CN108629095A - A kind of modeling method of gearbox of wind turbine bearing temperature - Google Patents
A kind of modeling method of gearbox of wind turbine bearing temperature Download PDFInfo
- Publication number
- CN108629095A CN108629095A CN201810364855.0A CN201810364855A CN108629095A CN 108629095 A CN108629095 A CN 108629095A CN 201810364855 A CN201810364855 A CN 201810364855A CN 108629095 A CN108629095 A CN 108629095A
- Authority
- CN
- China
- Prior art keywords
- neural network
- model
- bearing temperature
- network model
- data
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/17—Mechanical parametric or variational design
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Physics & Mathematics (AREA)
- Geometry (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computational Mathematics (AREA)
- Wind Motors (AREA)
Abstract
The invention discloses a kind of modeling methods of gearbox of wind turbine bearing temperature, including step:1) running of wind generating set data are obtained;2) aspect of model database is established;3) the feedforward neural network model structure of box bearing temperature is determined;4) neural network model is trained;5) neural network model is tested.The present invention is directed to by acquiring a large amount of unit history datas, and it is calculated and is analyzed, box bearing temperature model is established, the relationship of the input parameter and box bearing temperature of reflection unit operation operating mode, the final zone of reasonableness for obtaining the box bearing temperature under different operating modes are established.
Description
Technical field
The present invention relates to the technical fields of wind-power electricity generation, refer in particular to a kind of modeling of gearbox of wind turbine bearing temperature
Method.
Background technology
The gear-box of Wind turbines is the key that influence the big component of Wind turbines normal operation, is the low of connection impeller drive
The component of the high speed shaft of fast axis and generator.The box bearing temperature of Wind turbines is to reflect gear-box whether normal operation
Key parameter, with the change of running of wind generating set operating mode, box bearing temperature also can be with change.
Judge that the conventional method of box bearing temperature anomaly is that the threshold value of an early warning is set to box bearing temperature,
This threshold value is fixed value, is adjusted in real time not according to running of wind generating set operating mode.Therefore this threshold range is larger, no
It can reflect the health status of unit in time.
Box bearing temperature model is to be reflected under different duty parameter inputs, and box bearing temperature should be at
Reasonable temperature range.It establishes box bearing temperature model and can get the box bearing temperature under unit difference operating condition
Zone of reasonableness, discover the abnormality of gear-box in time, avoid abnormal extension, reduce because being damaged caused by gear-box exception
It loses.
Invention content
The shortcomings that it is an object of the invention to overcome the prior art and deficiency, it is proposed that a kind of reliable Wind turbines gear
The modeling method of axle box bearing temperature, it is intended to by acquiring a large amount of unit history datas, and be calculated and be analyzed, establish tooth
Roller box bearing temperature model is established the relationship of the input parameter and box bearing temperature of reflection unit operation operating mode, is finally obtained
Obtain the zone of reasonableness of the box bearing temperature under different operating modes.
To achieve the above object, technical solution provided by the present invention is:A kind of gearbox of wind turbine bearing temperature
Modeling method includes the following steps:
1) running of wind generating set data are obtained
Using wind power plant data collecting system acquire Wind turbines whole year operation data, including set gear box oil temperature,
Fuel pump outlet pressure, environment temperature, unit generation power, wind speed, generator speed, gear case oil position data and box bearing
Temperature;
2) aspect of model database is established
First, gathered data is pre-processed, removes the unreasonable data of value range;Then gathered data is carried out again
Normalized set is calculated unit and is averaged generated output, mean wind speed, average generator speed and average gear case oil position,
And be averaged generated output, mean wind speed, average generator of gear-box oil temperature, fuel pump outlet pressure, environment temperature, unit is turned
As input data, box bearing temperature establishes aspect of model database as output data for speed, average gear-box oil level;
3) the feedforward neural network model structure for determining box bearing temperature, is specifically to determine the hidden of neural network model
Containing number and neuronal quantity layer by layer, activation primitive is determined;
4) neural network model is trained
Using one group of inputoutput data as a sample, using a part of sample in aspect of model database as training
Sample, a part of sample is as test sample;The model training for the neural network model that training sample is determined for step 3),
To obtain the weights and amount of bias of neural network model;
5) neural network model is tested
If the input of test sample is xm, export as ym, according to the Artificial Neural Network Structures and step 4) of step 3) institute
The weights and amount of bias for obtaining neural network model are input with xm, calculate the output ya of neural network model;Enable err=| ym-
Ya | be model bias, if the sample proportion that model bias is more than preset value is more than preset ratio p%, return to step 3) again
It determines Artificial Neural Network Structures, and executes step 4) and step 5) successively;If model bias is more than the sample proportion of preset value
Less than or equal to preset ratio p%, then the neural network model determined with the Artificial Neural Network Structures of step 3), step 4) is weighed
Value and amount of bias form box bearing temperature model.
Compared with prior art, the present invention having the following advantages that and advantageous effect:
1, the annual data unit operation of program covering, basic to cover Wind turbines major part operating mode, the model of foundation is accurate
True property is high.
2, on the basis of initial data, using the more data characteristicses of statistic information extraction, model can be improved in the program
Accuracy.
3, the box bearing temperature model of the program can reflect Wind turbines in different operating condition lower gear axle box bearings
The zone of reasonableness of temperature can reflect whether box bearing is abnormal under various operating modes, and gear-box is avoided to magnify extremely, reduces
Unit fault time.
Description of the drawings
Fig. 1 is the logic flow schematic diagram of the method for the present invention.
Specific implementation mode
The present invention is further explained in the light of specific embodiments.
As shown in Figure 1, the modeling method for the gearbox of wind turbine bearing temperature that the present embodiment is provided, concrete condition
It is as follows:
1) running of wind generating set data are obtained
1 year operation data of Wind turbines that the same type of wind power plant is acquired using wind power plant SCADA system, is adopted for every 1 minute
Collect one group of data, gathered data includes set gear box oil temperature, fuel pump outlet pressure, environment temperature, unit generation power, wind
Speed, generator speed, gear-box oil level and box bearing temperature.
2) aspect of model database is established
Set gear box oil temperature, fuel pump outlet pressure, environment temperature, unit generation power, wind speed, generator is set separately
The maximum value and minimum value of rotating speed, gear-box oil level and box bearing temperature, removal value range is not in maximum value and minimum
The data being worth in range.
Normalized set is carried out to gathered data, calculates sample mean, the average value of use can be from 10 minutes to 2
The average value of hour.By taking 10 minutes average value as an example, 1 sample of acquisition in each minute in 10 minutes, 10 of acquisition in 10 minutes
Sample is x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, then 10 minutes average value is:Xmean=(x1+x2+x3+x4+x5+
x6+x7+x8+x9+x10)/10.10 minutes average generated outputs of unit are calculated, 10 minutes mean wind speeds are averaged for 10 minutes
Generator speed and 10 minutes average gear-box oil levels.
By gear-box oil temperature, fuel pump outlet pressure, environment temperature, 10 minutes average generated outputs of unit, it is averaged within 10 minutes
Wind speed, 10 minutes average generator speeds, 10 minutes average gear-box oil levels are as input data, box bearing temperature conduct
Output data establishes aspect of model database.
3) the feedforward neural network model structure of box bearing temperature is determined
The hidden layer of neural network model is 1 layer and neuronal quantity is 5, using tanh S type functionsAs activation primitive.
4) neural network model is trained
Training sample accounting is between 50%-90%, for example, 80% sample in aspect of model database is as training sample
This, 20% sample is as test sample.The model training for the neural network model that training sample is determined for step 3), to
Obtain the weights and amount of bias of neural network model.
5) neural network model is tested
If the input of test sample is xm, export as ym, according to the Artificial Neural Network Structures and step 4) of step 3) institute
The weights and amount of bias for obtaining neural network model are input with xm, calculate the output ya of neural network model.Enable err=| ym-
Ya | it is model bias.Preset ratio p% can be arranged within the scope of 1%-10%, if for example, model bias is more than 5 sample
Ratio is more than 5%, then return to step 3) 1 is added to the implicit number of plies or 1 is added to neuronal quantity, and step 4) and step are executed successively
It is rapid 5).If sample proportion of the model bias more than 5 is less than or equal to 5%, with the Artificial Neural Network Structures of step 3), step
4) the neural network model weights and amount of bias determined form box bearing temperature model.
Embodiment described above is only the preferred embodiments of the invention, and but not intended to limit the scope of the present invention, therefore
Change made by all shapes according to the present invention, principle, should all cover within the scope of the present invention.
Claims (1)
1. a kind of modeling method of gearbox of wind turbine bearing temperature, which is characterized in that include the following steps:
1) running of wind generating set data are obtained
Wind turbines whole year operation data, including set gear box oil temperature, oil pump are acquired using the data collecting system of wind power plant
Outlet pressure, environment temperature, unit generation power, wind speed, generator speed, gear case oil position data and box bearing temperature
Degree;
2) aspect of model database is established
First, gathered data is pre-processed, removes the unreasonable data of value range;Then gathered data is counted again
Amount calculates, and unit is calculated and be averaged generated output, mean wind speed, average generator speed and the gear case oil position that is averaged, and
By gear-box oil temperature, fuel pump outlet pressure, environment temperature, unit be averaged generated output, mean wind speed, average generator speed,
For average gear-box oil level as input data, box bearing temperature establishes aspect of model database as output data;
3) the feedforward neural network model structure for determining box bearing temperature, is specifically to determine the hidden layer of neural network model
The number of plies and neuronal quantity, determine activation primitive;
4) neural network model is trained
Using one group of inputoutput data as a sample, using a part of sample in aspect of model database as training sample
This, a part of sample is as test sample;The model training for the neural network model that training sample is determined for step 3), from
And obtain the weights and amount of bias of neural network model;
5) neural network model is tested
If the input of test sample is xm, export as ym, according to god obtained by the Artificial Neural Network Structures and step 4) of step 3)
Weights through network model and amount of bias are input with xm, calculate the output ya of neural network model;Enable err=| ym-ya | be
Model bias, if the sample proportion that model bias is more than preset value is more than preset ratio p%, return to step 3) redefine god
Through network architecture, and step 4) and step 5) are executed successively;If model bias be more than preset value sample proportion be less than or
Equal to preset ratio p%, then with the determining neural network model weights of the Artificial Neural Network Structures of step 3), step 4) and partially
The amount of setting forms box bearing temperature model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810364855.0A CN108629095A (en) | 2018-04-23 | 2018-04-23 | A kind of modeling method of gearbox of wind turbine bearing temperature |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810364855.0A CN108629095A (en) | 2018-04-23 | 2018-04-23 | A kind of modeling method of gearbox of wind turbine bearing temperature |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108629095A true CN108629095A (en) | 2018-10-09 |
Family
ID=63694192
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810364855.0A Pending CN108629095A (en) | 2018-04-23 | 2018-04-23 | A kind of modeling method of gearbox of wind turbine bearing temperature |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108629095A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110173472A (en) * | 2019-05-15 | 2019-08-27 | 明阳智慧能源集团股份公司 | The test bed hydraulic system of wind turbine gearbox flexible pin shaft and its control method |
CN110378042A (en) * | 2019-07-23 | 2019-10-25 | 山东大学 | Gearbox of wind turbine oil temperature method for detecting abnormality and system based on SCADA data |
CN110907170A (en) * | 2019-11-30 | 2020-03-24 | 华能如东八仙角海上风力发电有限责任公司 | Wind turbine generator gearbox bearing temperature state monitoring and fault diagnosis method |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101872165A (en) * | 2010-06-13 | 2010-10-27 | 西安交通大学 | Method for fault diagnosis of wind turbines on basis of genetic neural network |
CN103234753A (en) * | 2013-04-11 | 2013-08-07 | 华北电力大学 | Gaussian process modeling based wind turbine shafting state monitoring method |
CN103758696A (en) * | 2014-01-09 | 2014-04-30 | 华北电力大学 | SCADA (supervisory control and data acquisition) temperature parameter based wind turbine set security evaluation method |
CN105787561A (en) * | 2016-03-22 | 2016-07-20 | 新疆金风科技股份有限公司 | Recurrent neural network model construction method and gearbox fault detection method and device |
CN107392304A (en) * | 2017-08-04 | 2017-11-24 | 中国电力科学研究院 | A kind of Wind turbines disorder data recognition method and device |
CN107797063A (en) * | 2017-09-19 | 2018-03-13 | 北京瑞盈同创智能技术研究院有限公司 | Running of wind generating set state estimation and method for diagnosing faults based on SCADA |
-
2018
- 2018-04-23 CN CN201810364855.0A patent/CN108629095A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101872165A (en) * | 2010-06-13 | 2010-10-27 | 西安交通大学 | Method for fault diagnosis of wind turbines on basis of genetic neural network |
CN103234753A (en) * | 2013-04-11 | 2013-08-07 | 华北电力大学 | Gaussian process modeling based wind turbine shafting state monitoring method |
CN103758696A (en) * | 2014-01-09 | 2014-04-30 | 华北电力大学 | SCADA (supervisory control and data acquisition) temperature parameter based wind turbine set security evaluation method |
CN105787561A (en) * | 2016-03-22 | 2016-07-20 | 新疆金风科技股份有限公司 | Recurrent neural network model construction method and gearbox fault detection method and device |
CN107392304A (en) * | 2017-08-04 | 2017-11-24 | 中国电力科学研究院 | A kind of Wind turbines disorder data recognition method and device |
CN107797063A (en) * | 2017-09-19 | 2018-03-13 | 北京瑞盈同创智能技术研究院有限公司 | Running of wind generating set state estimation and method for diagnosing faults based on SCADA |
Non-Patent Citations (4)
Title |
---|
MEIK SCHLECHTINGEN ET AL.: "Comparative analysis of neural network and regression based condition monitoring approaches for wind turbine fault detection", 《MECHANICAL SYSTEMS AND SIGNAL PROCESSING》 * |
丁佳煜 等: "基于轴承温度模型的风电机组故障预测研究", 《可再生能源》 * |
刘旭: "数据挖掘在风力发电机组故障诊断中的应用", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
许克磊: "基于数据的风电机组非正常工况预测研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110173472A (en) * | 2019-05-15 | 2019-08-27 | 明阳智慧能源集团股份公司 | The test bed hydraulic system of wind turbine gearbox flexible pin shaft and its control method |
CN110173472B (en) * | 2019-05-15 | 2024-03-26 | 明阳智慧能源集团股份公司 | Hydraulic system of wind power gear box flexible pin shaft test bench and control method thereof |
CN110378042A (en) * | 2019-07-23 | 2019-10-25 | 山东大学 | Gearbox of wind turbine oil temperature method for detecting abnormality and system based on SCADA data |
CN110907170A (en) * | 2019-11-30 | 2020-03-24 | 华能如东八仙角海上风力发电有限责任公司 | Wind turbine generator gearbox bearing temperature state monitoring and fault diagnosis method |
CN110907170B (en) * | 2019-11-30 | 2021-03-16 | 华能如东八仙角海上风力发电有限责任公司 | Wind turbine generator gearbox bearing temperature state monitoring and fault diagnosis method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108629095A (en) | A kind of modeling method of gearbox of wind turbine bearing temperature | |
CA2764451C (en) | Available power estimator | |
CN108507787A (en) | Wind power gear speed increase box fault diagnostic test platform based on multi-feature fusion and method | |
CN108334672A (en) | Variable Speed Wind Power Generator maximal wind-energy capture method based on effective wind speed estimation | |
CN106368908A (en) | Wind turbine generator set power curve testing method based on SCADA (supervisory control and data acquisition) system | |
CN106991508A (en) | A kind of running of wind generating set state identification method based on DBSCAN | |
CN105673322B (en) | Realize the variable element Nonlinear Feedback Control Method of wind energy conversion system MPPT controls | |
CN106704103B (en) | Wind turbine generator power curve acquisition method based on blade parameter self-learning | |
CN106870298A (en) | Blade root bolt fracture fault detection method based on machine learning | |
CN108874733A (en) | A kind of large-scale half direct-drive unit health state evaluation method | |
CN103712792A (en) | Fault diagnosis method for wind-power gear case | |
Sun et al. | A robust approach of multi-sensor fusion for fault diagnosis using convolution neural network | |
CN111287911A (en) | Wind turbine fatigue load early warning method and system | |
CN103852255B (en) | Based on neutral net wind power generating set typical case's drive failures intelligent diagnosing method | |
CN111814848B (en) | Self-adaptive early warning strategy design method for temperature faults of wind turbine generator | |
CN107103167A (en) | For the DEH speed-regulating functions diagnostic method and system of removal of load operating mode | |
CN114033617B (en) | Controllable wind power generation method and system with control parameters adjusted in self-adaptive mode | |
KR102226971B1 (en) | Method for fault diagnosis based on multiple variables and apparatus using the method | |
CN109779844A (en) | The acquisition methods and system of angle of fan leaves measurement error influence relationship | |
CN113111314B (en) | Multi-unit cooperation based health state evaluation method for grid-connected wind turbine generator | |
CN114607571A (en) | Offshore wind power gear box fault identification method and system based on lubricating system monitoring | |
Obdam et al. | Flight leader concept for wind farm load counting and performance assessment | |
CN110334951B (en) | Intelligent evaluation method and system for high-temperature capacity reduction state of wind turbine generator | |
Yu et al. | The fault diagnosis based on self-organizing fuzzy clustering elman network in wind turbines | |
CN117521282B (en) | Density-dependent fan parameterization method for wind farm climate characteristic simulation |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181009 |
|
RJ01 | Rejection of invention patent application after publication |