CN109583075A - Permanent magnet direct-drive wind-force machine military service quality evaluating method based on temperature parameter prediction - Google Patents
Permanent magnet direct-drive wind-force machine military service quality evaluating method based on temperature parameter prediction Download PDFInfo
- Publication number
- CN109583075A CN109583075A CN201811413999.7A CN201811413999A CN109583075A CN 109583075 A CN109583075 A CN 109583075A CN 201811413999 A CN201811413999 A CN 201811413999A CN 109583075 A CN109583075 A CN 109583075A
- Authority
- CN
- China
- Prior art keywords
- temperature
- energy conversion
- conversion system
- wind
- wind energy
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000006243 chemical reaction Methods 0.000 claims abstract description 95
- 238000012549 training Methods 0.000 claims abstract description 22
- 238000011156 evaluation Methods 0.000 claims abstract description 15
- 238000004364 calculation method Methods 0.000 claims abstract description 13
- 238000012423 maintenance Methods 0.000 claims abstract description 12
- 238000005070 sampling Methods 0.000 claims abstract description 7
- 238000000714 time series forecasting Methods 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 description 8
- 238000012544 monitoring process Methods 0.000 description 8
- 230000003862 health status Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 241000932075 Priacanthus hamrur Species 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 238000013441 quality evaluation Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 230000010355 oscillation Effects 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 208000025274 Lightning injury Diseases 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000000505 pernicious effect Effects 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06395—Quality analysis or management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/80—Management or planning
- Y02P90/82—Energy audits or management systems therefor
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Educational Administration (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Geometry (AREA)
- Evolutionary Computation (AREA)
- Computer Hardware Design (AREA)
- Primary Health Care (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Wind Motors (AREA)
Abstract
The invention discloses a kind of permanent magnet direct-drive wind-force machine military service quality evaluating methods based on temperature parameter prediction, it establishes and variable, base bearing temperature, cabin temperature, the temperature parameter time series predicting model that wheel hub temperature is autoregressive prediction variable is an externally input with wind turbine hub revolving speed, extraneous wind speed, environment temperature, active power of output, vane propeller-changing angle；By there is the uniformly random sampling put back to obtain 5 training sample subsets, 5 temperature parameter time series predicting models of stand-alone training；5 models are integrated in such a way that prediction result takes mean value, establish temperature parameter integrated predictive model；Temperature parameter prediction error calculation wind energy conversion system temperature volume according to integrated predictive model conquers labour quality index and carries out Real-Time Evaluation to wind energy conversion system military service quality.The scientific maintenance and efficient operation that the present invention can be permanent magnet direct-drive wind-force machine under harsh environments provide crucial technical guarantee.
Description
Technical field
The present invention relates to complex electromechanical systems status monitorings and assessment technology field, especially a kind of pre- based on temperature parameter
The permanent magnet direct-drive wind-force machine military service quality evaluating method of survey.
Background technique
Wind generating technology is most mature as current technology, most scale exploit condition green energy resource utilizes skill
Art has vast potential for future development.Critical equipment-wind energy conversion system of wind generating technology is the Mechatronic Systems of one and its complexity,
Usually be seated exurbs remote, have inconvenient traffic, bad environments and coastal or greater coasting area, frost, low pressure,
The severe natural environment such as sand and dust, lightning stroke causes day-to-day operation state-detection difficult, and maintenance cost is expensive.Develop wind energy conversion system state
Monitoring and evaluation technology grasps wind energy conversion system health status and development trend, for optimizing unit maintenance project, saves O&M cost
Expenditure avoids pernicious safety and production accident from occurring, and the competitiveness for promoting Wind Power Generation Industry has great importance.
Installation data acquisition is that the monitoring wind energy conversion system that current wind power plant generallys use is real-time with monitoring control (SCAD) system
The measure of operating status, it is desirable to take this to reach the target for improving wind power plant operational safety and economy.The system monitoring parameter is many
It is more, including temperature, wind speed, vibration, voltage, electric current, yaw angle, motor control etc., due to without installing in the unit additionally
Data collection system, wind energy conversion system manufacturer and wind power plant owner etc. are intended to realize wind by analyzing the collected number of SCADA system
Power machine health state evaluation.
In recent years, domestic and international expert is opened using the wind energy conversion system failure predication and status assessment field of SCAD system state data
Extensive work is opened up.North China Electric Power University Guo Peng etc. uses nonlinear state Eq technology as modeling method, to wind turbine
On the basis of group tower oscillation characteristic and its influence factor carry out careful analysis, tower oscillation model is established, intends opening to be subsequent
The wind generating set vibration status monitoring and Incipient Fault Diagnosis of exhibition are laid a good foundation；University of Iowa Kusiak et al. is logical
Analysis SCADA data is crossed, information relevant to key components and parts failures such as Wind turbines bearing, motors has been excavated, constructs number
It is used to disclose the associate feature between the vibration of main shaft and pylon and running of wind generating set parameter according to model.Beijing Jiaotong University Wang Wei
Etc. establishing the wind energy conversion system pitch-controlled system Partial Linear Models based on SCAD system data, and proposed accordingly based on data mining
Paddle change system of wind turbines deterioration state on-line identification method.Compared with the signals such as electrical, vibration, temperature signal be it is a kind of more
For stable status signal, it is reliable in short term to have wind energy conversion system of scholar's trial exploration based on SCAD system temperature signal estimation model
Property appraisal procedure, but these researchs do not account for the shadow that changes to the historic state of wind energy conversion system long period scale to temperature parameter
It rings, the precision of temperature parameter prediction model and the accuracy of wind energy conversion system state evaluation can not ensure.
Above-mentioned fruitful work advances to a certain extent develops Wind turbines status monitoring and assessment technology
Level, but status assessment and failure are pre- in a certain shorter period under a certain component of most concern wind energy conversion system or operating status
It warns, is not able to satisfy the practical interior integrality (comprising operating status and shutdown status) of being on active service the period entirely to wind energy conversion system of concern engineering and comments
The technical need estimated；The specific object of these researchs is usually double-feed type wind machine, for same widely applied permanent magnet direct-drive
Wind energy conversion system is related to less；Theoretical research stage is also rested essentially within, there are also a distances with practical application.The prior art can not be
Wind energy conversion system, especially scientific maintenance of the permanent magnet direct-drive wind-force machine under harsh environments and efficient operation provide solid technology
Support.
Summary of the invention
The technical problem to be solved by the present invention is in view of the shortcomings of the prior art, provide a kind of based on temperature parameter prediction
Permanent magnet direct-drive wind-force machine military service quality evaluating method, realize wind energy conversion system service phase full working scope dynamical health state comment in real time
Estimate, provides crucial technical guarantee for scientific maintenance of the permanent magnet direct-drive wind-force machine under harsh environments and efficient operation.
In order to solve the above technical problems, the technical scheme adopted by the invention is that: the technical scheme adopted by the invention is that:
A kind of permanent magnet direct-drive wind-force machine military service quality evaluating method based on temperature parameter prediction, comprising the following steps:
1) it establishes temperature parameter time series predicting model: establishing with wind turbine hub revolving speed, extraneous wind speed, environment temperature
Degree, active power of output, vane propeller-changing angle are an externally input variable, and base bearing temperature, cabin temperature, wheel hub temperature are to return certainly
Return the temperature parameter time series predicting model of predictive variable.
2) more prediction models it is trained with it is integrated: read from the acquisition of the data of wind energy conversion system with monitoring in (SCADA) system of control
Wind energy conversion system to be evaluated stablizes the status data being on active service the 1st year as training sample set, by there is the uniformly random sampling put back to obtain
5 sample sets are taken, the sample size of each sample set is the 3/10 of sample set, is independently trained using sample set
5 temperature parameter time series predicting models；5 trained models are integrated in such a way that prediction result takes mean value,
Establish temperature parameter integrated predictive model.
3) temperature parameter prediction conquers labour quality index with wind energy conversion system temperature volume: it is current to read wind energy conversion system from SCADA system
State parameter online predicts current time predictive variable using temperature parameter integrated predictive model；It is absolute according to prediction
Error calculation wind energy conversion system temperature volume conquers labour quality index.
4) wind energy conversion system military service quality evaluation: labour quality indicator value is conquered to wind energy conversion system military service quality according to wind energy conversion system temperature volume
Carry out Real-Time Evaluation.
In step 1, temperature parameter time series predicting model framework is band external input Nonlinear Time Series autoregression
Neural network prediction model, the external input vector X (t) of model are as follows:
X (t)=[x (t)1 x(t)2 x(t)3 x(t)4 x(t)5 x(t)6 x(t)7]
Wherein, t is current time.x(t)1、x(t)2、x(t)3、x(t)4、x(t)5、x(t)6、x(t)7Respectively t moment wind
The hub rotation speed of power machine, extraneous wind speed, environment temperature, active power of output, 1 variable pitch angle of blade, 2 variable pitch angle of blade and leaf
3 variable pitch angle of piece.
Autoregressive prediction vector Y (t) are as follows:
Y (t)=[y (t)1 y(t)2 y(t)3]
Wherein, y (t)1、y(t)2、y(t)3The respectively base bearing temperature, cabin temperature and wheel hub temperature of t moment wind energy conversion system
Degree.
Temperature parameter time series predicting model PM [] may be expressed as:
Wherein, a is that model considers the order that influences on Current Temperatures parameter of historical state data, the value range of a for [2,
10]。z(t)1——z(t)s1For the input parameter of model, z (t)1=x (t)1,z(t)2=x (t)2..., z (t)7=x (t)7, z
(t)8=x (t-1)1..., z (t)7(a+1)=x (t-a)7, z (t)7(a+1)+1=y (t-1)1..., z (t)s1=y (t-a)3；Wherein,
s1=10 × a+7.The output of model is the predicted value of autoregressive prediction vector
Temperature parameter time series predicting model PM [] is by 1 input layer L1, 1 output layer L3With 1 with the hidden of delay
L containing layer2It constitutes, L1Node layer number is the number s for inputting parameter1, L3Node layer number is the number 3, L of output parameter2Layer section
Point number s2Calculation formula are as follows:
Wherein, INT () is bracket function, and c is constant, and the value range of c is [3,6].
Temperature parameter time series predicting model hidden layer L2I-th of node weighting input L2biAre as follows:
Wherein, vijFor the connection weight between j-th of input layer and i-th of hidden layer node, θiIt is implicit for i-th
The amount of bias of node layer.
Hidden layer L2I-th of node weighting export OL2biAre as follows:
Wherein, e is natural constant.
Output layer L3K-th of node weighting input L3bkAnd outputAre as follows:
Wherein, wkhFor the connection weight between h-th of hidden layer node and k-th of output node layer, γkIt is defeated for k-th
The amount of bias of node layer out；For the autoregressive prediction vector predictors of prediction model outputK-th of element,That is t moment wind-force main bearing temperature, the predicted value of cabin temperature and wheel hub temperature.
In step 2, the training method of temperature parameter time series predicting model is the calculation of Levenberg-Marquardt iteration
Method, stopping criterion for iteration are as follows:
Wherein,For the real output value of output node k after the r times training, y (t)krTo export section after the r times training
The predictive variable value of the acquisition of the desired output of point k, i.e. data and supervisor control actual measurement, reads from training sample；er
For maximum allowable mean square error, er value in section [0.03,0.07]；Ger is that effective error reduces speed, and Ger is in section
[5×10-8, 1.5 × 10-7] in value.
In step 3, wind energy conversion system temperature volume conquers labour quality index and counts health indicator and temperature sign rail by temperature sign
Mark health indicator two secondary indexs are constituted.The military service quality evaluation angle of temperature sign statistics health indicator are as follows: a period of time
The statistics matching degree of interior wind energy conversion system temperature change and prediction model；The military service quality evaluation of temperature sign track health indicator
Angle are as follows: current time wind energy conversion system temperature parameter and the wind energy conversion system temperature parameter development track of temperature prediction model prediction deviate journey
Degree；The two complements one another.
It is prediction error criterion percentile, the method for acquisition that temperature sign, which counts the reference data that health indicator calculates,
Are as follows: applying step 2 obtain temperature parameter integrated predictive model to whole training sample set autoregressive predictions described in step 2 to
Prediction absolute error is predicted and calculated to amount, obtains normative forecast error sample set；Statistics calculates normative forecast error sample
The percentile manifold pc of collection1,pc2,…,pc99, as prediction error criterion percentile.The temperature sign of t moment counts
Health indicator IStCalculation method are as follows:
CStAt the time of reference record stateful with supervisor control for data acquisition in hb hours before current time t
The set of composition, hb value in section [4,72], the state parameter before moment t record t moment when less than hb hours
State parameter the reference record lacked is supplied.aqntFor with nt moment temperature parameter integrated predictive model absolute error entMost
Close prediction error criterion percentile pcpiSerial number.nqntCalculation method are as follows:
Aqn is the nonsingular threshold value of deviation, aqn value in section [67,95].
T moment temperature sign track health indicator ITtCalculation method are as follows:
Wherein, etFor t moment temperature parameter integrated predictive model absolute prediction error, arctan () is arc tangent letter
Number.
The prediction absolute error e of t moment temperature parameter integrated predictive modeltCalculation method are as follows:
Wherein, y (t)kFor the measured value of k-th of element of regression forecasting vector Y (t), y (t)1、y(t)2、y(t)3Respectively
T moment wind-force main bearing temperature, cabin temperature and the wheel hub temperature of data acquisition and supervisor control record.
In step 4, military service quality is evaluated with IS according to wind energy conversion system temperature sign quality indicator valuetBased on index,
ITtIndex assists IStIndex judges military service quality, interpretational criteria are as follows:
1)ISt∈ (0.95,1] when, wind energy conversion system military service quality is outstanding；ISt∈ (0.85,0.95] when, wind energy conversion system military service quality
Well；ISt∈ (0.7,0.85] when, wind energy conversion system military service quality is slightly degenerated, and need to be aroused attention；ISt∈ (0.5,0.7] when, wind-force
Machine military service quality gently degraded, needs to pay close attention to；IStWhen [0,0.5] ∈, wind energy conversion system military service quality heavy is degenerated, and need to be closed closely
Infuse wind energy conversion system or scheduled overhaul maintenance.
2)ITt∈ (0.8,1] when, without paying special attention to wind energy conversion system；ITt∈ (0.6,0.8] when, wind energy conversion system state need to be paid close attention to；
ITtWhen [0,0.6] ∈, wind energy conversion system or scheduled overhaul maintenance need to be paid close attention to.
Compared with prior art, the advantageous effect of present invention is that: fully consider wind energy conversion system long period scale
The influence that historic state changes temperature parameter, it is externally input non-linear by establishing permanent magnet direct-drive wind-force machine temperature parameter band
Time series Recurrent neural network prediction model, the accurate changing rule for grasping health status apparatus for lower wind machine parameter, in turn
According to observed temperature parameter and health status apparatus for lower wind machine estimates of parameters difference, accurately commenting to wind energy conversion system military service quality is realized
Estimate, the scientific maintenance that can be permanent magnet direct-drive wind-force machine under harsh environments and efficient operation provide crucial technical guarantee；
It include the full working scope dynamical health state online evaluation of shutdown status in the achievable wind energy conversion system service phase of the present invention, rather than only office
It is limited to operating status, grasp wind energy conversion system health status that can be more complete.The training sample source of prediction model is wind-force to be evaluated
1 year status data after machine formally begins one's duty constructs sample using the uniformly random sampling for having playback without further screening
Subset, and the uncertain influence to precision of prediction that the integrated mode of multiple prediction models inhibits training sample is respectively trained,
It can guarantee the validity of temperature parameter prediction while simplified model instructs operation difficulty.The present invention passes through building wind energy conversion system temperature
Sign military service quality index, wind energy conversion system military service quality information contain in prediction absolute error data, not clear enough is accurate,
Clearly, clear presentation, reduces the working strength and difficulty of user.The present invention can also be applied to other type wind-force
The dynamical health state online evaluation of the complex electromechanical systems such as machine, bullet train.
Detailed description of the invention
Fig. 1 is one embodiment of the invention method flow diagram；
Fig. 2 is the temperature parameter time series predicting model structural schematic diagram of one embodiment of the invention；Wherein, 1, input
Layer；2, hidden layer；3, output layer；
Fig. 3 is the prediction error criterion percentile scatter plot of one embodiment of the invention；
Fig. 4 be according to the embodiment of the present invention wind energy conversion system military service second year temperature volume conquer labour quality index calculated result with
Disorderly closedown schematic diagram data；The a of Fig. 4 is prediction absolute error；The b of Fig. 4 is temperature sign statistics health indicator；The c of Fig. 4
It is temperature sign track health indicator；The d of Fig. 4 is wind energy conversion system disorderly closedown data；
Fig. 5 be according to the embodiment of the present invention wind energy conversion system military service third year temperature volume conquer labour quality index calculated result with
Disorderly closedown schematic diagram data；The a of Fig. 5 is prediction absolute error；The b of Fig. 5 is temperature sign statistics health indicator；The c of Fig. 5
It is temperature sign track health indicator；The d of Fig. 5 is wind energy conversion system disorderly closedown data.
Specific embodiment
As shown in Fig. 1, one embodiment of the invention detection method process is as follows:
First, it establishes with wind turbine hub revolving speed, extraneous wind speed, environment temperature, active power of output, vane propeller-changing angle
It is an externally input variable, base bearing temperature, cabin temperature, the temperature parameter time series that wheel hub temperature is autoregressive prediction variable
Prediction model.Model structure is referring to attached drawing 2, external input vector X (t) are as follows:
X (t)=[x (t)1 x(t)2 x(t)3 x(t)4 x(t)5 x(t)6 x(t)7]
Wherein, t is current time.x(t)1、x(t)2、x(t)3、x(t)4、x(t)5、x(t)6、x(t)7Respectively t moment wind
The hub rotation speed of power machine, extraneous wind speed, environment temperature, active power of output, 1 variable pitch angle of blade, 2 variable pitch angle of blade and leaf
3 variable pitch angle of piece.
Autoregressive prediction vector Y (t) are as follows:
Y (t)=[y (t)1 y(t)2 y(t)3]
Wherein, y (t)1、y(t)2、y(t)3The respectively base bearing temperature, cabin temperature and wheel hub temperature of t moment wind energy conversion system
Degree.
In the present embodiment, the sampling period of wind energy conversion system SCADA system is 10 minutes, 40 minutes historic state numbers before consideration
According to the influence to Current Temperatures parameter, therefore, order a value is 3, inputs the number s of parameter1=10 × 3+7=37, temperature ginseng
Number time series predicting model PM [] may be expressed as:
Wherein, z (t)1=x (t)1,z(t)2=x (t)2..., z (t)7=x (t)7, z (t)8=x (t-1)1..., z (t)28
=x (t-3)7, z (t)29=y (t-1)1..., z (t)37。
Model PM [] hidden layer L2Node number s2Calculation formula are as follows:
In the present embodiment, constant c value 4, hidden layer L2Node number is 10.Input, output layer node number difference
For 37 and 3.
After establishing temperature parameter time series forecasting mould, wind energy conversion system to be evaluated is read from wind energy conversion system SCADA system and is stablized
The status data being on active service the 1st year is as training sample set.The evaluation object of the present embodiment is certain of southern china mountain wind field
Platform permanent magnet direct-drive wind-force machine, wind energy conversion system completion in-site installation in 2012,2013 start to stablize military service.Training sample set is should
The 1 day January in 2013 of wind energy conversion system SCADA system record starts the status data until on December 31st, 2013.
The present embodiment to training sample set by there is the uniformly random sampling put back to obtain 5 sample sets, each sample
The sample size of subset is the 3/10 of sample set, independently trains 5 temperature parameter time serieses pre- using sample set
Survey model.The maximum allowable mean square error er value 0.05 of termination of training iteration, effective error reduce speed Ger value 1 × 10-7.5 trained models are integrated in such a way that prediction result takes mean value, establish temperature parameter integrated predictive model.
After the temperature parameter integrated predictive model of acquisition, it should autoregressive prediction of the model to wind energy conversion system training sample set
Prediction absolute error is predicted and calculated to vector, obtains normative forecast error sample set；Statistics calculates normative forecast error sample
The percentile manifold pc of this collection1,pc2,…,pc99, as prediction error criterion percentile.Obtained prediction error mark
Quasi- percentile referring to attached drawing 3, by Fig. 3 also it can be seen that, in normative forecast error sample set, 80% element numerical value is taken the photograph 0.12
Within family name's degree, the precision of prediction of temperature parameter integrated predictive model of the invention is high.
When the present embodiment calculating wind energy conversion system temperature volume conquers labour quality index, parameter hb value is 48 hours, the non-surprise of deviation
Different threshold value aqn is taken as 82.The present embodiment wind energy conversion system 2014 and temperature volume in 2015 conquer labour quality index calculated result with it is non-
Orderly closedown data are referring to attached drawing 4 and attached drawing 5.Wherein, a of a and Fig. 5 of Fig. 4 are respectively that prediction in 2014 and 2015 is exhausted
To error；The b of the b and Fig. 5 of Fig. 4 are respectively 2014 and 2015 temperature sign statistics health indicator calculated result；；Fig. 4
C and the c of Fig. 5 be respectively 2014 and 2015 temperature sign track health indicator results；The d of the d and Fig. 5 of Fig. 4 distinguish
For 2014 and 2015 wind energy conversion system disorderly closedown data.As seen from the figure, 4000- in 2014 of SCADA system record
There are long-time disorderly closedown, disorderly closedowns for 7000 argument sequences and 45000 to 50000 argument sequence wind energy conversion systems in 2015
When will occur and occur, temperature sign counts health indicator IStNumerical value is substantially reduced, and most of moment is located at the present invention and is on active service
The military service quality gently degraded of evaluating criterion of quality and heavy-degraded section；Temperature sign track health indicator ITtAt these
Wind energy conversion system state need to be paid close attention to and need to pay close attention to wind energy conversion system or scheduled overhaul maintenance section by being also predominantly located at quarter.Wind energy conversion system state
Normal moment (shutdown without exception occurs or will occur), IStAnd ITtNumerical value is higher, be predominantly located in military service quality it is good,
Military service quality slightly degenerates or without paying special attention to wind energy conversion system section.The method of the present invention can effectively to wind energy conversion system military service quality into
Row evaluation.The a of a, Fig. 5 of the c and Fig. 4 of c, Fig. 5 of b, Fig. 4 of b, Fig. 5 of comparison diagram 4 are it is found that the wind energy conversion system that the present invention constructs
Temperature volume conquers labour quality index, especially temperature sign statistics health indicator can will predict it is containing in absolute error data,
Not enough accurate, clear, the clear presentation of clear wind energy conversion system military service quality information；Temperature sign track health indicator can be used as
The supplement auxiliary temperature sign statistics health indicator of meaning evaluates wind energy conversion system military service quality.
The present invention fully considers the influence that the historic state of wind energy conversion system long period scale changes temperature parameter, by building
The vertical externally input Nonlinear Time Series Recurrent neural network prediction model of permanent magnet direct-drive wind-force machine temperature parameter band, accurately
Grasp health status apparatus for lower wind machine parameter changing rule, and then join according to observed temperature parameter and health status apparatus for lower wind machine
Number estimated value difference, realizes the accurate evaluation to wind energy conversion system military service quality, can be permanent magnet direct-drive wind-force machine in harsh environments
Under scientific maintenance and efficient operation crucial technical guarantee is provided；The present invention can be achieved in wind energy conversion system service phase comprising shutting down shape
The full working scope dynamical health state online evaluation of state, rather than it is limited only to operating status, grasp wind energy conversion system that can be more complete is strong
Health state.The training sample source of prediction model is 1 year status data after wind energy conversion system to be evaluated formally begins one's duty, without into
The screening of one step constructs sample set using the uniformly random sampling for having playback, and the integrated side of multiple prediction models is respectively trained
Formula inhibits the uncertain influence to precision of prediction of training sample, can guarantee temperature while simplified model instructs operation difficulty
The validity of parameter prediction.The present invention conquers labour quality index by constructing wind energy conversion system temperature volume, will predict absolute error data
In contain, presentation that not clear enough wind energy conversion system military service quality information is accurate, clear, clear, reduce the work of user
Intensity and difficulty.The dynamic that the present invention can also be applied to the complex electromechanical systems such as other type wind energy conversion systems, bullet train is strong
Health state online evaluation.
Claims (10)
1. it is a kind of based on temperature parameter prediction permanent magnet direct-drive wind-force machine military service quality evaluating method, which is characterized in that including with
Lower step:
1) establishing with wind turbine hub revolving speed, extraneous wind speed, environment temperature, active power of output, vane propeller-changing angle is outside
Input variable, base bearing temperature, cabin temperature, the temperature parameter time series forecasting mould that wheel hub temperature is autoregressive prediction variable
Type；
2) read wind energy conversion system to be evaluated and stablize the status data be on active service the 1st year as training sample set, by have put back to it is uniform
Random sampling obtains 5 sample sets, and the sample size of each sample set is the 3/10 of sample set, using sample set point
5 temperature parameter time series predicting models of other stand-alone training；To 5 trained predictions in such a way that prediction result takes mean value
Model is integrated, and temperature parameter integrated predictive model is established；
3) read wind energy conversion system current state parameter, using temperature parameter integrated predictive model online to current time predictive variable into
Row prediction；Wind energy conversion system temperature volume, which is calculated, according to prediction absolute error conquers labour quality index；
4) labour quality indicator value is conquered according to wind energy conversion system temperature volume and Real-Time Evaluation is carried out to wind energy conversion system military service quality.
2. the permanent magnet direct-drive wind-force machine military service quality evaluating method according to claim 1 based on temperature parameter prediction,
It is characterized in that, the external input vector X (t) of the temperature parameter time series predicting model are as follows: X (t)=[x (t)1 x(t)2 x
(t)3 x(t)4 x(t)5 x(t)6 x(t)7]；Wherein, t is current time；x(t)1、x(t)2、x(t)3、x(t)4、x(t)5、x
(t)6、x(t)7The respectively hub rotation speed of t moment wind energy conversion system, extraneous wind speed, environment temperature, active power of output, 1 variable pitch of blade
3 variable pitch angle of angle, 2 variable pitch angle of blade and blade；
Autoregressive prediction variable Y (t) are as follows: Y (t)=[y (t)1 y(t)2 y(t)3]；Wherein, y (t)1、y(t)2、y(t)3Respectively
Base bearing temperature, cabin temperature and the wheel hub temperature of t moment wind energy conversion system.
3. the permanent magnet direct-drive wind-force machine military service quality evaluating method according to claim 2 based on temperature parameter prediction,
It is characterized in that, the temperature parameter time series predicting model PM [] indicates are as follows:
Wherein, a is the order that model considers that historical state data influences Current Temperatures parameter；z(t)1——For model
Input parameter, z (t)1=x (t)1,z(t)2=x (t)2..., z (t)7=x (t)7, z (t)8=x (t-1)1..., z (t)7(a+1)=x
(t-a)7, z (t)7(a+1)+1=y (t-1)1...,Wherein, s1=10 × a+7；The output of model is autoregression
The predicted value of predicted vectorThe value range of a is [2,10].
4. the permanent magnet direct-drive wind-force machine military service quality evaluating method according to claim 3 based on temperature parameter prediction,
It is characterized in that, the temperature parameter time series predicting model PM [] includes 1 input layer L1, 1 output layer L3With 1 band
The hidden layer L of delay2, L1Node layer number is the number s for inputting parameter1, L3Node layer number is the number 3, L of output parameter2
Node layer number s2Calculation formula are as follows:Wherein, INT () is bracket function, and c is normal
Number, the value range of c are [3,6]；The hidden layer L2I-th of node weighting input L2biAre as follows:Wherein, vijFor the connection weight between j-th of input layer and i-th of hidden layer node, θi
For the amount of bias of i-th of hidden layer node；Hidden layer L2I-th of node weighting export OL2biAre as follows:Its
In, e is natural constant；Output layer L3K-th of node weighting input L3bkAnd outputAre as follows:Wherein, wkhFor the company between h-th of hidden layer node and k-th of output node layer
Meet weight, γkThe amount of bias of node layer is exported for k-th；For the autoregressive prediction vector predictors of prediction model outputK-th of element,That is t moment wind-force main bearing temperature, cabin temperature and wheel hub temperature
Predicted value.
5. the permanent magnet direct-drive wind-force machine military service quality evaluating method according to claim 4 based on temperature parameter prediction,
It being characterized in that, the training method of the temperature parameter time series predicting model is Levenberg-Marquardt iterative algorithm,
Stopping criterion for iteration are as follows:
Wherein,For the real output value of output node k after the r times training, y (t)krFor output node k after the r times training
Desired output；Er is maximum allowable mean square error, er value in section [0.03,0.07]；Ger is that effective error reduces speed
Degree, Ger is in section [5 × 10-8, 1.5 × 10-7] in value.
6. the permanent magnet direct-drive wind-force machine military service quality evaluating method according to claim 1 based on temperature parameter prediction,
It is characterized in that, it includes temperature sign statistics health indicator and temperature sign track that the wind energy conversion system temperature volume, which conquers labour quality index,
Health indicator.
7. the permanent magnet direct-drive wind-force machine military service quality evaluating method according to claim 1 based on temperature parameter prediction,
It is characterized in that, t moment temperature sign counts health indicator IStCalculation method are as follows:
CStThe collection formed at the time of reference record stateful with supervisor control for data acquisition in hb hours before current time t
It closes, hb value in section [4,72], the state parameter before moment t records the state parameter of t moment when less than hb hours
The reference record lacked is supplied；aqntFor with nt moment temperature parameter integrated predictive model absolute error entImmediate prediction
Error criterion percentile pcpiSerial number,Aqn is the nonsingular threshold value of deviation, and aqn is in section
Value in [67,95].
8. the permanent magnet direct-drive wind-force machine military service quality evaluating method according to claim 1 based on temperature parameter prediction,
It is characterized in that, temperature sign track health indicator IT described in t momenttCalculation method are as follows:Wherein,
etFor t moment temperature parameter integrated predictive model absolute prediction error, arctan () is arctan function.
9. the permanent magnet direct-drive wind-force machine military service quality evaluating method according to claim 1 based on temperature parameter prediction,
It is characterized in that, the prediction absolute error e of temperature parameter integrated predictive model described in t momenttCalculation method are as follows:Wherein, y (t)kFor the measured value of k-th of element of regression forecasting vector Y (t), y (t)1、y(t)2、y
(t)3Respectively t moment wind-force main bearing temperature, cabin temperature and the wheel hub temperature of data acquisition and supervisor control record
Degree.
10. the permanent magnet direct-drive wind-force machine military service quality evaluating method according to claim 1 based on temperature parameter prediction,
It is characterized in that, it is quasi- to conquer the evaluation that labour quality indicator value carries out Real-Time Evaluation to wind energy conversion system military service quality according to wind energy conversion system temperature volume
Then are as follows:
1)ISt∈ (0.95,1] when, wind energy conversion system military service quality is outstanding；ISt∈ (0.85,0.95] when, wind energy conversion system military service quality is good
It is good；ISt∈ (0.7,0.85] when, wind energy conversion system military service quality is slightly degenerated, and need to be aroused attention；ISt∈ (0.5,0.7] when, wind energy conversion system
Military service quality gently degraded, needs to pay close attention to；IStWhen [0,0.5] ∈, wind energy conversion system military service quality heavy is degenerated, and needs to pay close attention to
Wind energy conversion system or scheduled overhaul maintenance；
2)ITt∈ (0.8,1] when, without paying special attention to wind energy conversion system；ITt∈ (0.6,0.8] when, wind energy conversion system state need to be paid close attention to；ITt
When [0,0.6] ∈, wind energy conversion system or scheduled overhaul maintenance need to be paid close attention to.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811413999.7A CN109583075B (en) | 2018-11-26 | 2018-11-26 | Permanent magnet direct-drive wind turbine service quality evaluation method based on temperature parameter prediction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811413999.7A CN109583075B (en) | 2018-11-26 | 2018-11-26 | Permanent magnet direct-drive wind turbine service quality evaluation method based on temperature parameter prediction |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109583075A true CN109583075A (en) | 2019-04-05 |
| CN109583075B CN109583075B (en) | 2022-12-02 |
Family
ID=65924430
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811413999.7A Active CN109583075B (en) | 2018-11-26 | 2018-11-26 | Permanent magnet direct-drive wind turbine service quality evaluation method based on temperature parameter prediction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109583075B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110544003A (en) * | 2019-07-18 | 2019-12-06 | 湖南科技大学 | temperature prediction-based wind power plant wind turbine generator set frequency converter state evaluation method |
| CN111080039A (en) * | 2020-03-17 | 2020-04-28 | 浙江上风高科专风实业有限公司 | Fan cluster fault prediction method and system |
| CN111553032A (en) * | 2020-04-27 | 2020-08-18 | 湖北文理学院 | Blisk milling temperature prediction method, blisk milling temperature prediction device, blisk milling temperature prediction equipment and storage medium |
| CN116611741A (en) * | 2023-07-14 | 2023-08-18 | 湖南省计量检测研究院 | Construction method and system of service quality index system based on wind power equipment |
| CN117111661A (en) * | 2023-08-31 | 2023-11-24 | 杭州泰龙净化设备工程有限公司 | Centralized control system and method for production workshops |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106845826A (en) * | 2017-01-18 | 2017-06-13 | 西安交通大学 | A kind of cold continuous rolling production line military service quality state appraisal procedure based on PCA Cpk |
| JP2018036196A (en) * | 2016-09-01 | 2018-03-08 | 富士電機株式会社 | Wind speed prediction device, wind speed prediction system, wind speed predication method and program |
| CN108680358A (en) * | 2018-03-23 | 2018-10-19 | 河海大学 | A kind of Wind turbines failure prediction method based on bearing temperature model |
-
2018
- 2018-11-26 CN CN201811413999.7A patent/CN109583075B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2018036196A (en) * | 2016-09-01 | 2018-03-08 | 富士電機株式会社 | Wind speed prediction device, wind speed prediction system, wind speed predication method and program |
| CN106845826A (en) * | 2017-01-18 | 2017-06-13 | 西安交通大学 | A kind of cold continuous rolling production line military service quality state appraisal procedure based on PCA Cpk |
| CN108680358A (en) * | 2018-03-23 | 2018-10-19 | 河海大学 | A kind of Wind turbines failure prediction method based on bearing temperature model |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110544003A (en) * | 2019-07-18 | 2019-12-06 | 湖南科技大学 | temperature prediction-based wind power plant wind turbine generator set frequency converter state evaluation method |
| CN110544003B (en) * | 2019-07-18 | 2023-09-15 | 湖南科技大学 | Wind turbine generator system frequency converter state evaluation method based on temperature prediction |
| CN111080039A (en) * | 2020-03-17 | 2020-04-28 | 浙江上风高科专风实业有限公司 | Fan cluster fault prediction method and system |
| CN111080039B (en) * | 2020-03-17 | 2023-10-20 | 浙江上风高科专风实业有限公司 | Fan cluster fault prediction method and system |
| CN111553032A (en) * | 2020-04-27 | 2020-08-18 | 湖北文理学院 | Blisk milling temperature prediction method, blisk milling temperature prediction device, blisk milling temperature prediction equipment and storage medium |
| CN111553032B (en) * | 2020-04-27 | 2023-09-15 | 湖北文理学院 | Blisk milling temperature prediction method, device, equipment and storage medium |
| CN116611741A (en) * | 2023-07-14 | 2023-08-18 | 湖南省计量检测研究院 | Construction method and system of service quality index system based on wind power equipment |
| CN117111661A (en) * | 2023-08-31 | 2023-11-24 | 杭州泰龙净化设备工程有限公司 | Centralized control system and method for production workshops |
| CN117111661B (en) * | 2023-08-31 | 2024-05-24 | 杭州泰龙净化设备工程有限公司 | Centralized control system and method for production workshops |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109583075B (en) | 2022-12-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109583075A (en) | Permanent magnet direct-drive wind-force machine military service quality evaluating method based on temperature parameter prediction | |
| Terciyanlı et al. | Enhanced nationwide wind-electric power monitoring and forecast system | |
| CN104131950B (en) | Partitioning determination method for threshold value of temperature characteristic quantity of wind generating set | |
| CN108388962B (en) | Wind power prediction system and method | |
| CN105809287A (en) | High-voltage transmission line icing process integrated prediction method | |
| CN105894391A (en) | Wind turbine torque control performance evaluation method based on SCADA operation data extraction | |
| CN104573876A (en) | Wind power plant short-period wind speed prediction method based on time sequence long memory model | |
| CN110378504B (en) | Photovoltaic power generation climbing event probability prediction method based on high-dimensional Copula technology | |
| CN110544003B (en) | Wind turbine generator system frequency converter state evaluation method based on temperature prediction | |
| CN111878322B (en) | Wind power generator device | |
| CN111814849B (en) | DA-RNN-based wind turbine generator set key component fault early warning method | |
| CN114033617B (en) | Controllable wind power generation method and system with control parameters adjusted in self-adaptive mode | |
| CN111814848A (en) | Self-adaptive early warning strategy design method for temperature fault of wind turbine generator | |
| CN111396266A (en) | GBRT-based wind turbine generator bearing fault early warning method | |
| CN208431106U (en) | Wind turbines yaw system monitors mechanism on-line | |
| CN112039051A (en) | Real-time modeling method for accessing double-fed wind driven generator into substation bus load | |
| CN103065049A (en) | Real-time prediction and calculation method of wind power and based on coif5 small wave real-time decomposition | |
| Spiridonakos et al. | Wind turbines structural identification framework for the representation of both short-and long-term variability | |
| CN111794921B (en) | Onshore wind turbine blade icing diagnosis method based on migration component analysis | |
| CN114048767A (en) | Fault monitoring and early warning method for wind power master control system | |
| CN112668807A (en) | Wind speed prediction method for wind power plant power prediction | |
| Hsu et al. | Predicting internal energy consumption of a wind turbine using semi-supervised deep learning | |
| CN112749820A (en) | Wind power prediction method and system | |
| CN114021992B (en) | Evaluation method for real-time output performance of large wind turbine generator | |
| CN113946608B (en) | Wind generating set running state monitoring method based on online RBF neural network |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |