CN103544539A - Method for predicting variables of users on basis of artificial neural networks and D-S (Dempster-Shafer) evidence theory - Google Patents
Method for predicting variables of users on basis of artificial neural networks and D-S (Dempster-Shafer) evidence theory Download PDFInfo
- Publication number
- CN103544539A CN103544539A CN201310474134.2A CN201310474134A CN103544539A CN 103544539 A CN103544539 A CN 103544539A CN 201310474134 A CN201310474134 A CN 201310474134A CN 103544539 A CN103544539 A CN 103544539A
- Authority
- CN
- China
- Prior art keywords
- data
- neural network
- neural networks
- users
- acquiring
- 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
Abstract
The invention provides a method for predicting variables of users on the basis of artificial neural networks and a D-S (Dempster-Shafer) evidence theory. The method includes steps of 1), acquiring data such as power consumption conditions and power utilization habit of the users; 2), acquiring original sample data of the users, randomly and equally dividing the original sample data into n equal portions and preprocessing the acquired data; 3), initializing n RBF (radial basis function) neural networks, setting network parameters and performing offline learning training on the networks; 4), establishing feature-level fusion models on the basis of the neural networks after training on the RBF neural networks is completed; 5), acquiring detection samples, preprocessing data and inputting the preprocessed data into each neural network; 6), acquiring n groups of results outputted by the neural networks, and normalizing each group of results to generate a belief function of a group of evidences; 7), fusing the evidences by the aid of the D-S evidence theory according to combination rules; 8), acquiring a final belief function and predicting the variables of the users according to judgment rules.
Description
Technical field
The present invention relates to a kind of user's variable quantity Forecasting Methodology based on artificial neural network and D-S evidence theory.
Background technology
Along with the development of electricity market, user improves constantly the demand of electric power, and the demand of electrical power services is also improved constantly.Power consumer enormous amount, electric power need to be from user's electricity consumption situation and characteristic, Accurate Prediction user variable quantity, thus formulate meticulous service program and scheme.For this reason, the present invention proposes a kind of power consumer variable quantity Forecasting Methodology based on artificial neural network and D-S evidence theory.
Summary of the invention
A kind of user's variable quantity Forecasting Methodology based on artificial neural network and D-S evidence theory of the present invention, comprises the steps:
1) gather the data such as user power utilization amount situation and consumption habit;
2) obtain user's raw sample data, be divided at random n decile, the data that gather are carried out to pre-service;
3) initialization n RBF neural network, arranges network parameter, carries out internet off-line learning training;
4) RBF neural metwork training complete after the feature level Fusion Model of foundation based on neural network;
5) to detect sample, after data pre-service as the input of each neural network;
6) obtain n group neural network Output rusults, to every group of result normalization, as the belief function of one group of evidence;
7), according to composition rule, utilize D-S evidence theory to merge these evidences; And
8) obtain final belief function, according to judgment rule, make the prediction of user's variable quantity.
According to the method, predict power consumer variable quantity, obtained good Social benefit and economic benefit.
Accompanying drawing explanation
Fig. 1 is the schematic diagram of the power consumer variable quantity Forecasting Methodology based on data fusion;
Fig. 2 is the structure of RBF neural network;
Fig. 3 is the schematic diagram of the blending decision method based on neural network and D-S evidence theory.
Fig. 4 is the user's changed measurement test result according to an embodiment.
Embodiment
Data fusion is a kind of automated information comprehensive treatment technique, utilizes the complementarity of multi-source data and the information that Intelligent Computation Technology is comprehensively analyzed data, thereby the characteristic of description object is more accurately made and identified accurately and judge object.In the present invention, adopt data fusion method to study power consumer variable quantity, implementation process as shown in Figure 1: first, using user power utilization custom and situation as inputting data, and carry out the pre-service of data; Subsequently, from mass data, carry out the extraction of electricity consumption user characteristics and normalization; Finally adopt a kind of Forecasting Methodology based on neural network and D-S evidence theory, dope power consumer variable quantity.
The RBF neural network basic structure that the present invention adopts as shown in Figure 2, is used RBF function as the activation function of hidden unit, forms hidden layer space, and between input vector and hidden layer, weights are 1, and hidden layer adopts learning algorithm adjustment to the weights between output layer.Whole RBF network has been realized the Nonlinear Mapping that is input to output, can be applied to the analysis of different nonlinear relationships, and network output is simultaneously again linear relationship with being connected between weights, and the training speed of network is fast like this.
D-S evidence theory is the probabilistic common method of a kind of processing, by setting up the corresponding relation between proposition and set, the size of identification uncertainty is described, by different evidences are upgraded to belief function with D-S composition rule with basic probability assignment function or belief function.
In the present invention, adopt a kind of blending decision method, as shown in Figure 3.Adopt neural network to carry out feature level fusion, adopt D-S evidence to carry out decision level fusion, first by a plurality of neural networks, user power utilization amount situation and consumption habit are done after rough handling, each network Output rusults normalization is afterwards as an evidence, D-S evidence theory merges these evidences, according to making electricity consumption customer volume Forecasting Methodology after the synthetic processing of decision rule.
Like this, just, set up a kind of power consumer amount Forecasting Methodology based on neural network and D-S evidence theory fusion.Its concrete steps comprise: gather the data such as user power utilization amount situation and consumption habit; Obtain user's raw sample data, be divided at random n decile, data are carried out to pre-service; The initialization of n RBF neural network, arranges network parameter, internet off-line learning training; RBF neural metwork training completes, and sets up the feature level Fusion Model based on neural network; To detect sample, after data pre-service as the input of each neural network; Obtain n group neural network Output rusults, every group of result normalization, as the belief function of one group of evidence; According to composition rule, utilize D-S evidence theory to merge these evidences; Obtain final belief function, according to judgment rule, make the prediction of user's variable quantity.
According to an embodiment, in order to verify the validity of customer volume Forecasting Methodology of the present invention, chosen the data of a certain power supply section here and carried out training and testing.First, randomly draw 300 days customer volume data of power supply section as the training sample of neural network, subsequently, chosen in addition the validity that 30 days power consumption data are tested the method.
The data of emulation testing are as follows:
Measured data value sequence=[282 303 316 308 289 297 301 322 289 317 309 299 292 301 304 290 309 294 301 309 296 298 312 298 320 327 317 302 312 293].
Predicted data value sequence=[288 309 321 304 290 290 300 319 284 312 308 293 293 301 307 293 311 287 295 306 296 300 311 303 323 334 317 299 306 295].
As shown in Figure 4, result shows the result of test, and the method predicated error is less than 3%, has good accuracy.
Although described the present invention about specific embodiment, the present invention is intended to be limited.Those skilled in the technology concerned are easy to modify, improve and are out of shape, and are intended to all these modifications, improvement and distortion to be included in the scope of claim of the present invention.
Claims (1)
1. the user's variable quantity Forecasting Methodology based on artificial neural network and D-S evidence theory, comprises the steps:
1) gather the data such as user power utilization amount situation and consumption habit;
2) obtain user's raw sample data, be divided at random n decile, the data that gather are carried out to pre-service;
3) initialization n RBF neural network, arranges network parameter, carries out internet off-line learning training;
4) RBF neural metwork training complete after the feature level Fusion Model of foundation based on neural network;
5) to detect sample, after data pre-service as the input of each neural network;
6) obtain n group neural network Output rusults, to every group of result normalization, as the belief function of one group of evidence;
7), according to composition rule, utilize D-S evidence theory to merge these evidences; And
8) obtain final belief function, according to judgment rule, make the prediction of user's variable quantity.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310474134.2A CN103544539A (en) | 2013-10-12 | 2013-10-12 | Method for predicting variables of users on basis of artificial neural networks and D-S (Dempster-Shafer) evidence theory |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310474134.2A CN103544539A (en) | 2013-10-12 | 2013-10-12 | Method for predicting variables of users on basis of artificial neural networks and D-S (Dempster-Shafer) evidence theory |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103544539A true CN103544539A (en) | 2014-01-29 |
Family
ID=49967973
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310474134.2A Pending CN103544539A (en) | 2013-10-12 | 2013-10-12 | Method for predicting variables of users on basis of artificial neural networks and D-S (Dempster-Shafer) evidence theory |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103544539A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104317910A (en) * | 2014-10-27 | 2015-01-28 | 国家电网公司 | Data processing method and device |
CN104954185A (en) * | 2015-06-09 | 2015-09-30 | 中国石油大学(华东) | Cloud computing load prediction method based on deep belief network |
CN108376300A (en) * | 2018-03-02 | 2018-08-07 | 江苏电力信息技术有限公司 | A kind of user power utilization behavior prediction method based on probability graph model |
CN108764520A (en) * | 2018-04-11 | 2018-11-06 | 杭州电子科技大学 | A kind of water quality parameter prediction technique based on multilayer circulation neural network and D-S evidence theory |
WO2020147265A1 (en) * | 2019-01-14 | 2020-07-23 | 南京信息工程大学 | Mobile electronic commerce recommendation method and system employing multisource information fusion |
CN113763356A (en) * | 2021-09-08 | 2021-12-07 | 国网江西省电力有限公司电力科学研究院 | Target detection method based on visible light and infrared image fusion |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6032125A (en) * | 1996-11-07 | 2000-02-29 | Fujitsu Limited | Demand forecasting method, demand forecasting system, and recording medium |
CN101034389A (en) * | 2007-03-19 | 2007-09-12 | 江西省电力科学研究院 | Electrical energy power quality disturbance automatic identification method and system based on information fusion |
CN102395135A (en) * | 2011-10-25 | 2012-03-28 | 江苏省邮电规划设计院有限责任公司 | VLR (Visitor Location Register) user number predicting method based on gray system model neural network |
CN103147972A (en) * | 2013-03-19 | 2013-06-12 | 北京化工大学 | Reciprocating-type compressor fault diagnosis method based on multi-sensor information fusion |
-
2013
- 2013-10-12 CN CN201310474134.2A patent/CN103544539A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6032125A (en) * | 1996-11-07 | 2000-02-29 | Fujitsu Limited | Demand forecasting method, demand forecasting system, and recording medium |
CN101034389A (en) * | 2007-03-19 | 2007-09-12 | 江西省电力科学研究院 | Electrical energy power quality disturbance automatic identification method and system based on information fusion |
CN102395135A (en) * | 2011-10-25 | 2012-03-28 | 江苏省邮电规划设计院有限责任公司 | VLR (Visitor Location Register) user number predicting method based on gray system model neural network |
CN103147972A (en) * | 2013-03-19 | 2013-06-12 | 北京化工大学 | Reciprocating-type compressor fault diagnosis method based on multi-sensor information fusion |
Non-Patent Citations (1)
Title |
---|
郭润龙: "基于RBF神经网络与D-S证据理论的数据融合方法及其应用研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》, no. 1, 15 December 2011 (2011-12-15) * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104317910A (en) * | 2014-10-27 | 2015-01-28 | 国家电网公司 | Data processing method and device |
CN104954185A (en) * | 2015-06-09 | 2015-09-30 | 中国石油大学(华东) | Cloud computing load prediction method based on deep belief network |
CN104954185B (en) * | 2015-06-09 | 2018-08-28 | 青岛邃智信息科技有限公司 | A kind of cloud computing load predicting method based on depth confidence net |
CN108376300A (en) * | 2018-03-02 | 2018-08-07 | 江苏电力信息技术有限公司 | A kind of user power utilization behavior prediction method based on probability graph model |
CN108764520A (en) * | 2018-04-11 | 2018-11-06 | 杭州电子科技大学 | A kind of water quality parameter prediction technique based on multilayer circulation neural network and D-S evidence theory |
CN108764520B (en) * | 2018-04-11 | 2021-11-16 | 杭州电子科技大学 | Water quality parameter prediction method based on multilayer cyclic neural network and D-S evidence theory |
WO2020147265A1 (en) * | 2019-01-14 | 2020-07-23 | 南京信息工程大学 | Mobile electronic commerce recommendation method and system employing multisource information fusion |
CN113763356A (en) * | 2021-09-08 | 2021-12-07 | 国网江西省电力有限公司电力科学研究院 | Target detection method based on visible light and infrared image fusion |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103544539A (en) | Method for predicting variables of users on basis of artificial neural networks and D-S (Dempster-Shafer) evidence theory | |
CN111104981B (en) | Hydrological prediction precision evaluation method and system based on machine learning | |
CN107590565A (en) | A kind of method and device for building building energy consumption forecast model | |
CN106055918B (en) | Method for identifying and correcting load data of power system | |
CN111046581B (en) | Power transmission line fault type identification method and system | |
CN105740991B (en) | Climate change prediction method and system based on improved BP neural network fitting of multiple climate modes | |
CN108920863B (en) | Method for establishing energy consumption estimation model of robot servo system | |
CN112488395A (en) | Power distribution network line loss prediction method and system | |
CN103942461A (en) | Water quality parameter prediction method based on online sequential extreme learning machine | |
CN113205207A (en) | XGboost algorithm-based short-term power consumption load fluctuation prediction method and system | |
CN104318332A (en) | Power load predicting method and device | |
CN107358318A (en) | Based on GM(1,1)The urban power consumption Forecasting Methodology of model and Grey Markov chain predicting model | |
CN111369070A (en) | Envelope clustering-based multimode fusion photovoltaic power prediction method | |
CN110232483A (en) | Deep learning load forecasting method, device and terminal device | |
CN113901977A (en) | Deep learning-based power consumer electricity stealing identification method and system | |
CN104112062B (en) | The acquisition methods of wind-resources distribution based on interpolation method | |
CN109598052B (en) | Intelligent ammeter life cycle prediction method and device based on correlation coefficient analysis | |
CN112686481A (en) | Runoff forecasting method and processor | |
CN114548509A (en) | Multi-type load joint prediction method and system for multi-energy system | |
CN112215398A (en) | Power consumer load prediction model establishing method, device, equipment and storage medium | |
CN106295877B (en) | Method for predicting electric energy consumption of smart power grid | |
CN115358437A (en) | Power supply load prediction method based on convolutional neural network | |
CN105139287A (en) | Comprehensive electric energy metering device state assessment method integrating safety domain | |
CN112347531B (en) | Brittle marble Dan Sanwei crack propagation path prediction method and system | |
CN116707331B (en) | Inverter output voltage high-precision adjusting method and system based on model prediction |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | 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: 20140129 |
|
RJ01 | Rejection of invention patent application after publication |