CN105760617A - Calculation method applied to multi-parameter fault prediction and judgment indexes of wind generating set - Google Patents

Calculation method applied to multi-parameter fault prediction and judgment indexes of wind generating set Download PDF

Info

Publication number
CN105760617A
CN105760617A CN201610129865.7A CN201610129865A CN105760617A CN 105760617 A CN105760617 A CN 105760617A CN 201610129865 A CN201610129865 A CN 201610129865A CN 105760617 A CN105760617 A CN 105760617A
Authority
CN
China
Prior art keywords
fault
vector
wind
data
power generating
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
Application number
CN201610129865.7A
Other languages
Chinese (zh)
Inventor
姚万业
刘敬智
杨金彭
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
North China Electric Power University
Original Assignee
North China Electric Power University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by North China Electric Power University filed Critical North China Electric Power University
Priority to CN201610129865.7A priority Critical patent/CN105760617A/en
Publication of CN105760617A publication Critical patent/CN105760617A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00Computer-aided design [CAD]
    • G06F30/30Circuit design
    • G06F30/36Circuit design at the analogue level
    • G06F30/367Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/06Electricity, gas or water supply
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F2117/00Details relating to the type or aim of the circuit design
    • G06F2117/02Fault tolerance, e.g. for transient fault suppression
    • YGENERAL 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
    • Y04INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04SSYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Abstract

The invention relates to a calculation method applied to a multi-parameter fault prediction and judgment indexes of a wind generating set, belonging to the technical field wind power generation. The method comprises the following steps: (1) collecting and processing of historical operation data of the wind generating set; (2) carrying out fault characteristic selection of the wind generating set, namely carrying out characteristic selection on faults of key components of the wind generating set, and reasonably selecting a fault characteristic quantity according to the weights of various related parameters; (3) modeling by a nonlinear method, namely carrying out modeling by the nonlinear method and determining a fault prediction model; and (4) carrying out combination of judgment indexes and fault prediction. According to the method provided by the invention, multiple characteristic parameters are combined again according to the linear proportion of the weights to obtain the fault prediction and comprehensive judgment indexes, so that the fault prediction accuracy is improved.

Description

A kind of computational methods of the many reference amounts failure predication Judging index being applied to wind power generating set
Technical field
The present invention relates to wind power generating set failure prediction method, particularly to the computational methods of a kind of many reference amounts failure predication Judging index being applied to wind power generating set.Belong to technical field of wind power generation.
Background technology
Wind-power electricity generation is as a kind of emerging clean energy resource generation mode, and technology is not yet ripe, Wind turbines Frequent Troubles, has a strong impact on its safe and reliable operation.
In order to find the incipient fault hidden danger of unit in time, formulating rational Maintenance and Repair plan, Wind turbines failure predication technology has progressed into the research field of scholar.
Owing to the operating mode of Wind turbines is complicated, undulatory property is strong, is difficult to carry out approximate description by specific functional form in modeling process, therefore, generally uses nonlinear mapping method to be trained model reaching corresponding model accuracy.
In failure predication, nonlinear mapping method (BP network, support vector machine etc.) is utilized to set up the forecast model of Wind turbines particular elements fault, model is trained by the history data first using each operating mode, make model cover the normal working space of gear-box, then analyze the variation tendency of output variable corresponding to real-time running data.When Wind turbines particular elements misoperation, the dynamic characteristic of output variable necessarily deviates compared to normal working space, the residual distribution characteristic of output variable value and real variable value necessarily changes, if the confidence interval of the average of residual error or standard deviation exceeds threshold value set in advance, then giving a warning or warning message, prompting operations staff checks unit operation situation.This is the basic ideas of Wind turbines failure predication technology.
But, close coupling, the nonlinear system that electromagnetism, machinery, control, communication etc. are integrated is melted owing to Wind turbines is one, any one locally tiny fault all can be undertaken propagating, spread, accumulate and amplifying by system, operational factor is various simultaneously, dependency between parameter is strong, and single parameter may not necessarily accurately reflect the fault type of unit.
At present, prior art has proposed multiparameter failure predication strategy, yet with the weight of multiple parameters not being taken into account so that model still haves much room for improvement in prediction accuracy.
By consulting literatures, the specific fault of Wind turbines is carried out feature selection by existing approach application relief algorithm (or other feature selection approach), output variable using fault characteristic value as least square method supporting vector machine model, thus being predicted specific fault.When calculating failure predication Judging index, owing to being multiparameter failure predication, it is therefore necessary to each Failure Characteristic Parameter is carried out efficient combination, thus obtaining combination Judging index.The method, using multiple fault characteristic value as output observation vector a, judges set state by calculating the distance of itself and actual measurement vector b.Formula is as follows:
δ = Σ i = 1 n - ( a i - b i ) 2 - - - ( 1 )
Wherein, δ is the distance of two vector a and b of definition, and n is output observation vector dimension, aiAnd biThe respectively value of output observation vector and actual measurement vector i-th dimension variable.
Owing to the weight of multiple fault characteristic value is not taken into account, it is defaulted as the contribution margin that each characteristic quantity adjusts the distance identical, runs counter to the fact so that model prediction accuracy is relatively low.
Number of patent application is 201310259857.0, it is a kind of based on normalized Wind turbines fault early warning method that name is called that the Chinese invention patent of " based on normalized Wind turbines fault early warning method " discloses, by setting up Wind turbines fault model, the fault that Wind turbines is occurred carries out grade evaluation and test, and carries out early warning according to fault level.Specifically it is expressed as the running parameter rule of thumb selecting each fault of Wind turbines and is normalized, running status historical data alarm threshold value previously according to Wind turbines arranges malfunction grade threshold, weight then according to the variable quantity statistics gained of each running parameter affecting malfunction grade threshold, determines fault level by following formula:
Although this kind of method takes weight into account, model has many uncertainties: first running parameter and threshold value are all rule of thumb gained, it does not have the standard determined, there will be deviation unavoidably;It addition, the computational methods of weight are based primarily upon the running parameter variable quantity when fault occurs, owing to Wind turbines parts are various, 26S Proteasome Structure and Function is complicated, and the change of parameter exists time delay, and delay time differs so that weight calculation exists bigger error;Meanwhile, this kind of weighing computation method lacks rigorous theoretical basis, thus affecting the precision of fault pre-alarming model.
Therefore it provides a kind of precision that can improve wind power generating set anomalous identification, accurately determining fault type, the computational methods reducing the many reference amounts failure predication Judging index of erroneous judgement probability just become the technical barrier that this technical field is badly in need of solving.
Summary of the invention
It is an object of the invention to provide a kind of precision that can improve wind power generating set anomalous identification, accurately determine fault type, reduce the computational methods of the many reference amounts failure predication Judging index of erroneous judgement probability.
The above-mentioned purpose of the present invention reaches by the following technical programs:
A kind of computational methods of the many reference amounts failure predication Judging index being applied to wind power generating set, its step is as follows:
(1) collection of wind power generating set history data and process
Gather the history data of wind power generating set normal condition, the data gathered are screened and are normalized;
(2) wind driven generator set failure feature selects
The fault of wind power generating set critical component is carried out feature selection, selects fault characteristic value according to the weight size reasonable of each associated arguments;
(3) nonlinear method modeling
Model is set up with nonlinear method, select suitable input vector and output vector, input vector should meet all service conditions or cover all operating conditions, output vector is original assessment target faults characteristic quantity or other corresponding indexs therefrom, model is trained, establishes failure predication model;
(4) combination Judging index and failure predication: using multiple fault characteristic value as output observation vector a, judges set state by calculating the distance of itself and actual measurement vector b, adds fault characteristic value weight coefficient when computed range simultaneously;
Formula is as follows:
δ = Σ i = 1 n ( c i Σ i = 1 n c i ( a i - b i ) 2 ) - - - ( 3 )
Wherein, δ is the distance of two vector a and b of definition, and n is output observation vector dimension, aiAnd biThe respectively value of output observation vector and actual measurement vector i-th dimension variable, ciStandardized weight for i-th Failure Characteristic Parameter;
After combination Judging index is formulated, choose a large amount of normal history data covering all operating conditions, by input variable input model, output exports observation vector accordingly, calculate the distance δ of output observation vector and actual measurement vector, ask for the maximum MAX of δ, it is determined that its span is [0, MAX], according to statistics rule, using maximum MAX as prediction threshold value, by the corresponding input variable input model of real time data, export distance δ, if distance δ is less than MAX, then unit is properly functioning, otherwise, there is potential faults in unit.
Preferably, described step (1) is specific as follows:
1) collection of history data:
Gathering the history data of wind power generating set normal condition, the data gathered are screened, the data of several situations are rejected below: (A) Wind turbines disorderly closedown;(B) Wind turbines is safeguarded and is shut down;(C) wind speed is below incision wind speed, and Wind turbines is not grid-connected;(D) wind speed exceeds cut-out wind speed, blower fan off-grid;(E) Wind turbines open machine process and open machine after a period of time, at this moment likely gearbox temperature is too low, and blower fan is in limit power rating automatically;(F) Wind turbines stopping process;(G) Wind turbines people is limited power rating.It addition, the properly functioning data all operating conditions of uniform fold unit;
2) process of history data:
By step 1) the data obtained is normalized, and the scope of all operational factors is limited between [0,1], adopts linear transfor function to be normalized, and formula is as follows:
y = x - x m i n x m a x - x min - - - ( 4 )
Wherein, y is the value of consult volume after standardization, and x is the value of consult volume before standardization, xmaxFor the max-thresholds that parameter sets in SCADA system, xminIt is 0;
After normalized, obtain affecting the standardized value of the variable parameter that certain critical component runs.
Beneficial effect:
The present invention proposes the computational methods of a kind of many reference amounts failure predication Judging index being applied to wind power generating set, according to the linear accounting of weight, multiple characteristic parameters is reconfigured, obtains failure predication Comprehensive Evaluation index, improve the degree of accuracy of failure predication.
Below by the drawings and specific embodiments, the present invention is described in detail.It should be understood that described embodiment only relates to the preferred embodiments of the invention, without departing from the spirit and scope of the present invention situation, selection and the process of various parameters are all possible.
Accompanying drawing explanation
Fig. 1 is 1# pitch position sensor fault LSSVM network structure in the embodiment of the present invention 1.
Fig. 2 is the embodiment of the present invention 1 failure predication flow chart.
Fig. 3 is that the present invention proposes method and existing method failure predication control procedure chart.
Detailed description of the invention
Embodiment 1
A kind of computational methods of the many reference amounts failure predication Judging index being applied to wind power generating set, its step is as follows:
(1) history data collection and process
1) history data collection:
Gathered the history data of a certain wind power generating set normal condition of certain wind power plant by SCADA system, the data gathered are screened.The data of several situations should give rejecting below: (A) Wind turbines disorderly closedown;(B) Wind turbines is safeguarded and is shut down;(C) wind speed is below incision wind speed, and Wind turbines is not grid-connected;(D) wind speed exceeds cut-out wind speed, blower fan off-grid;(E) Wind turbines open machine process and open machine after a period of time, at this moment likely gearbox temperature is too low, and blower fan is in limit power rating automatically;(F) Wind turbines stopping process;(G) Wind turbines people is limited power rating.It addition, properly functioning data answer all operating conditions of uniform fold unit.
2) history data acquisition process: in data process, because the order of magnitude difference of each operational factor is bigger, for Accurate Prediction corresponding failure, needs are normalized, the scope of all operational factors is limited to [0,1] between, adopting linear transfor function to be normalized, formula is as follows:
y = x - x m i n x m a x - x m i n - - - ( 4 )
Wherein, y is the value of consult volume after standardization, and x is the value of consult volume before standardization, xmaxFor the max-thresholds that parameter sets in SCADA system, xminIt is 0;
(2) wind driven generator set failure feature selects
The fault of wind power generating set critical component is carried out feature selection, judge whether unit exists potential faults by analyzing the variation tendency of fault characteristic value, the method that fault signature selects is not limited to, according to the present invention, as long as the parameters quantitative values to specific fault sensitivity can be obtained, it is possible to the synthetic determination index calculating method of the application present invention.
A kind of conventional fault signature system of selection is set forth below so that the practical application of the inventive method is expanded on further.
Below the specific fault of the progressive deterioration of a certain wind power generating set is carried out feature selection, for instance the 1# pitch position sensor fault in pitch-controlled system, uses the calculating of relief algorithm to obtain result as follows:
Table 11# pitch position sensor fault feature selection weight
Fault characteristic value is selected according to the weight size reasonable of each associated arguments.Fault characteristic value is descending as can be seen from Table 1 is followed successively by 1# pitch motor temperature (DEG C), 2# pitch motor temperature (DEG C), 3# pitch motor temperature (DEG C), 1# change oar capacitance voltage (VDC) etc., wherein 1# pitch motor temperature (DEG C), 2# pitch motor temperature (DEG C), 3# pitch motor temperature (DEG C), 1# become oar capacitance voltage (VDC) standardized weight sum be 0.74352, illustrate these four operational factors to 1# pitch position sensor fault reflection sensitiveer.Affect classifier performance in order to avoid dimension is higher, the parameter weight after the oar capacitance voltage of 1# change simultaneously compares front four differences relatively big (at least one order of magnitude), and shared weight is only small, therefore can ignore.
Total principle that selects of fault signature is: selecting parameter to try one's best under few premise, be selected to the main parameters of faults type according to weight.
(3) nonlinear method modeling
Non-linear modeling method is a lot, enumerates a kind of conventional least square method supporting vector machine (LSSVM) Modeling Theory and illustrates.
LSSVM applies non-linear map very widely at present, is mainly reflected in classification, recurrence aspect.Failure predication is to realize on the basis returned.Initially set up normal LSSVM computing network: select suitable input vector and output vector, input vector should meet all service conditions or cover all operating conditions, output vector is original assessment target characteristic selected amount or other corresponding indexs therefrom, and model is trained.
Also according to above-mentioned fault special case, set up and train the 1# pitch position sensor fault forecast model of Wind turbines.Using wind speed, active power and wind speed round as input vector, select fault characteristic value 1# pitch motor temperature (DEG C) obtained, 2# pitch motor temperature (DEG C), 3# pitch motor temperature (DEG C) and 1# change oar capacitance voltage (VDC) as output vector using fault signature, setting up LSSVM model, network structure is shown in Fig. 1: 1# pitch position sensor fault LSSVM network structure.
(4) combination Judging index and failure predication
It is similar to existing method, owing to being multiparameter failure predication, it is therefore necessary to each Failure Characteristic Parameter is carried out efficient combination, thus obtaining combination Judging index.Using multiple fault characteristic value as output observation vector a, judge set state by calculating the distance of itself and actual measurement vector b, add fault characteristic value weight coefficient simultaneously when computed range.
Formula is as follows:
δ = Σ i = 1 n ( c i Σ i = 1 n c i ( a i - b i ) 2 ) - - - ( 3 )
Wherein, δ is output observation vector and the distance of actual observation vector, and n is output observation vector dimension, aiAnd biThe respectively value of output observation vector and actual measurement vector i-th dimension variable, ciStandardized weight for i-th Failure Characteristic Parameter.
After combination Judging index is formulated, choose a large amount of normal history data covering all operating conditions, by input variable input model, output exports observation vector accordingly, calculate the distance δ of output observation vector and actual measurement vector, ask for the maximum MAX of δ, it is determined that its span is [0, MAX], according to statistics rule, using maximum MAX as prediction threshold value, by the corresponding input variable input model of real time data, export distance δ, if distance δ is less than MAX, then unit is properly functioning, otherwise, there is potential faults in unit.Corresponding failure predication flow chart is as shown in Figure 2.
Model with least square method supporting vector machine (LSSVM), initially set up normal LSSVM computing network: (input vector should cover all operating conditions as much as possible to select suitable input vector and output vector, output vector is original assessment target characteristic selected amount or other corresponding indexs therefrom), utilize the historical data after normalized that sample is trained, obtain optimal parameter, set up LSSVM Early-warning Model;Then real-time running data is monitored, corresponding input vector is inputted above-mentioned model, it is thus achieved that model output valve;Distance finally by computation model output valve Yu actual measured value judges set state, add fault characteristic value weight coefficient when computed range simultaneously, if distance is less than the threshold value set, illustrate that it is in normal condition, if it exceeds the threshold value set, then belong to abnormality, alert.
Comparative examples
Gather relevant variable data before certain wind power plant a certain wind power generating set fault (the corresponding above-mentioned LSSVM model of fault).
Data message before the following is the fault of collection: fault type is 1# pitch position sensor fault, fault time is 10:08:49 on May 13rd, 2015, and data sampling period is 7 seconds.Data before putting 10:08:49 fault time are theoretic properly functioning data, and list is observed and without exception data.Choose the fault above-mentioned forecast model of first few minutes data application its state is evaluated and tested, it is judged that whether unit exists potential faults.Comparing with existing model, the checking present invention proposes the practical application effect of model simultaneously.
Same threshold calculations rule, owing to the δ computing formula of two kinds of methods is different, the threshold value MAX obtained is different, in order to compare more intuitively, threshold value is standardized, and formula is as follows:
δ i j ′ = δ i j MAX i i = 1 , 2 - - - ( 5 )
Wherein, i represents two kinds of methods (1 represents the present invention proposes method, and 2 represent existing method);Time point before j representing fault, δ 'ijRepresent the distance after i-th kind of method jth time point standardization;δijIt it is the distance before i-th kind of method jth time point standardization;MAXiIt it is the threshold value of i-th kind of method.
Choose front 33 the time point data of fault, namely from 10:04:42 to 10:08:42 on May 13rd, 2015.Obtain the present invention and propose method and existing method failure predication control procedure chart, as shown in Figure 3.
As shown in Figure 3, before fault occurs, the present invention proposes method and is found range at the 25th time point (10:07:45) and exceeded threshold value from δ, and hereafter time point distance δ become fluctuation ascendant trend, illustrate that now unit exists potential faults, it was predicted that fault type is 1# pitch position sensor fault.It is advanced by nearly one minute compared to now methodical 30th predicted time point (10:08:21), and distance δ-value is relatively bigger.Illustrate that the present invention proposes method and improves a lot in failure predication performance, timely and accurately specific fault is achieved prediction.
In order to improve the accuracy of Wind turbines failure predication, existing multiparameter failure predication is combined Judging index and proposes to improve by the present invention, obtains considering the computing formula of the combination Judging index of weight.This formula can be applicable to the prediction of progressive fault, as long as obtaining the fault characteristic value weight coefficient of corresponding failure.
The existing failure prediction method of Wind turbines is proposed to improve by the present invention, relatively low for existing method one-parameter precision of prediction, multiparameter does not consider the drawback of weight, propose the computational methods of the Multi-parameter Combined Tool Judging index considering weight, apply it to the precision that can improve unit anomalous identification in the failure predication of Wind turbines and accurately determine the type of fault, reducing the probability of erroneous judgement.Accurately identifying of abnormality is conducive to staff to make halt instruction or other effective measures in time, reduces the damage that unit is caused by misoperation, the Accurate Prediction of fault type, is conducive to the carrying out of follow-up specific aim trouble shooting maintenance work.

Claims (3)

1. being applied to computational methods for the many reference amounts failure predication Judging index of wind power generating set, its step is as follows:
(1) collection of wind power generating set history data and process
Gather the history data of wind power generating set normal condition, the data gathered are screened and are normalized;
(2) wind driven generator set failure feature selects
The fault of wind power generating set critical component is carried out feature selection, selects fault characteristic value according to the weight size reasonable of each associated arguments;
(3) nonlinear method modeling
Model is set up with nonlinear method, select suitable input vector and output observation vector, input vector should meet all service conditions or cover all operating conditions, output observation vector is original assessment target faults characteristic quantity or other corresponding indexs therefrom, model is trained, establishes failure predication model;
(4) combination Judging index and failure predication: using multiple fault characteristic value as exporting observation vector a, set state is judged by calculating the distance of itself and actual measurement vector b, add fault characteristic value weight coefficient when computed range, formula is as follows simultaneously:
δ = Σ i = 1 n ( c i Σ i = 1 n c i ( a i - b i ) 2 )
Wherein, δ is output observation vector and the distance of actual observation vector, and n is output observation vector dimension, aiAnd biThe respectively value of output observation vector and actual measurement vector i-th dimension variable, ciWeight for i-th Failure Characteristic Parameter;
After combination Judging index is formulated, choose a large amount of normal history data covering all operating conditions, by input variable input model, output exports observation vector accordingly, calculate the distance δ of output observation vector and actual measurement vector, ask for the maximum MAX of δ, it is determined that its span is [0, MAX], according to statistics rule, using maximum MAX as prediction threshold value, by the corresponding input variable input model of real time data, export distance δ, if distance δ is less than MAX, then unit is properly functioning, otherwise, there is potential faults in unit.
2. the computational methods of the many reference amounts failure predication Judging index being applied to wind power generating set according to claim 1, it is characterised in that: specifically comprising the following steps that of described step (1)
1) collection of history data:
Gathering the history data of wind power generating set normal condition, the data gathered are screened, the data of several situations are rejected below: (A) Wind turbines disorderly closedown;(B) Wind turbines is safeguarded and is shut down;(C) wind speed is below incision wind speed, and Wind turbines is not grid-connected;(D) wind speed exceeds cut-out wind speed, blower fan off-grid;(E) Wind turbines open machine process and open machine after a period of time;(F) Wind turbines stopping process;(G) Wind turbines people is limited power rating, it addition, the properly functioning data all operating conditions of uniform fold unit;
2) process of history data:
By step 1) the data obtained is normalized, and the scope of all operational factors is limited between [0,1], adopts linear transfor function to be normalized, and formula is as follows:
y = x - x m i n x m a x - x m i n
Wherein y is the value of consult volume after standardization, and x is the value of consult volume before standardization, xmaxFor the max-thresholds that parameter sets in SCADA system, xminIt is 0;
After normalized, obtain affecting the standardized value of the variable parameter that certain critical component runs.
3. the computational methods of the many reference amounts failure predication Judging index being applied to wind power generating set according to claim 2, it is characterised in that: nonlinear method described in described step (3) is least square method supporting vector machine method.
CN201610129865.7A 2016-03-07 2016-03-07 Calculation method applied to multi-parameter fault prediction and judgment indexes of wind generating set Pending CN105760617A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610129865.7A CN105760617A (en) 2016-03-07 2016-03-07 Calculation method applied to multi-parameter fault prediction and judgment indexes of wind generating set

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610129865.7A CN105760617A (en) 2016-03-07 2016-03-07 Calculation method applied to multi-parameter fault prediction and judgment indexes of wind generating set

Publications (1)

Publication Number Publication Date
CN105760617A true CN105760617A (en) 2016-07-13

Family

ID=56331803

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610129865.7A Pending CN105760617A (en) 2016-03-07 2016-03-07 Calculation method applied to multi-parameter fault prediction and judgment indexes of wind generating set

Country Status (1)

Country Link
CN (1) CN105760617A (en)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106249148A (en) * 2016-08-30 2016-12-21 重庆勤智科技有限公司 Generator carbon brush failure analysis methods
CN106407589A (en) * 2016-09-29 2017-02-15 北京岳能科技股份有限公司 Wind turbine state evaluation and prediction method and system
CN106777606A (en) * 2016-12-02 2017-05-31 上海电机学院 A kind of gearbox of wind turbine failure predication diagnosis algorithm
CN106951997A (en) * 2017-03-24 2017-07-14 新疆金风科技股份有限公司 The method and apparatus for predicting the failure of blower fan
CN106959652A (en) * 2017-05-08 2017-07-18 北京百度网讯科技有限公司 Intelligent control method and device
CN106991508A (en) * 2017-05-25 2017-07-28 华北电力大学 A kind of running of wind generating set state identification method based on DBSCAN
CN107655668A (en) * 2017-09-20 2018-02-02 上海振华重工(集团)股份有限公司 The acquisition method of the accident analysis data of harbour machinery
CN108932531A (en) * 2018-07-09 2018-12-04 朱卫列 Equipment state on-line checking algorithm based on model feature value
CN110349289A (en) * 2019-06-21 2019-10-18 东软集团股份有限公司 Failure prediction method, device, storage medium and electronic equipment
CN111287912A (en) * 2020-02-18 2020-06-16 广州特种承压设备检测研究院 Fault diagnosis method for variable pitch system of wind driven generator
CN112380782A (en) * 2020-12-07 2021-02-19 重庆忽米网络科技有限公司 Rotating equipment fault prediction method based on mixed indexes and neural network

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103344914A (en) * 2013-06-26 2013-10-09 中能电力科技开发有限公司 Wind turbine generation unit fault early warning method based on normalization
CN105134510A (en) * 2015-09-18 2015-12-09 北京中恒博瑞数字电力科技有限公司 State monitoring and failure diagnosis method for wind generating set variable pitch system

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103344914A (en) * 2013-06-26 2013-10-09 中能电力科技开发有限公司 Wind turbine generation unit fault early warning method based on normalization
CN105134510A (en) * 2015-09-18 2015-12-09 北京中恒博瑞数字电力科技有限公司 State monitoring and failure diagnosis method for wind generating set variable pitch system

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
李学伟: "基于数据挖掘的风电机组状态预测及变桨系统异常识别", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106249148B (en) * 2016-08-30 2019-07-05 重庆勤智科技有限公司 Generator carbon brush failure analysis methods
CN106249148A (en) * 2016-08-30 2016-12-21 重庆勤智科技有限公司 Generator carbon brush failure analysis methods
CN106407589A (en) * 2016-09-29 2017-02-15 北京岳能科技股份有限公司 Wind turbine state evaluation and prediction method and system
CN106407589B (en) * 2016-09-29 2020-01-21 北京岳能科技股份有限公司 Fan state evaluation and prediction method and system
CN106777606A (en) * 2016-12-02 2017-05-31 上海电机学院 A kind of gearbox of wind turbine failure predication diagnosis algorithm
CN106951997A (en) * 2017-03-24 2017-07-14 新疆金风科技股份有限公司 The method and apparatus for predicting the failure of blower fan
CN106951997B (en) * 2017-03-24 2020-11-10 新疆金风科技股份有限公司 Method and device for predicting fault of fan
CN106959652A (en) * 2017-05-08 2017-07-18 北京百度网讯科技有限公司 Intelligent control method and device
CN106959652B (en) * 2017-05-08 2019-08-16 北京百度网讯科技有限公司 Intelligent control method, device and readable computer storage medium
CN106991508A (en) * 2017-05-25 2017-07-28 华北电力大学 A kind of running of wind generating set state identification method based on DBSCAN
CN107655668B (en) * 2017-09-20 2019-10-29 上海振华重工(集团)股份有限公司 The acquisition method of the accident analysis data of harbour machinery
CN107655668A (en) * 2017-09-20 2018-02-02 上海振华重工(集团)股份有限公司 The acquisition method of the accident analysis data of harbour machinery
CN108932531A (en) * 2018-07-09 2018-12-04 朱卫列 Equipment state on-line checking algorithm based on model feature value
CN110349289A (en) * 2019-06-21 2019-10-18 东软集团股份有限公司 Failure prediction method, device, storage medium and electronic equipment
CN111287912A (en) * 2020-02-18 2020-06-16 广州特种承压设备检测研究院 Fault diagnosis method for variable pitch system of wind driven generator
CN112380782A (en) * 2020-12-07 2021-02-19 重庆忽米网络科技有限公司 Rotating equipment fault prediction method based on mixed indexes and neural network

Similar Documents

Publication Publication Date Title
CN105760617A (en) Calculation method applied to multi-parameter fault prediction and judgment indexes of wind generating set
CN105846780A (en) Decision tree model-based photovoltaic assembly fault diagnosis method
CN102934312B (en) Energy production system and control thereof
CN104483575A (en) Self-adaptive load event detection method for noninvasive power monitoring
CN106777984A (en) A kind of method that photovoltaic array Working state analysis and fault diagnosis are realized based on density clustering algorithm
CN107742053A (en) Wind turbines abnormality recognition method and device
CN104732276A (en) On-line diagnosing method for faults of metering production facility
CN103758696A (en) SCADA (supervisory control and data acquisition) temperature parameter based wind turbine set security evaluation method
CN108154291B (en) Power supply plant performance index evaluation method and system based on strategy library
CN105205569A (en) Draught fan gear box state on-line evaluation model building method and on-line evaluation method
CN103912448B (en) A kind of regional wind power power of the assembling unit characteristic monitoring method
CN103278771B (en) Identification method for abnormal state of three-phase asynchronous motor at industrial site
CN104952000A (en) Wind turbine operating state fuzzy synthetic evaluation method based on Markov chain
CN104807644A (en) Fault early warning method and system for wind generation set variable-pitch system
CN103711645B (en) Based on the wind power generating set state evaluating method of modeling parameters signature analysis
CN105508149B (en) Fault detection method and device for wind power generating set
CN102062832B (en) Power system online prewarning method based on microdisturbance signal low-frequency oscillation mode identification
CN102289731A (en) Method for maintaining state of power transmission equipment based on system risk
CN108133311A (en) A kind of Wind turbines fault mode risk assessment and analysis method for reliability
CN103925155A (en) Self-adaptive detection method for abnormal wind turbine output power
CN106815771A (en) A kind of long-term evaluation method of wind power plant load
CN110926809B (en) Big data analysis-based wind turbine generator transmission chain fault early warning method
CN104819107B (en) The diagnostic method and system of a kind of Wind turbines power curve drift extremely
CN105024644A (en) Performance evaluation system and method of photovoltaic system
CN105825002A (en) Method for modeling dynamic equivalence of wind power farm based on dynamic grey-relevancy analysis method

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20160713