, c is clusters number,
degree of membership matrix, and agreement
time, have
.V
pebe worth littlely, cluster result is overstepping the bounds of propriety bright, and minimum desired value correspondence best cluster numbers;
(c) Xie-Beni refers to that Xie X .L and Beni G. A two people are according to the geometry of data set, the Validity Index proposing in 1991; Validity Index V
xbcan weigh compactness in class and the degree of separation between class, find an equilibrium point in class between compactness and class between degree of separation, its value is less, and the cluster result of acquisition is better; Its expression is:
(3)
Wherein: n is number of samples, c is clusters number,
for sample j,
be degree of membership matrix, m is smoothing parameter, represents the blur level of degree of membership matrix U, and the cluster centre that V is every class, in above formula
be used for weighing the compactness in class, be worth littlely, data similarity is compacter more greatly in class;
be used for weighing the separation degree between class and class, the dissimilarity between larger class and class is larger, and between class, degree of separation is better;
4) according to the cluster result under best clusters number c and inter-object distance minimum principle, select the sample at the most close Mei Lei center as the training sample of the cluster of RBF neural network, wherein RBF refers to radial basis function; Calculate all samples in each classification to the Distance matrix D IST of the cluster centre of this classification
i, i=1 wherein, 2 ..., c; From Distance matrix D IST
ithe h of a middle chosen distance minimum sample, as one group of sample, is i by setting its corresponding classification number, forms training sample set;
5) adopt the secondary classification model of RBF neural network transformer station part throttle characteristics, utilize neural network described in the training sample set pair forming to train, input data are
the load structure of individual transformer station is than sample, and output data are the classification number under this sample, and the neural network that recycling trains is classified than sample to all transformer stations load structure, realizes the load classification to transformer station.
In sum, the method of a kind of substation load characteristics classification based on efficiency index FCM and RBF neural network provided by the invention, by method of fuzzy cluster analysis, by sample data cluster, be different numbers, according to three kinds of Cluster Validity Indexes, select best clusters number, obtain the cluster result under best clusters number, respectively in each class according to the training sample of selecting one group of sample as RBF neural network apart from minimum principle, by setting the classification number corresponding with each group sample, the training of realization to RBF neural network, by the neural network training, realize all transformer stations load structure is classified than sample, effectively improved the accuracy of the load classification of transformer station, economic benefit and social benefit have been significantly improved.
It should be noted that above embodiment is only used for technical scheme of the present invention is described, be not used in restriction the present invention, protection scope of the present invention is defined by the claims; By embodiment, to detailed explanation of the present invention, those skilled in the art can modify or be equal to replacement this, this modification or be equal to replacement and should think within the scope of the claims in the present invention.
Claims (1)
1. the substation load characteristics classification method based on efficiency index FCM and RBF neural network, is characterized in that the method comprises the following steps:
1) choose the proper vector of the load classification of transformer station, with the load structure of transformer station, be compared to the proper vector of substation load characteristics classification;
2) proper vector of substation load characteristics classification is carried out to a cluster analysis, utilize method of fuzzy cluster analysis to try to achieve the Data classification result under different clusters number, it is a class that the sample with similar part throttle characteristics is gathered;
3) according to three kinds of Cluster Assessment target functions, try to achieve the division factor V of all Data classifications
pc, divide closely related V
peand Xie-Beni Validity Index V
xb, the numerical value of three kinds of Validity Indexes under comprehensive more different clusters number, determines best clusters number, obtains fuzzy membership matrix and all kinds of cluster centre under best clusters number;
4) according to the cluster result under best clusters number and inter-object distance minimum principle, calculate the distance of the cluster centre of all samples in each cluster classification and this classification, choose all kinds of in cluster centre apart from a minimum h sample as one group of sample, by setting the classification number corresponding with each group sample, form training sample set;
5) adopt RBF neural network transformer station part throttle characteristics secondary classification model, described in the training sample set pair that utilization forms, neural network is trained, the load structure ratio that input data are transformer station, output data are the classification number under this sample, the neural network that recycling trains is classified than sample to all transformer stations load structure, realizes the load classification to transformer station.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310612449.9A CN103646354A (en) | 2013-11-28 | 2013-11-28 | Effective index FCM and RBF neural network-based substation load characteristic categorization method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310612449.9A CN103646354A (en) | 2013-11-28 | 2013-11-28 | Effective index FCM and RBF neural network-based substation load characteristic categorization method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103646354A true CN103646354A (en) | 2014-03-19 |
Family
ID=50251564
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310612449.9A Pending CN103646354A (en) | 2013-11-28 | 2013-11-28 | Effective index FCM and RBF neural network-based substation load characteristic categorization method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103646354A (en) |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104318489A (en) * | 2014-10-21 | 2015-01-28 | 广东电网有限责任公司电力科学研究院 | Transformer grouping method based on load characteristic analysis |
CN104809255A (en) * | 2015-05-21 | 2015-07-29 | 国家电网公司 | Load shape acquisition method and system |
CN105139025A (en) * | 2015-07-30 | 2015-12-09 | 东南大学 | Non-linear analysis method based online intelligent identification method for flow pattern of gas-solid fluidized bed |
CN105678766A (en) * | 2016-01-06 | 2016-06-15 | 福州大学 | Fuzzy c-means image segmentation method based on local neighborhood and global information |
CN106599138A (en) * | 2016-12-05 | 2017-04-26 | 南京信息工程大学 | Variety identification method for electrical appliances |
CN106600059A (en) * | 2016-12-13 | 2017-04-26 | 北京邮电大学 | Intelligent power grid short-term load predication method based on improved RBF neural network |
CN107977771A (en) * | 2017-11-07 | 2018-05-01 | 国家电网公司 | A kind of substation's characteristic analysis method based on multivariate clustering model Yu two-phase analyzing method correction algorithm |
CN109344919A (en) * | 2018-12-09 | 2019-02-15 | 国网江苏省电力有限公司扬州供电分公司 | A kind of construction method for the interruptible load characteristic monitoring model adapting to net load interaction |
CN109426813A (en) * | 2018-11-02 | 2019-03-05 | 中电科新型智慧城市研究院有限公司 | The customized interest point extraction method of remote sensing images based on fuzzy clustering and neural network model |
CN109492048A (en) * | 2019-01-21 | 2019-03-19 | 国网河北省电力有限公司经济技术研究院 | A kind of extracting method, system and the terminal device of power consumer electrical characteristics |
CN109670536A (en) * | 2018-11-30 | 2019-04-23 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | A kind of local discharge signal clustering method in the case of multi-source electric discharge and interference superposition |
CN109871997A (en) * | 2019-02-14 | 2019-06-11 | 国网宁夏电力有限公司经济技术研究院 | A kind of temperature-lowering load combination calculating system adapting to complex environment |
CN110322369A (en) * | 2019-07-03 | 2019-10-11 | 厦门理工学院 | A kind of building load optimum combination determines method, terminal device and storage medium |
CN111028004A (en) * | 2019-11-28 | 2020-04-17 | 国网吉林省电力有限公司 | Market assessment analysis method based on big data technology |
CN111242217A (en) * | 2020-01-13 | 2020-06-05 | 支付宝实验室(新加坡)有限公司 | Training method and device of image recognition model, electronic equipment and storage medium |
CN111695623A (en) * | 2020-06-09 | 2020-09-22 | 中国电力科学研究院有限公司 | Large-scale battery energy storage system group modeling method, system and equipment based on fuzzy clustering and readable storage medium |
CN111737924A (en) * | 2020-08-17 | 2020-10-02 | 国网江西省电力有限公司电力科学研究院 | Method for selecting typical load characteristic transformer substation based on multi-source data |
CN112308299A (en) * | 2020-10-19 | 2021-02-02 | 新奥数能科技有限公司 | Sample data extraction method and device for power system load prediction model |
CN112541528A (en) * | 2020-12-02 | 2021-03-23 | 国家电网有限公司 | Power transmission and transformation project cost prediction index optimization method based on fuzzy clustering |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR100963188B1 (en) * | 2008-05-16 | 2010-06-14 | 삼성물산 주식회사 | System for measuring the distributed electric power in home |
CN103390200A (en) * | 2013-07-18 | 2013-11-13 | 国家电网公司 | Photovoltaic power station electricity generation output power forecasting method based on similar days |
-
2013
- 2013-11-28 CN CN201310612449.9A patent/CN103646354A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR100963188B1 (en) * | 2008-05-16 | 2010-06-14 | 삼성물산 주식회사 | System for measuring the distributed electric power in home |
CN103390200A (en) * | 2013-07-18 | 2013-11-13 | 国家电网公司 | Photovoltaic power station electricity generation output power forecasting method based on similar days |
Non-Patent Citations (1)
Title |
---|
姜勇: "基于模糊聚类的神经网络短期负荷预测方法", 《电网技术》 * |
Cited By (27)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104318489A (en) * | 2014-10-21 | 2015-01-28 | 广东电网有限责任公司电力科学研究院 | Transformer grouping method based on load characteristic analysis |
CN104809255A (en) * | 2015-05-21 | 2015-07-29 | 国家电网公司 | Load shape acquisition method and system |
CN105139025A (en) * | 2015-07-30 | 2015-12-09 | 东南大学 | Non-linear analysis method based online intelligent identification method for flow pattern of gas-solid fluidized bed |
CN105678766B (en) * | 2016-01-06 | 2019-01-22 | 福州大学 | A kind of fuzzy c-means image partition method based on local neighborhood and global information |
CN105678766A (en) * | 2016-01-06 | 2016-06-15 | 福州大学 | Fuzzy c-means image segmentation method based on local neighborhood and global information |
CN106599138A (en) * | 2016-12-05 | 2017-04-26 | 南京信息工程大学 | Variety identification method for electrical appliances |
CN106600059A (en) * | 2016-12-13 | 2017-04-26 | 北京邮电大学 | Intelligent power grid short-term load predication method based on improved RBF neural network |
CN106600059B (en) * | 2016-12-13 | 2020-07-24 | 北京邮电大学 | Intelligent power grid short-term load prediction method based on improved RBF neural network |
CN107977771A (en) * | 2017-11-07 | 2018-05-01 | 国家电网公司 | A kind of substation's characteristic analysis method based on multivariate clustering model Yu two-phase analyzing method correction algorithm |
CN107977771B (en) * | 2017-11-07 | 2021-05-07 | 国家电网公司 | Transformer substation characteristic analysis method based on multi-element clustering model and two-stage clustering correction algorithm |
CN109426813A (en) * | 2018-11-02 | 2019-03-05 | 中电科新型智慧城市研究院有限公司 | The customized interest point extraction method of remote sensing images based on fuzzy clustering and neural network model |
CN109426813B (en) * | 2018-11-02 | 2022-06-24 | 中电科新型智慧城市研究院有限公司 | Remote sensing image user-defined interest point extraction method based on fuzzy clustering and neural network model |
CN109670536A (en) * | 2018-11-30 | 2019-04-23 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | A kind of local discharge signal clustering method in the case of multi-source electric discharge and interference superposition |
CN109670536B (en) * | 2018-11-30 | 2023-05-30 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | Partial discharge signal clustering method under multi-source discharge and interference superposition condition |
CN109344919A (en) * | 2018-12-09 | 2019-02-15 | 国网江苏省电力有限公司扬州供电分公司 | A kind of construction method for the interruptible load characteristic monitoring model adapting to net load interaction |
CN109492048A (en) * | 2019-01-21 | 2019-03-19 | 国网河北省电力有限公司经济技术研究院 | A kind of extracting method, system and the terminal device of power consumer electrical characteristics |
CN109871997A (en) * | 2019-02-14 | 2019-06-11 | 国网宁夏电力有限公司经济技术研究院 | A kind of temperature-lowering load combination calculating system adapting to complex environment |
CN110322369B (en) * | 2019-07-03 | 2021-10-15 | 厦门理工学院 | Building load optimal combination determination method, terminal device and storage medium |
CN110322369A (en) * | 2019-07-03 | 2019-10-11 | 厦门理工学院 | A kind of building load optimum combination determines method, terminal device and storage medium |
CN111028004A (en) * | 2019-11-28 | 2020-04-17 | 国网吉林省电力有限公司 | Market assessment analysis method based on big data technology |
CN111242217A (en) * | 2020-01-13 | 2020-06-05 | 支付宝实验室(新加坡)有限公司 | Training method and device of image recognition model, electronic equipment and storage medium |
CN111695623A (en) * | 2020-06-09 | 2020-09-22 | 中国电力科学研究院有限公司 | Large-scale battery energy storage system group modeling method, system and equipment based on fuzzy clustering and readable storage medium |
CN111695623B (en) * | 2020-06-09 | 2024-05-10 | 中国电力科学研究院有限公司 | Group modeling method, system, equipment and readable storage medium for large-scale battery energy storage system based on fuzzy clustering |
CN111737924A (en) * | 2020-08-17 | 2020-10-02 | 国网江西省电力有限公司电力科学研究院 | Method for selecting typical load characteristic transformer substation based on multi-source data |
CN112308299A (en) * | 2020-10-19 | 2021-02-02 | 新奥数能科技有限公司 | Sample data extraction method and device for power system load prediction model |
CN112308299B (en) * | 2020-10-19 | 2024-04-19 | 新奥数能科技有限公司 | Sample data extraction method and device for power system load prediction model |
CN112541528A (en) * | 2020-12-02 | 2021-03-23 | 国家电网有限公司 | Power transmission and transformation project cost prediction index optimization method based on fuzzy clustering |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103646354A (en) | Effective index FCM and RBF neural network-based substation load characteristic categorization method | |
Ardakanian et al. | On identification of distribution grids | |
Ye et al. | A data-driven bottom-up approach for spatial and temporal electric load forecasting | |
CN103390200A (en) | Photovoltaic power station electricity generation output power forecasting method based on similar days | |
CN105163326B (en) | A kind of cell clustering method and system based on wireless network traffic feature | |
CN104573947A (en) | Comprehensive evaluation method for low-voltage transformer areas of regional intelligent distribution network | |
CN110059356A (en) | A kind of bulk power grid intelligent control system and method based on big data and artificial intelligence | |
CN104376371B (en) | A kind of distribution based on topology is layered load forecasting method | |
CN104408667A (en) | Comprehensive assessment method and system of power quality | |
CN101976842A (en) | Method for automatically acquiring key sections based on electrical partitioning | |
Dong et al. | A short-term power load forecasting method based on k-means and SVM | |
CN106503851A (en) | A kind of improved Short-Term Load Forecasting Method based on wavelet analysises | |
Wang et al. | Actuator placement for enhanced grid dynamic performance: A machine learning approach | |
CN104933505A (en) | Decision and evaluation method for intelligent power distribution network group based on fuzzy assessment | |
CN105528671B (en) | A kind of power grid Multidimensional Awareness based on big data and safety estimation system and method | |
CN107301604A (en) | Multi-model fusion estimation system | |
CN103762579A (en) | Direct-current power emergency control performance index calculation method for improving transient power angle stability of power grid | |
CN108959958A (en) | A kind of method for secret protection and system being associated with big data | |
Chan et al. | Multiple classifier system for short term load forecast of microgrid | |
CN110059913A (en) | A kind of quantitative estimation method counted and the power failure of future-state is planned | |
CN109034500A (en) | A kind of mid-term electric load forecasting method of multiple timings collaboration | |
CN107966600A (en) | A kind of electricity anti-theft system and its electricity anti-theft method based on deep learning algorithm | |
CN105303268A (en) | Wind power generation output power prediction method based on similarity theory | |
Miraftabzadeh et al. | K-means and alternative clustering methods in modern power systems | |
CN110490220A (en) | A kind of bus load discrimination method and system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140319 |