CN111794921A - Onshore wind turbine generator blade icing diagnosis method based on migration component analysis - Google Patents
Onshore wind turbine generator blade icing diagnosis method based on migration component analysis Download PDFInfo
- Publication number
- CN111794921A CN111794921A CN202010864520.2A CN202010864520A CN111794921A CN 111794921 A CN111794921 A CN 111794921A CN 202010864520 A CN202010864520 A CN 202010864520A CN 111794921 A CN111794921 A CN 111794921A
- Authority
- CN
- China
- Prior art keywords
- icing
- wind turbine
- data
- turbine generator
- component analysis
- 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
Images
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
- F03D—WIND MOTORS
- F03D80/00—Details, components or accessories not provided for in groups F03D1/00 - F03D17/00
- F03D80/40—Ice detection; De-icing means
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/70—Wind energy
- Y02E10/72—Wind turbines with rotation axis in wind direction
Abstract
The invention discloses a onshore wind turbine icing diagnosis method based on migration component analysis, which comprises the following steps of 1: forming a model training data set; step 2: selecting the characteristics; and step 3: performing repeated iteration training on the data processed by the migration component analysis method by adopting a machine learning algorithm to obtain a wind turbine blade icing state diagnosis model; and 4, step 4: and deploying and applying an online blade icing state diagnosis model. According to the onshore wind turbine icing diagnosis method based on migration component analysis, the migration component analysis method is adopted, the marginal probability distribution between the modeling unit and the target unit is minimized, and the distribution difference between data is reduced. The method provided by the invention has strong generalization capability, can effectively reduce the difference between the icing data of different units and improve the diagnosis accuracy of the blade icing of the wind turbine.
Description
The technical field is as follows:
the invention relates to the field of state monitoring and fault diagnosis of an onshore wind turbine generator, in particular to an onshore wind turbine generator blade icing diagnosis method based on migration component analysis.
Background art:
wind energy is widely concerned worldwide because of its characteristics of abundant reserves, cleanness, no pollution, green and sustainability, etc. In actual operation, the generating efficiency of the wind turbine generator is often limited by environmental factors, wherein the blade icing problem under low-temperature and humid environment is particularly prominent. When the blades are frozen, damage of different degrees can be caused, and the generating efficiency of the wind turbine generator is reduced to influence the income of a wind power plant; and if so, the blade is induced to break, so that the safety of the wind turbine generator and the operation and maintenance personnel of the wind power plant is endangered. The traditional wind turbine state monitoring system and the control system can not diagnose and early warn the early icing state of the blades, when the unit is shut down due to icing, the large-area icing phenomenon of the blades of the wind turbine often occurs, and the safety of the wind turbine and operation and maintenance personnel is seriously endangered. An accurate blade early icing state diagnosis method is urgently needed in a wind power plant. The existing icing diagnosis method mostly uses a machine learning algorithm to take historical SCADA data of one or more wind turbines as input, take blade icing state data as output, and fit an optimal icing state diagnosis model through repeated iterative training. The method has accurate diagnosis capability for wind turbines with same or similar SCADA data distribution, can accurately identify the blade icing fault, but has poor generalization capability for the wind turbines with larger SCADA data distribution difference, and cannot accurately identify the blade icing fault.
The invention content is as follows:
aiming at the defects of the prior art, the invention provides a onshore wind turbine blade icing diagnosis method based on migration component analysis, aiming at the problem that different wind turbine SCADA data distributions are different, the migration component analysis method is adopted to process a source domain (training set wind turbine SCADA data) and a target domain (wind turbine SCADA data to be diagnosed), and the edge probability distribution distance between the source domain and the target domain is minimized; and performing model training and diagnosis by using a machine learning method based on the data processed by the migration component analysis method. The method provided by the invention can reduce the distribution difference among the SCADA data of different wind turbine generators so as to realize accurate icing diagnosis of the wind turbine generators in different distributions, and provides a reliable method basis for the research in the fields of onshore wind turbine generator operation maintenance, state monitoring, fault diagnosis and the like.
The invention is implemented by the following technical scheme: a method for diagnosing icing of a land-based wind turbine based on migration component analysis comprises the following steps:
step 1: monitoring and recording the icing state of the wind turbine generator, and adding an icing label and a time continuity label to SCADA data corresponding to the SCADA system to form a model training data set;
step 2: transient state and time sequence characteristics of the SCADA data are calculated, and a greedy algorithm is used for characteristic selection;
and step 3: minimizing the difference between the characteristic data sets of the wind turbine generator set in the training set and the wind turbine generator set to be diagnosed by using a migration component analysis method, and performing repeated iterative training on the data processed by the migration component analysis method by using a machine learning algorithm to obtain a wind turbine generator set blade icing state diagnosis model;
and 4, step 4: and deploying and applying an online blade icing state diagnosis model.
Further, the step 1 comprises:
step 11: selecting a wind turbine generators in an icing period and in a good running state to form a reference generator set for training a model;
step 12: observing and recording the icing time period and the normal time period of each unit in the reference unit set in the icing period;
step 13: b characteristic variables with large blade icing correlation are selected from the original SCADA data, and the characteristic variables comprise: wind speed, variable pitch motor temperature, generator rotating speed, blade speed, net side active power, ng5 temperature, wind angle, acceleration in x direction, acceleration in y direction, 25-second average wind direction angle, yaw position, environment temperature, yaw speed, cabin temperature, blade angle and ng5 charger direct current, and the variables and the timestamp in the icing period are derived;
step 14: adding an icing state label to the SCADA derived in the step 13 according to the icing time period and normal time period data obtained in the step 12, wherein the setting mode of the icing state label is as shown in the following table:
step 15: setting a time continuity threshold value m seconds, and performing time continuity judgment on the data processed in the step 14; the judgment method is as follows: if the time interval between two continuous points is less than m seconds, judging that the two points belong to the same continuous time period; if the time interval between two continuous points is greater than m seconds, judging that the two points belong to different continuous time periods; different time continuity labels are added for data at different time periods.
Further, the step 2 comprises:
step 21: dividing the SCADA data in the normal state into n wind speed intervals with equal width at equal intervals according to the wind speed by adopting a bin method, and calculating the average wind speed and power in each wind speed interval;
step 22: fitting the average wind speed and the average power data by adopting a Savitzky-Golay smoothing algorithm to form a theoretical power curve and a function;
step 23: transient characteristics influencing icing are constructed, and the newly constructed transient characteristics and a formula are as follows:
severity of icing Wβ
In formula (1): prealIs the actual power, PpredThe theoretical response power is obtained according to the fitted power curve;
wind speed squared V2
In formula (2): v∞Is the wind speed;
wind energy comprehensive utilization rate Ctotal
In formula (3): prealIs the actual power, PpredThe theoretical response power is obtained according to the fitted power curve;
step 24: calculating the sequence characteristics such as the icing severity and the maximum value, standard deviation, average change rate and sum of the comprehensive utilization rate of wind energy in a continuous time period by adopting a sliding window method, and setting the size of a sliding window as h and the step length as i;
step 25: and eliminating invalid or unfavorable characteristics for model training by adopting a greedy algorithm through repeated iterative training, and finally forming a characteristic data set.
Further, the step 3 comprises:
step 31: carrying out the operations of the step 1 and the step 2 on the training set wind turbine generator, and eliminating invalid data in the data according to the blade icing state label obtained in the step 11 to finally obtain a characteristic data set and an icing state label of the training set wind turbine generator;
step 32: carrying out all the operations of the steps 12 to 15 and the step 2 in the step 1 on the wind turbine to be diagnosed to obtain a characteristic data set of the wind turbine to be diagnosed;
step 33: and carrying out normalization operation on the data in the characteristic data set, wherein the normalization formula is as follows:
step 34: processing the feature data sets in the steps 31 and 32 by using a migration component analysis method, and establishing a feature mapping phi to minimize the edge probability distribution distance between the mapped training set wind turbine generator set and the wind turbine generator set data to be diagnosed;
step 35: and training a wind turbine generator blade icing state diagnosis model through multiple iterations by using a machine learning algorithm A with the characteristic data set as input and the blade icing state label as output.
Further, the step 4 comprises:
step 41: setting a time sensitivity threshold value k seconds, and combining the updated SCADA data with historical SCADA data to form a new wind turbine generator set data set to be diagnosed when the duration of the updated SCADA data exceeds k seconds;
step 42: performing all the operations of the steps 11 to 15 and the step 2 on the new wind turbine generator set data set to be diagnosed to form a new wind turbine generator set characteristic data set to be diagnosed;
step 43: performing the operation in the step 3 on the training set wind turbine generator characteristic data set and the updated wind turbine generator characteristic data set to be diagnosed;
step 44: and diagnosing the blade icing state in the updating period according to the training model obtained in the step 43, and issuing early warning according to the icing diagnosis state.
According to the onshore wind turbine icing diagnosis method based on migration component analysis, the migration component analysis method is adopted, the marginal probability distribution between the modeling unit and the target unit is minimized, and the distribution difference between data is reduced. The method provided by the invention has strong generalization capability, can be used by combining different algorithms, can effectively reduce the difference between the icing data of different units, improves the diagnosis accuracy of the icing of the blades of the wind turbine generator, and has application value.
Description of the drawings:
FIG. 1 is a power curve fitting curve diagram of a wind turbine;
FIG. 2 is a graph of greedy algorithm feature selection quantity versus diagnostic accuracy.
The specific implementation mode is as follows:
the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Example 1:
the embodiment relates to a land wind turbine blade icing diagnosis method based on migration component analysis, which comprises the following specific steps of:
step 1: monitoring and recording the icing state of the wind turbine generator, and adding an icing label and a time continuity label to the corresponding SCADA data to form a model training data set;
step 11: selecting 1 wind turbine generator set in an icing period and in a good running state to form a reference set for training a model;
step 12: observing and recording the icing time period and the normal time period of each unit in the reference unit set in the icing period;
step 13: the 26 characteristic variables with large blade icing correlation are selected from the original SCADA data, and comprise: wind speed, generator rotating speed, net side active power, wind alignment angle, average wind direction angle, yaw position, yaw speed, blade angle, variable pitch motor temperature, blade speed, ng5 temperature, x-direction acceleration, y-direction acceleration, environment temperature, cabin temperature and ng5 charger direct current, and the variables and the timestamp in the icing period are derived;
step 14: adding an icing state label to the SCADA derived in the step 13 according to the icing time period and normal time period data obtained in the step 12, wherein the setting mode of the icing state label is as shown in the following table:
step 15: setting a time continuity threshold value of 3000 seconds, and performing time continuity judgment on the data processed in the step 14; the judgment method is as follows: if the time interval between two continuous points is less than 3000 seconds, judging that the two points belong to the same continuous time period; if the time interval between two continuous points is more than 3000 seconds, judging that the two points belong to different continuous time periods; adding different time continuity labels to data at different time periods;
step 2: calculating transient state and time sequence characteristics of the SCADA data according to the data obtained in the step 15, and performing characteristic selection by using a greedy algorithm, wherein the method comprises the following specific steps of:
step 21: dividing the SCADA data in a normal state into 180 wind speed intervals with equal width at equal intervals according to wind speed by adopting a bin method, and calculating the average wind speed and power in each wind speed interval;
step 22: fitting the average wind speed and average power data by adopting a Savitzky-Golay smoothing algorithm to form a theoretical power curve (shown in figure 1) and a function;
step 23: transient characteristics influencing icing are constructed, and the newly constructed transient characteristics and a formula are as follows:
severity of icing Wβ
In formula (1): prealIs the actual power, PpredIs the theoretical response power obtained according to the fitted power curve.
Wind speed squared V2
In formula (2): v∞Is the wind speed.
Wind energy comprehensive utilization rate Ctotal
In formula (3): prealIs the actual power, PpredIs the theoretical response power obtained according to the fitted power curve.
Step 24: calculating the icing severity and the maximum value, standard deviation, average change rate, sum and other time sequence characteristics of the comprehensive utilization rate of wind energy in a continuous time period by adopting a sliding window method (the size of a sliding window is set to be 50, and the step length is set to be 1);
step 25: determining the optimal feature quantity (as shown in figure 2) by adopting a greedy algorithm through multiple iterative training, and eliminating the invalid or unfavorable features for model training to finally form a feature data set;
and step 3: the method comprises the following steps of minimizing the difference between the distribution of a feature data set of a training set wind turbine generator and a wind turbine generator to be diagnosed by using a migration component analysis method, and performing repeated iterative training on data processed by the migration component analysis method by using a machine learning algorithm to obtain a wind turbine generator blade icing diagnosis model, and specifically comprises the following steps:
step 31: carrying out the operations of the step 1 and the step 2 on the training set wind turbine generator, and eliminating invalid data in the data according to the blade icing state label obtained in the step 11 to finally obtain a characteristic data set and an icing state label of the training set wind turbine generator;
step 32: carrying out all the operations of the steps 12 to 15 and the step 2 in the step 1 on the wind turbine to be diagnosed to obtain a characteristic data set of the wind turbine to be diagnosed;
step 33: and carrying out normalization operation on the data in the characteristic data set, wherein the normalization formula is as follows:
step 34: processing the feature data sets in the steps 31 and 32 by using a migration component analysis method, and establishing a feature mapping phi to minimize the edge probability distribution distance between the mapped training set wind turbine generator set and the wind turbine generator set data to be diagnosed;
step 35: taking the characteristic data set as input and the blade icing state label as output, and adopting machine learning algorithms such as linear discriminant analysis, k-nearest neighbor and naive Bayes to train a wind turbine generator blade icing state diagnosis model through multiple iterations;
and 4, step 4: the method comprises the following steps of applying and deploying an online blade icing state diagnosis model:
step 41: setting a time sensitivity threshold value for 600 seconds, and merging the updated SCADA data with historical SCADA data when the duration of the updated SCADA data exceeds 600 seconds to form a new wind turbine generator set data set to be diagnosed;
step 42: performing all the operations of the steps 11 to 15 and the step 2 on the new wind turbine generator set data set to be diagnosed to form a new wind turbine generator set characteristic data set to be diagnosed;
step 43: performing the operation in the step 3 on the training set wind turbine generator characteristic data set and the updated wind turbine generator characteristic data set to be diagnosed;
step 44: diagnosing the blade icing state in the updating period according to the training model obtained in the step 43, and issuing early warning according to the icing diagnosis state;
and dividing the original data set into a training set and a test set according to the ratio of 8:2, wherein the training set data is used for training the model, and the test set is used for testing the performance of the model. The accuracy and generalization ability of the migration component analysis methods presented herein were analyzed using three example comparisons.
Example 2:
firstly, obtaining an icing diagnosis model based on a naive Bayes algorithm through multiple iterations by combining training set data with the naive Bayes algorithm, and inputting test set data into the trained model to obtain a diagnosis accuracy index; then, the migration component analysis method described in embodiment 1 is used to adjust the data distribution difference between the training set and the test set to obtain a post-migration training set and a post-migration test set, the post-migration training set data is combined with the naive bayes algorithm to obtain an icing diagnosis model based on the naive bayes algorithm through multiple iterations, and the post-migration training set data is input into the trained model to obtain a diagnosis accuracy index.
Example 3: firstly, obtaining an icing diagnosis model based on a linear discriminant analysis algorithm through multiple iterations by combining training set data with the linear discriminant analysis algorithm, and inputting test set data into the trained model to obtain a diagnosis accuracy index; then, the migration component analysis method described in embodiment 1 is used to adjust the data distribution difference between the training set and the test set to obtain a post-migration training set and a post-migration test set, the post-migration training set data is combined with a linear discriminant algorithm to obtain an icing diagnosis model based on the linear discriminant algorithm through multiple iterations, and the post-migration training set data is input into the trained model to obtain a diagnosis accuracy index.
Example 4: firstly, combining training set data with a k-nearest neighbor algorithm to obtain an icing diagnosis model based on the k-nearest neighbor algorithm through multiple iterations, and inputting test set data into a trained model to obtain a diagnosis accuracy index; then, the migration component analysis method described in embodiment 1 is used to adjust the data distribution difference between the training set and the test set to obtain a post-migration training set and a post-migration test set, the freezing diagnosis model based on the k-nearest neighbor algorithm is obtained through multiple iterations by combining the post-migration training set data with the k-nearest neighbor algorithm, and the post-migration training set data is input into the trained model to obtain a diagnosis accuracy index.
The results of the three examples are summarized in the following table:
through the result comparison analysis of different algorithms, after the migration component analysis method is used, the diagnosis accuracy of the three algorithms is improved to different degrees, and the following conclusion can be obtained: 1) the method has better generalization capability and can be used by combining different algorithms; 2) the method provided by the invention can improve the diagnosis accuracy of different algorithms, and has application value.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (5)
1. A land wind turbine icing diagnosis method based on migration component analysis is characterized by comprising the following steps:
step 1: monitoring and recording the icing state of the wind turbine generator, and adding an icing label and a time continuity label to SCADA data corresponding to the SCADA system to form a model training data set;
step 2: transient state and time sequence characteristics of the SCADA data are calculated, and a greedy algorithm is used for characteristic selection;
and step 3: minimizing the difference between the characteristic data sets of the wind turbine generator set in the training set and the wind turbine generator set to be diagnosed by using a migration component analysis method, and performing repeated iterative training on the data processed by the migration component analysis method by using a machine learning algorithm to obtain a wind turbine generator set blade icing state diagnosis model;
and 4, step 4: and deploying and applying an online blade icing state diagnosis model.
2. The method for diagnosing icing of onshore wind turbine generators based on migration component analysis as claimed in claim 1, wherein the step 1 comprises:
step 11: selecting a wind turbine generators in an icing period and in a good running state to form a reference generator set for training a model;
step 12: observing and recording the icing time period and the normal time period of each unit in the reference unit set in the icing period;
step 13: b characteristic variables with large blade icing correlation are selected from the original SCADA data, and the characteristic variables comprise: wind speed, variable pitch motor temperature, generator rotating speed, blade speed, net side active power, ng5 temperature, wind angle, acceleration in x direction, acceleration in y direction, 25-second average wind direction angle, yaw position, environment temperature, yaw speed, cabin temperature, blade angle and ng5 charger direct current, and the variables and the timestamp in the icing period are derived;
step 14: adding an icing state label to the SCADA derived in the step 13 according to the icing time period and normal time period data obtained in the step 12, wherein the setting mode of the icing state label is as shown in the following table:
step 15: setting a time continuity threshold value m seconds, and performing time continuity judgment on the data processed in the step 14; the judgment method is as follows: if the time interval between two continuous points is less than m seconds, judging that the two points belong to the same continuous time period; if the time interval between two continuous points is greater than m seconds, judging that the two points belong to different continuous time periods; different time continuity labels are added for data at different time periods.
3. The method for diagnosing icing of onshore wind turbine generator set based on migration component analysis as claimed in claim 2, wherein the step 2 comprises
Step 21: dividing the SCADA data in the normal state into n wind speed intervals with equal width at equal intervals according to the wind speed by adopting a bin method, and calculating the average wind speed and power in each wind speed interval;
step 22: fitting the average wind speed and the average power data by adopting a Savitzky-Golay smoothing algorithm to form a theoretical power curve and a function;
step 23: transient characteristics influencing icing are constructed, and the newly constructed transient characteristics and a formula are as follows:
severity of icing Wβ
In formula (1): prealIs the actual power, PpredThe theoretical response power is obtained according to the fitted power curve;
wind speed squared V2
In formula (2): v∞Is the wind speed;
wind energy comprehensive utilization rate Ctotal
In formula (3): prealIs the actual power, PpredAccording to the fitting workThe theoretical power of the transmitter is obtained by the rate curve;
step 24: calculating the sequence characteristics such as the icing severity and the maximum value, standard deviation, average change rate and sum of the comprehensive utilization rate of wind energy in a continuous time period by adopting a sliding window method, and setting the size of a sliding window as h and the step length as i;
step 25: and eliminating invalid or unfavorable characteristics for model training by adopting a greedy algorithm through repeated iterative training, and finally forming a characteristic data set.
4. The method for diagnosing icing of onshore wind turbine generators based on migration component analysis as claimed in claim 3, wherein the step 3 comprises:
step 31: carrying out the operations of the step 1 and the step 2 on the training set wind turbine generator, and eliminating invalid data in the data according to the blade icing state label obtained in the step 11 to finally obtain a characteristic data set and an icing state label of the training set wind turbine generator;
step 32: carrying out all the operations of the steps 12 to 15 and the step 2 in the step 1 on the wind turbine to be diagnosed to obtain a characteristic data set of the wind turbine to be diagnosed;
step 33: and carrying out normalization operation on the data in the characteristic data set, wherein the normalization formula is as follows:
step 34: processing the feature data sets in the steps 31 and 32 by using a migration component analysis method, and establishing a feature mapping phi to minimize the edge probability distribution distance between the mapped training set wind turbine generator set and the wind turbine generator set data to be diagnosed;
step 35: and training a wind turbine generator blade icing state diagnosis model through multiple iterations by using a machine learning algorithm A with the characteristic data set as input and the blade icing state label as output.
5. The method for diagnosing icing of onshore wind turbine generators based on migration component analysis as claimed in claim 4, wherein the step 4 comprises:
step 41: setting a time sensitivity threshold value k seconds, and combining the updated SCADA data with historical SCADA data to form a new wind turbine generator set data set to be diagnosed when the duration of the updated SCADA data exceeds k seconds;
step 42: performing all the operations of the steps 11 to 15 and the step 2 on the new wind turbine generator set data set to be diagnosed to form a new wind turbine generator set characteristic data set to be diagnosed;
step 43: performing the operation in the step 3 on the training set wind turbine generator characteristic data set and the updated wind turbine generator characteristic data set to be diagnosed;
step 44: and diagnosing the blade icing state in the updating period according to the training model obtained in the step 43, and issuing early warning according to the icing diagnosis state.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010864520.2A CN111794921B (en) | 2020-08-25 | 2020-08-25 | Onshore wind turbine blade icing diagnosis method based on migration component analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010864520.2A CN111794921B (en) | 2020-08-25 | 2020-08-25 | Onshore wind turbine blade icing diagnosis method based on migration component analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111794921A true CN111794921A (en) | 2020-10-20 |
| CN111794921B CN111794921B (en) | 2023-04-18 |
Family
ID=72834659
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010864520.2A Active CN111794921B (en) | 2020-08-25 | 2020-08-25 | Onshore wind turbine blade icing diagnosis method based on migration component analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111794921B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112594125A (en) * | 2020-11-29 | 2021-04-02 | 上海电机学院 | Automatic-shrinkage wind power generation blade and control method thereof |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120226485A1 (en) * | 2011-03-03 | 2012-09-06 | Inventus Holdings, Llc | Methods for predicting the formation of wind turbine blade ice |
| CN105464912A (en) * | 2016-01-27 | 2016-04-06 | 国电联合动力技术有限公司 | Method and device for detecting freezing of wind generating set blades |
| CN109209790A (en) * | 2018-10-09 | 2019-01-15 | 浙江运达风电股份有限公司 | A kind of wind electricity blade icing estimation method based on data modeling |
| WO2019086287A1 (en) * | 2017-10-30 | 2019-05-09 | fos4X GmbH | Method for forecasting the yield of a wind farm under icing conditions |
-
2020
- 2020-08-25 CN CN202010864520.2A patent/CN111794921B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120226485A1 (en) * | 2011-03-03 | 2012-09-06 | Inventus Holdings, Llc | Methods for predicting the formation of wind turbine blade ice |
| CN105464912A (en) * | 2016-01-27 | 2016-04-06 | 国电联合动力技术有限公司 | Method and device for detecting freezing of wind generating set blades |
| WO2019086287A1 (en) * | 2017-10-30 | 2019-05-09 | fos4X GmbH | Method for forecasting the yield of a wind farm under icing conditions |
| CN109209790A (en) * | 2018-10-09 | 2019-01-15 | 浙江运达风电股份有限公司 | A kind of wind electricity blade icing estimation method based on data modeling |
Non-Patent Citations (1)
| Title |
|---|
| 李宁波等: "基于SCADA数据的风机叶片结冰检测方法", 《发电技术》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112594125A (en) * | 2020-11-29 | 2021-04-02 | 上海电机学院 | Automatic-shrinkage wind power generation blade and control method thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111794921B (en) | 2023-04-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109751206B (en) | Fan blade icing fault prediction method and device and storage medium | |
| CN105134510A (en) | State monitoring and failure diagnosis method for wind generating set variable pitch system | |
| CN110362045B (en) | Marine doubly-fed wind turbine generator fault discrimination method considering marine meteorological factors | |
| CN111091298B (en) | Wind power plant flow field coupling characteristic evaluation and intelligent grouping method and system | |
| CN106570790B (en) | Wind power plant output data restoration method considering wind speed data segmentation characteristics | |
| CN103925155A (en) | Self-adaptive detection method for abnormal wind turbine output power | |
| Chen et al. | Learning deep representation for blades icing fault detection of wind turbines | |
| CN112417612B (en) | Wind power gear box degradation state tracking and failure aggregation risk assessment method | |
| CN115013258B (en) | Intelligent soft measurement method for wind speed in front of wind turbine generator | |
| Shi et al. | Study of wind turbine fault diagnosis and early warning based on SCADA data | |
| CN113612237A (en) | Method for positioning resonance-induced subsynchronous oscillation source in offshore wind farm | |
| CN111794921B (en) | Onshore wind turbine blade icing diagnosis method based on migration component analysis | |
| CN116771610A (en) | Method for adjusting fault evaluation value of variable pitch system of wind turbine | |
| CN110273818B (en) | Fan blade icing fault monitoring method based on shaft transformation thickness classification | |
| CN112215281A (en) | Fan blade icing fault detection method | |
| CN115898787A (en) | Method and device for dynamically identifying static yaw error of wind turbine generator | |
| CN113847214B (en) | Icing detection method for large wind generating set | |
| Bao et al. | Iterative modeling of wind turbine power curve based on least‐square B‐spline approximation | |
| CN114320773A (en) | Wind turbine generator fault early warning method based on power curve analysis and neural network | |
| Souza et al. | Evaluation of data based normal behavior models for fault detection in wind turbines | |
| Ye et al. | Using SCADA data fusion by swarm intelligence for wind turbine condition monitoring | |
| CN113048012A (en) | Wind turbine generator yaw angle identification method and device based on Gaussian mixture model | |
| Guo et al. | Early Fault Detection of Wind Turbine Gearbox Based on Adam-trained LSTM | |
| Peng et al. | Short-Term Prediction of Generator Blade Ice Fault Based on Multi-AN | |
| CN117728587B (en) | Real-time monitoring system and method for operation data of new energy power generation equipment |
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 |