CN108629441A - Prediction technique and device based on clustering and the improved fan noise of small echo - Google Patents
Prediction technique and device based on clustering and the improved fan noise of small echo Download PDFInfo
- Publication number
- CN108629441A CN108629441A CN201810202707.9A CN201810202707A CN108629441A CN 108629441 A CN108629441 A CN 108629441A CN 201810202707 A CN201810202707 A CN 201810202707A CN 108629441 A CN108629441 A CN 108629441A
- Authority
- CN
- China
- Prior art keywords
- data
- training
- small echo
- clustering
- parameter
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Human Resources & Organizations (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Game Theory and Decision Science (AREA)
- Biomedical Technology (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Development Economics (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A kind of prediction technique and device based on clustering and the improved fan noise of small echo, method include:Obtain wind turbine training parameter, wherein the training parameter includes the geometric parameter and performance parameter of wind turbine;The training parameter is analyzed, wherein the analysis includes clustering and correlation analysis;It is established according to the result of analysis and is based on small echo improved BP-NN model, and carry out neural metwork training;Obtain data to be predicted, and data to be predicted are combined with the training parameter, and prediction data is treated according to the neural network and is predicted, the training precision that neural network can be improved by way of data clusters effectively reduces the prediction error of neural network by the improved BP neural network of small echo.
Description
Technical field
The present invention relates to technical field of data prediction, more particularly to a kind of to be made an uproar based on clustering and the improved wind turbine of small echo
The prediction of the prediction technique of sound, a kind of non-transitory readable storage medium storing program for executing and a kind of clustering and the improved fan noise of small echo
Device.
Background technology
Fan design is designed generally according to the requirement of client by similar theory, is made an uproar in the related technology to wind turbine
Sound carry out prediction be typically based on fluid or mechanical mechanism, still, the relevant technologies the problem is that, the prediction based on certain features
The feature of whole whole can not be characterized, therefore the noise figure forecasting inaccuracy of wind turbine can be caused true, and then wind turbine sample usually occurs
The noise figure of machine needs the case where rectification higher than customer requirement, causes model machine needs to be supplemented with money, not only causes economic loss, also prolong
Project process is delayed, therefore, the relevant technologies need to improve.
Invention content
The prediction technique based on clustering and the improved fan noise of small echo that the object of the present invention is to provide a kind of, can
Fan noise value is accurately predicted in the fan design stage.
Another aspect of the present invention embodiment proposes a kind of non-transitory readable storage medium storing program for executing.
Another aspect of the invention embodiment proposes a kind of prediction dress based on clustering and the improved fan noise of small echo
It sets.
To solve the above problems, the first aspect of the present invention, which provides one kind, being based on clustering and the improved wind turbine of small echo
The prediction technique of noise, includes the following steps:Obtain wind turbine training parameter, wherein the training parameter includes the geometry of wind turbine
Parameter and performance parameter;The training parameter is analyzed, wherein the analysis includes clustering and correlation analysis;
It is established according to the result of analysis and is based on small echo improved BP-NN model, and carry out neural metwork training;It obtains to be predicted
Data, and data to be predicted are combined with the training parameter, and prediction data is treated according to the neural network and is carried out in advance
It surveys.
Further, carrying out analysis to the training parameter includes:The training parameter is classified and clustered
Analysis, to obtain multiple classification;The correlation between the training parameter and output parameter each classified is calculated, and is ranked up.
Further, described established according to the result of the analysis is based on small echo improved BP-NN model, goes forward side by side
Row neural metwork training includes:It is established according to the result of the analysis and is based on the improved BP neural network of small echo;According to the instruction
Practice data to be trained the improved BP neural network of the small echo, to obtain the corresponding neural network model of each cluster.
Further, data to be predicted are obtained, and data to be predicted are combined with the training parameter, and to be predicted
Data carry out prediction:Data to be tested are obtained, data to be tested are combined with training parameter, to obtain new input data
Combination;Clustering is carried out to the new input data combination, to judge the cluster belonging to the data to be tested;It will be to be checked
Measured data inputs the corresponding neural network of cluster belonging to it, to carry out noise-predictive to data to be tested.
According to another aspect of the present invention, a kind of non-transitory readable storage medium storing program for executing, including it is stored thereon with computer
Program realizes the prediction side based on clustering and the improved fan noise of small echo when the program is executed by processor
Method.
According to another aspect of the invention, a kind of prediction meanss based on clustering and the improved fan noise of small echo,
It is characterised in that it includes:Acquisition module, for obtaining wind turbine training parameter, wherein the training parameter includes the geometry of wind turbine
Parameter and performance parameter;Analysis module, for analyzing the training parameter, wherein the analysis includes clustering
And correlation analysis;Neural network module is based on small echo improved BP-NN model for being established according to the result of analysis,
And carry out neural metwork training;Prediction module, for obtaining data to be predicted, and by data to be predicted and the training parameter group
It closes, and prediction data is treated according to the neural network and is predicted.
According to the prediction technique based on clustering and the improved fan noise of small echo that present example proposes, by obtaining
Wind turbine training parameter is taken, then training parameter is analyzed, is established based on the improved BP nerves of small echo according to the result of analysis
Network model obtains data to be predicted, and data to be predicted is combined with training parameter, and according to neural network to be predicted
Data are predicted.The prediction technique of the embodiment of the present invention can improve neural network by way of data clusters as a result,
Training precision effectively reduces the prediction error of neural network by the improved BP neural network of small echo.
Description of the drawings
Fig. 1 is the flow chart of the prediction technique based on clustering and the improved fan noise of small echo;
Fig. 2 is the waveform diagram for a variety of transmission functions that the embodiment of the present invention proposes;
Fig. 3 is the block diagram of the prediction meanss based on clustering and the improved fan noise of small echo.
Specific implementation mode
In order to make the objectives, technical solutions and advantages of the present invention clearer, With reference to embodiment and join
According to attached drawing, the present invention is described in more detail.It should be understood that these descriptions are merely illustrative, and it is not intended to limit this hair
Bright range.In addition, in the following description, descriptions of well-known structures and technologies are omitted, to avoid this is unnecessarily obscured
The concept of invention.
Schematic diagram of a layer structure according to the ... of the embodiment of the present invention is shown in the accompanying drawings.These figures are not drawn to scale
, wherein for purposes of clarity, some details are magnified, and some details may be omitted.It is shown in the drawings various
Region, the shape of layer and the relative size between them, position relationship are merely exemplary, in practice may be public due to manufacture
Difference or technology restriction and be deviated, and those skilled in the art may be additionally designed as required with not similar shape
Shape, size, the regions/layers of relative position.
Obviously, described embodiments are some of the embodiments of the present invention, instead of all the embodiments.Based on the present invention
In embodiment, every other implementation obtained by those of ordinary skill in the art without making creative efforts
Example, shall fall within the protection scope of the present invention.
In the description of the present invention, it should be noted that term " first ", " second ", " third " are used for description purposes only,
It is not understood to indicate or imply relative importance.
As long as in addition, technical characteristic involved in invention described below different embodiments non-structure each other
It can be combined with each other at conflict.
Hereinafter reference will be made to the drawings is more fully described the present invention.In various figures, identical element is using similar attached
Icon is remembered to indicate.For the sake of clarity, the various pieces in attached drawing are not necessarily to scale.
In Application of Neural Network, most noticeable is training precision and precision of prediction and generalization ability.The present invention
One of contribution is classify to input sample and test sample by clustering to improve sample quality, is neural network instruction
White silk is ready;Contribution second is that pre- to carry out network by the improved BP neural network transmission functions of wavelet function Morlet
It surveys, to improve the precision of fan noise value prediction by clustering strategy and small echo improvement.
Below with reference to the accompanying drawings the pre- based on clustering and the improved fan noise of small echo of the embodiment of the present invention described
Survey method.
Fig. 1 is the stream according to the prediction technique based on clustering and the improved fan noise of small echo of the embodiment of the present invention
Cheng Tu.As shown in Figure 1, the prediction technique based on clustering and the improved fan noise of small echo of the embodiment of the present invention, including
Following steps:
S1:Obtain wind turbine training parameter.
Wherein, training parameter includes the geometric parameter and performance parameter of wind turbine.
A specific embodiment according to the present invention, the data provided using the really holding central laboratory of connection, wherein for
The training parameter of analysis includes flow, total head, rotating speed, power, efficiency and air outlet velocity, i.e. 6 performance parameters and impeller is straight
Chord length, blade tip established angle, blade tip chord length, the number of blade in established angle, leaf in diameter, hub ratio, blade root established angle, blade root chord length, leaf
With guide vane number, i.e. 10 geometric parameters.Wherein, 16 training parameters are sorted with said sequence.
It should be understood that only having only one output parameter, the i.e. sound pressure level of fan noise in present example.
S2:Training parameter is analyzed.Wherein, analysis includes clustering and correlation analysis.
According to one embodiment of present invention, carrying out analysis to training parameter includes:
S101:Classified to training parameter and carry out clustering, to obtain multiple classification.
It should be noted that training parameter number more at most divide cluster it is bigger, as a result, can according to classification the case where from each point
A certain proportion of sample is chosen in cluster as training parameter.According to a particular embodiment of the invention, training parameter is 200,
Training parameter can be divided into 8 clusters.
It should be understood that carry out clustering to training parameter has remarkable efficacy in terms of improving sample quality.In reality
In the operation of border, number of samples can cause the poor fitting of neural network very little, and number of samples cross can at most cause training be fitted and
It cannot get rational neural network model.Therefore, the embodiment of the present invention to training parameter by carrying out Cluster Classification, so as to every
A class carries out the training of neural network respectively.
S102:The correlation between the training parameter and output parameter each classified is calculated, and is ranked up.
Specifically, 16 training parameters and output parameter are normalized according to the correlation properties of training parameter,
Then correlation analysis is carried out, and the importance of training parameter is ranked up according to the size of correlation.
Wherein, correlation is the statistical indicator of correlativity level of intimate between reflecting variable, is implemented in the present invention
Correlation is indicated using Pearson correlation coefficient in example.According to a particular embodiment of the invention, according to 16 training parameters and
Output parameter relevance ranking is:[4 7621 15 10 5 16 98 11 12 3 13 14], wherein maximum correlation
Coefficient is 0.7294, variable serial number 4 (i.e. performance parameter power), shows that with the maximally related variable of noise be power, the second phase
The variable of pass is geometric parameter impeller diameter, and the relevant variable of third is air outlet velocity, and so on.
S3:It is established according to the result of analysis and is based on small echo improved BP-NN model, and carry out neural metwork training.
According to one embodiment of present invention, it is established according to the result of analysis and is based on the improved BP neural network mould of small echo
Type, and carry out neural metwork training and include:It is established according to the result of the analysis and is based on the improved BP neural network of small echo;According to
Training data is trained the improved BP neural network of small echo, to obtain the corresponding neural network model of each cluster.
It should be noted that in an embodiment of the present invention, the input parameter of neural network is 16 training parameters, output
Parameter is noise sound pressure level.
It should also be noted that, in a specific embodiment of the present invention, the input parameter of the input layer of neural network should be 7
And output parameter is 1, to ensure that neural network is not in over-fitting and poor fitting.That is, in embodiments of the present invention
Above-mentioned and input parameter of the 7 forward parameters of output parameter relevance ranking as input layer can be chosen.
There are many transmission functions of BP neural network.Wherein, Log-sigmoid and Tan-sigmoid (Sigmoid, S types
Function) expression formula of basic function is:
Wherein, x is variable, and e is an irrational number, is approximately equal to 2.718281828.As shown in Fig. 2, Log-sigmoid types
The domain of function is R, and codomain is (0,1);The domain of Tan-sigmoid type transmission functions is R, and codomain is (- 1 ,+1);Line
The domain and codomain of property transmission function purelin is R, wherein R is real number.
Include at least one hidden layer in BP neural network, sigmoid type transmission functions can be used as the nerve in hidden layer
First transmission function, linear transfer function are used in the neuron of output layer, and the output of whole network can be arbitrary value.
Based on this, present example proposition substitutes hidden layer transmission function with small echo Morlet functions, with to BP nerves
Network is improved.
Wherein, small echo Morlet basic functions expression formula is:
Wherein, x is independent variable, and e is an irrational number, and it is to show wavelet scale coefficient to be approximately equal to 2.718281828, a.It is small
A values are different in wave Morlet basic functions, then the training precision and predictive ability of the BP neural network built also differ.In this hair
The purpose of bright embodiment is to improve precision of prediction, therefore compare the error condition that 4 kinds of transmission functions predict 8 unknown points, such as
Shown in table 1.
Table 1
Specifically, as a=0.49, the predictive ability of BP neural network and the generalized ability of neural network are most strong.
Based on this, the embodiment of the present invention establishes BP nerve nets using small echo Morlet functions as hidden layer transmission function
Network.
On the other hand, training data is subjected to integrity analysis, in embodiments of the present invention, training parameter can be 201,
It is 161 remaining after integrity analysis, and 161 training parameters are numbered, it, can according to clustering analysis result
8 data for being belonging respectively to each cluster are extracted as test data:[68 83 103 115 129 147 153 159],
153 remaining datas are trained the improved BP neural network of small echo as training data, according to training data, each to obtain
It is a to cluster corresponding neural network model.
That is, 161 training parameters are carried out clustering, 8 Cluster Classifications are obtained, one is extracted to each cluster
For a data as test data, remaining 153 training data is input to the improved BP neural network of small echo, and is divided as unit of class
It is other that neural network is trained, to obtain neural network model corresponding with each class.
S4:Data to be predicted are obtained, and data to be predicted are combined with training parameter, and are treated according to neural network pre-
Measured data is predicted.
According to one embodiment of present invention, data to be tested are obtained, data to be tested are combined with training parameter, to obtain
Obtain input data combination newly;Clustering is carried out to new input data combination, to judge the cluster belonging to data to be tested;
Data to be tested are inputted into the corresponding neural network of cluster belonging to it, to carry out noise-predictive to data to be tested.
In embodiments of the present invention, data to be predicted can be the test data extracted in above-mentioned steps.
Test data and training data are combined, combined with obtaining new input parameter, and to new input parameter
Combination carries out clustering, then judges the cluster belonging to test data, test data is inputted to the cluster belonging to it respectively
Corresponding neural network model, you can noise-predictive is carried out to test data.
NI180059
Wherein, the cluster belonging to data to be predicted can be determined by for cycles.
Based on this, above-mentioned test data is verified, as shown in table 2:
Table 2
Wherein, unique record number is the number of test data, and BP networks are to be passed using Tan-sigmoid as hidden layer
Using small echo Morlet functions as the neural network of hidden layer transmission function, i.e., the neural network of delivery function, wavelet network are
The improved BP neural network of small echo of the present embodiment.
It is found that the worst error absolute value of the prediction result of the improved BP neural network of small echo is respectively less than 1% substantially, i.e.,
The prediction result of speech, the improved BP neural network of small echo is better than Tan-sigmoid type BP neural networks.
Prediction technique according to the ... of the embodiment of the present invention based on clustering and the improved fan noise of small echo as a result, energy
It is enough that clustering is carried out to data to be detected, noise-predictive is then carried out according to the corresponding neural network model of affiliated cluster,
To not only avoid extraction test parameter blindness, but also can to avoid in the case where sample size is big by the excessive institute of sample
The training overfitting problem brought can also enhance measuring accuracy and generalized ability while effectively improving training precision.
The embodiment of the present invention also proposed a kind of non-transitory readable storage medium storing program for executing, be stored thereon with computer program, should
The prediction technique based on clustering and the improved fan noise of small echo is realized when program is executed by processor.
Fig. 3 is the block diagram of the prediction meanss based on clustering and the improved fan noise of small echo.Such as Fig. 3 institutes
Show, the prediction meanss based on clustering and the improved fan noise of small echo of the embodiment of the present invention include:Acquisition module 10 divides
Analyse module 20, neural network module 30 and prediction module 40.
Wherein, acquisition module 10 is for obtaining wind turbine training parameter, wherein training parameter include wind turbine geometric parameter and
Performance parameter;Analysis module 20 is for analyzing training parameter, wherein analysis includes clustering and correlation analysis;
Neural network module 30, which is used to be established according to the result of analysis, is based on small echo improved BP-NN model, and carries out nerve net
Network training;Prediction module 40 combines data to be predicted with training parameter for obtaining data to be predicted, and according to nerve
Network handles prediction data is predicted.
Prediction meanss according to the ... of the embodiment of the present invention based on clustering and the improved fan noise of small echo, pass through acquisition
Module obtains wind turbine training parameter, and analysis module analyzes training parameter, and neural network module is built according to the result of analysis
Be based on small echo improved BP-NN model, and carries out neural metwork training, and then prediction module obtains data to be predicted,
And combine data to be predicted with training parameter, prediction data prediction is treated according to neural network.The embodiment of the present invention as a result,
Prediction meanss can improve the training precision of neural network by way of data clusters, pass through the improved BP neural network of small echo
Effectively reduce the prediction error of neural network.
It should be understood that the above-mentioned specific implementation mode of the present invention is used only for exemplary illustration or explains the present invention's
Principle, but not to limit the present invention.Therefore, that is done without departing from the spirit and scope of the present invention is any
Modification, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.In addition, appended claims purport of the present invention
Covering the whole variations fallen into attached claim scope and boundary or this range and the equivalent form on boundary and is repairing
Change example.
In the above description, the technical details such as composition, the etching of each layer are not described in detail.But
It will be appreciated by those skilled in the art that can be by various means in the prior art, to form layer, the region of required shape
Deng.In addition, in order to form same structure, those skilled in the art can be devised by and process as described above not fully phase
Same method.
The present invention is described above by reference to the embodiment of the present invention.But these embodiments are used for the purpose of saying
Bright purpose, and be not intended to limit the scope of the invention.The scope of the present invention is limited by appended claims and its equivalent.
The scope of the present invention is not departed from, those skilled in the art can make a variety of substitutions and modifications, these substitutions and modifications should all be fallen
Within the scope of the present invention.
Although embodiments of the present invention are described in detail, it should be understood that, without departing from the present invention's
In the case of spirit and scope, can embodiments of the present invention be made with various changes, replacement and change.
Obviously, the above embodiments are merely examples for clarifying the description, and does not limit the embodiments.It is right
For those of ordinary skill in the art, can also make on the basis of the above description it is other it is various forms of variation or
It changes.There is no necessity and possibility to exhaust all the enbodiments.And it is extended from this it is obvious variation or
It changes still within the protection scope of the invention.
It should be understood by those skilled in the art that, the embodiment of the present invention can be provided as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, the present invention can be used in one or more wherein include computer usable program code computer
The computer program production implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.)
The form of product.
The present invention be with reference to according to the method for the embodiment of the present invention, the flow of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that can be realized by computer program instructions every first-class in flowchart and/or the block diagram
The combination of flow and/or box in journey and/or box and flowchart and/or the block diagram.These computer programs can be provided
Instruct the processor of all-purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine so that the instruction executed by computer or the processor of other programmable data processing devices is generated for real
The device for the function of being specified in present one flow of flow chart or one box of multiple flows and/or block diagram or multiple boxes.
These computer program instructions, which may also be stored in, can guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works so that instruction generation stored in the computer readable memory includes referring to
Enable the manufacture of device, the command device realize in one flow of flow chart or multiple flows and/or one box of block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device so that count
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, in computer or
The instruction executed on other programmable devices is provided for realizing in one flow of flow chart or multiple flows and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Claims (6)
1. a kind of prediction technique based on clustering and the improved fan noise of small echo, which is characterized in that include the following steps:
Obtain wind turbine training parameter, wherein the training parameter includes the geometric parameter and performance parameter of wind turbine;
The training parameter is analyzed, wherein the analysis includes clustering and correlation analysis;
It is established according to the result of analysis and is based on small echo improved BP-NN model, and carry out neural metwork training;
Data to be predicted are obtained, and data to be predicted are combined with the training parameter, and are treated according to the neural network
Prediction data is predicted.
2. the prediction technique according to claim 1 based on clustering and the improved fan noise of small echo, feature exist
In carrying out analysis to the training parameter includes:
Classified to the training parameter and carry out clustering, to obtain multiple classification;
The correlation between the training parameter and output parameter each classified is calculated, and is ranked up.
3. the prediction technique according to claim 1 based on clustering and the improved fan noise of small echo, feature exist
In described established according to the result of the analysis is based on small echo improved BP-NN model, and carries out neural metwork training
Including:
It is established according to the result of the analysis and is based on the improved BP neural network of small echo;
The improved BP neural network of the small echo is trained according to the training data, to obtain the corresponding god of each cluster
Through network model.
4. the prediction technique according to claim 1 based on clustering and the improved fan noise of small echo, feature exist
In obtaining data to be predicted, and data to be predicted are combined with the training parameter, and treat prediction data and carry out prediction packet
It includes:
Data to be tested are obtained, data to be tested are combined with training parameter, are combined with obtaining new input data;
Clustering is carried out to the new input data combination, to judge the cluster belonging to the data to be tested;
Data to be tested are inputted into the corresponding neural network of cluster belonging to it, to carry out noise-predictive to data to be tested.
5. a kind of non-transitory readable storage medium storing program for executing, is stored thereon with computer program, which is characterized in that the program is by processor
The prediction side based on clustering and the improved fan noise of small echo as described in any in claim 1-4 is realized when execution
Method.
6. a kind of prediction meanss based on clustering and the improved fan noise of small echo, which is characterized in that including:
Acquisition module, for obtaining wind turbine training parameter, wherein the training parameter includes the geometric parameter and performance ginseng of wind turbine
Number;
Analysis module, for analyzing the training parameter, wherein the analysis includes clustering and correlation point
Analysis;
Neural network module is based on small echo improved BP-NN model for being established according to the result of analysis, and carries out god
Through network training;
Prediction module is combined for obtaining data to be predicted, and by data to be predicted with the training parameter, and according to described
Neural network is treated prediction data and is predicted.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810202707.9A CN108629441A (en) | 2018-03-13 | 2018-03-13 | Prediction technique and device based on clustering and the improved fan noise of small echo |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810202707.9A CN108629441A (en) | 2018-03-13 | 2018-03-13 | Prediction technique and device based on clustering and the improved fan noise of small echo |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108629441A true CN108629441A (en) | 2018-10-09 |
Family
ID=63706217
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810202707.9A Pending CN108629441A (en) | 2018-03-13 | 2018-03-13 | Prediction technique and device based on clustering and the improved fan noise of small echo |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108629441A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109707658A (en) * | 2019-02-28 | 2019-05-03 | 苏州尼昂科技有限公司 | Method for determination of performance parameter, device and the electronic equipment of blower |
CN112465347A (en) * | 2020-11-26 | 2021-03-09 | 湖南科技大学 | Method for cooperatively predicting roof stability based on cluster analysis and improved neural network |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102938093A (en) * | 2012-10-18 | 2013-02-20 | 安徽工程大学 | Wind power forecasting method |
CN103425867A (en) * | 2013-06-24 | 2013-12-04 | 武汉大学 | Short-term wind power combination prediction method |
-
2018
- 2018-03-13 CN CN201810202707.9A patent/CN108629441A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102938093A (en) * | 2012-10-18 | 2013-02-20 | 安徽工程大学 | Wind power forecasting method |
CN103425867A (en) * | 2013-06-24 | 2013-12-04 | 武汉大学 | Short-term wind power combination prediction method |
Non-Patent Citations (3)
Title |
---|
史文雯: ""城市轨道交通短时客流预测与最优客运能力调配问题的研究"", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
程静等: ""基于回归分析与BP神经网络的风机噪声预测"", 《噪声与振动控制》 * |
魏桢等: ""基于Elman神经网络的风电场噪声预测模型"", 《水电能源科学》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109707658A (en) * | 2019-02-28 | 2019-05-03 | 苏州尼昂科技有限公司 | Method for determination of performance parameter, device and the electronic equipment of blower |
CN112465347A (en) * | 2020-11-26 | 2021-03-09 | 湖南科技大学 | Method for cooperatively predicting roof stability based on cluster analysis and improved neural network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105512799B (en) | Power system transient stability evaluation method based on mass online historical data | |
CN107577605A (en) | A kind of feature clustering system of selection of software-oriented failure prediction | |
US9037279B2 (en) | Clustering for prediction models in process control and for optimal dispatching | |
CN108344812A (en) | A kind of Diagnosis Method of Transformer Faults, device and storage medium | |
CN107360026B (en) | Distributed message middleware performance prediction and modeling method | |
CN110689070B (en) | Training method and device of business prediction model | |
CN109241849B (en) | Feature decomposition selection and fault diagnosis method for main engine of intelligent power plant steam turbine | |
CN108763096A (en) | Software Defects Predict Methods based on depth belief network algorithm support vector machines | |
CN108629441A (en) | Prediction technique and device based on clustering and the improved fan noise of small echo | |
CN111062425A (en) | Unbalanced data set processing method based on C-K-SMOTE algorithm | |
CN111898443A (en) | Flow monitoring method for wire feeding mechanism of FDM type 3D printer | |
CN109426655A (en) | Data analysing method, device, electronic equipment and computer readable storage medium | |
Tohyama | Varieties of Asian welfare capitalisms and the influence of globalization | |
CN110046593A (en) | The complex electric energy quality disturbance recognition methods of S-transformation and random forest is improved based on segmentation | |
CN110222384A (en) | A kind of Analysis of Bridge Cracks method, apparatus and storage medium | |
CN116564409A (en) | Machine learning-based identification method for sequencing data of transcriptome of metastatic breast cancer | |
CN112215398A (en) | Power consumer load prediction model establishing method, device, equipment and storage medium | |
CN117454721B (en) | Wind power plant wake superposition effect evaluation method and medium based on digital simulation experiment | |
JP5614297B2 (en) | Index calculation program and method, and design support apparatus | |
JP2017198620A (en) | Abnormality diagnosis device and abnormality diagnosis method | |
CN110826587B (en) | Vortex shaft engine fault detection method based on improved weighted support vector machine | |
CN116222753A (en) | Rotor system fault sensitivity feature extraction method and system | |
JP4994199B2 (en) | Machine learning apparatus and machine learning method | |
CN114077924A (en) | Wind power prediction gale data enhancement method considering extreme gale weather | |
CN112035533B (en) | System resource scheduling method and device based on multi-parameter quantization strategy feedback |
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 |