CN116412091A - Real-time deicing method and system for wind generating set blades based on machine learning - Google Patents
Real-time deicing method and system for wind generating set blades based on machine learningDownload PDFInfo
- Publication number
- CN116412091A CN116412091ACN202310228069.9ACN202310228069ACN116412091ACN 116412091 ACN116412091 ACN 116412091ACN 202310228069 ACN202310228069 ACN 202310228069ACN 116412091 ACN116412091 ACN 116412091A
- Authority
- CN
- China
- Prior art keywords
- deicing
- fan blade
- time
- ice
- real
- 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.)
- Withdrawn
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
Landscapes
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Sustainable Development (AREA)
- Sustainable Energy (AREA)
- Chemical & Material Sciences (AREA)
- Combustion & Propulsion (AREA)
- Mechanical Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Structures Of Non-Positive Displacement Pumps (AREA)
Abstract
Description
Technical Field
The invention relates to the field of real-time deicing, in particular to a real-time deicing method and system for blades of a wind generating set based on machine learning.
Background
At present, the total installed quantity of the fans in China is the first place in the world, and the situation of year-by-year growth is presented, wherein wind power plants in China are mainly distributed in wet and cold areas such as three north, and the fan blades in the areas have serious icing phenomenon. The blade icing causes great harm to the mechanical performance and the pneumatic performance of the fan, and seriously influences the output efficiency of the fan. Wind power generation has a relatively late start time in China, and domestic researches on fan blade deicing technology are relatively few and immature, so that the ice disaster prevention and control of the fan blades become one of the problems which must be solved in the development process of the wind power industry in order to ensure the normal and stable operation of a wind power plant.
However, the existing deicing technology is easy to cause mechanical damage to the blade, the vibration effect of the position near the blade root is poor, the deicing reliability is low, and on the other hand, the instantaneous electric power used in the deicing process is large, so that irreversible damage can be caused to equipment.
Therefore, the invention provides a real-time deicing method and a real-time deicing system for blades of a wind generating set based on machine learning.
Disclosure of Invention
The invention provides a real-time deicing method and a real-time deicing system for blades of a wind generating set based on machine learning, which are used for establishing an ice block coverage model by processing and analyzing real-time data of the blades of a fan so as to judge ice blocks and obtain a corresponding accurate deicing scheme, so that the deicing reliability is improved, and meanwhile, the power consumption in the deicing process is reduced and the damage to equipment is reduced.
The invention provides a real-time deicing method for blades of a wind generating set based on machine learning, which comprises the following steps:
step 1: acquiring real-time data information of different parts of each fan blade in the target wind generating set based on a preset sensor;
step 2: processing the acquired real-time data information based on machine learning, and establishing a first ice block coverage model of each fan blade based on a preset algorithm;
step 3: determining key ice cube attributes of corresponding fan blades based on the first ice cube covering model, and transmitting the key ice cube attributes to an intelligent management terminal for analysis;
step 4: based on the analysis result, predicting the ice cube index, determining a deicing scheme, and realizing real-time deicing of the target fan blade.
In one possible implementation manner, acquiring real-time data information of different parts of each fan blade in the target wind turbine generator system based on a preset sensor includes:
step 11: acquiring real-time sensor data of different parts of a target fan blade in a wind generating set based on a plurality of preset different types of sensors;
step 12: and filling the sensor data into a preset sensor data table, and carrying out standardized processing on the sensor data based on a standard data format of a corresponding data type to obtain real-time data information.
In one possible implementation, the processing of the acquired real-time data information based on machine learning and the establishment of the first ice covering model of each fan blade based on a preset algorithm includes:
step 21: carrying out information analysis on the acquired real-time data information;
step 22: screening the data algorithms of the same type which can be matched in the algorithm database based on the information type and the information analysis result of the real-time data information;
substituting the analysis result into the same type of data algorithm, and determining the matching degree of the analysis result and the same type of data algorithm;
step 23: screening an algorithm with the highest matching value based on the obtained matching degree, and taking the algorithm as a preset algorithm of the current fan blade;
step 24: based on the preset algorithm, an original ice block coverage model of the current fan blade is established;
step 25: and based on the historical data information of the current fan blade, testing and adjusting the original ice block coverage model to obtain an adjusted first ice block coverage model.
In one possible implementation manner, determining the key ice cube attribute of the corresponding fan blade based on the first ice cube coverage model, and transmitting the key ice cube attribute to the intelligent management terminal for analysis, including:
step 31: determining key ice cube attributes corresponding to the current fan blade based on the first ice cube covering model;
step 32: transmitting the key ice block attributes to an intelligent management terminal based on the key ice block attributes, and matching the key ice block attributes with preset ice block attributes;
step 33: determining a standard ice cube key attribute with the highest matching degree with the ice cube key attribute of the current fan blade and an ice cube grade corresponding to the standard ice cube key attribute based on the matching result;
step 34: and performing a first analysis on the ice of the current fan blade based on the ice grade and the ice attribute of the same ice grade corresponding to the standard ice key attribute.
In one possible implementation, predicting an ice cube index based on the analysis results, determining a de-icing scheme to achieve real-time de-icing of the target fan blade, comprising:
step 41: the first analysis result is corresponding to a preset analysis table, and the index of the matched ice cubes is used as a first prediction index of the ice cubes corresponding to the fan blades;
step 42: determining a corresponding first deicing scheme based on a first prediction index, and simulating the first deicing scheme on the basis of the first deicing scheme in an intelligent management terminal, so that the first deicing scheme is adjusted, and a second deicing scheme is obtained;
step 43: and carrying out real-time deicing on the corresponding positions of the fan blades based on the second deicing scheme.
In one possible implementation manner, determining a corresponding first deicing scheme based on the first prediction index, and simulating at the intelligent management terminal based on the scheme, so as to adjust the first deicing scheme to obtain a second deicing scheme, including:
step 421: comparing the index range based on the first predictive index to an index scheme table;
if the first prediction index is in a first preset range of the index scheme table, working based on the blade rotation inertia force of the fan blade and the corresponding ultrasonic module to serve as a first deicing scheme of the current fan blade;
if the first prediction index is in a second preset range of the index scheme table, working based on the blade rotation inertia force of the fan blade and the corresponding laser module to serve as a first deicing scheme of the current fan blade;
otherwise, based on the blade rotation inertia force of the fan blade, the corresponding laser module and the ultrasonic module, the fan blade is used as a first deicing scheme of the current fan blade;
step 422: simulating in the intelligent management terminal based on the predicted first deicing scheme, and predicting deicing time of the corresponding fan blade;
wherein T is0 The influence time of the common influence factors; t (T)1 The deicing time of the fan blade when the first prediction index is in a first preset range of the index scheme table is set; t (T)2 The deicing time of the fan blade when the first prediction index is in a second preset range of the index scheme table is set; t (T)3 Deicing time of the fan blade when the first prediction index is out of a second preset range of the index scheme table; t (T)s The deicing time of the fan blade in the standard state is set; beta1 、β2 、β3 、β4 Time conversion coefficients for each of the common influencing factors; h is a1 The maximum ice coating thickness of the current fan blade is set; h is a2 Is the minimum ice thickness; h is a0 The thickness of the ice coating is used for influencing the change point of the deicing time; alpha1 First influence index of ice coating thickness on deicing time；α2 A second impact index of ice coating thickness on deicing time; t is t1 The maximum environmental temperature of the current fan blade is the environmental; t is t2 Is the minimum ambient temperature; t is t0 The environmental temperature is the environmental temperature which influences the change point of the deicing time; mu (mu)1 A first impact index of ambient temperature on deicing time; mu (mu)2 A second impact index of ambient temperature on deicing time; v is the real-time working wind speed of the fan blade; vs The wind speed is the standard wind speed for the normal operation of the fan blade; l is the real-time liquid water content of the environment where the fan blade is located; l (L)s The standard liquid water content is the standard liquid water content of the environment where the fan blade is positioned; p (P)1 The heat release power of the fan blade under the influence of ultrasonic waves is provided; c (C)1 The heating time of the fan blade under the influence of ultrasonic waves is set; c is the specific heat capacity of ice; m is the mass of ice cubes on the target fan blade; omega is a temperature time conversion coefficient; p (P)2 The heat release power of the fan blade under the influence of laser is provided; c (C)2 The heating time of the fan blade under the influence of laser is set; epsilon1 、ε2 The influence weight of the current heating power on the current deicing time is given;
if the deicing time is within a first preset time range, determining that the current first deicing scheme is a second deicing scheme;
otherwise, data acquisition, algorithm matching and model establishment are carried out again, and a second deicing scheme is obtained.
In one possible implementation, predicting the ice cube index based on the analysis result, determining a deicing scheme, and after implementing real-time deicing of the target fan blade, further includes: the method for verifying the deicing result specifically comprises the following steps:
step 01: when the real-time deicing task is displayed, acquiring real-time data information of a preset sensor again, and performing standardized processing;
step 02: comparing the real-time data information after standardized processing with standard state data corresponding to the current fan blade to realize result verification;
if the checking result is within the preset normal range, judging that the current deicing task is completed, and transmitting the completed result to the intelligent management terminal;
otherwise, judging that the current deicing task is not completed, reestablishing a first ice block coverage model based on real-time data information of the fan blades with not completed deicing, analyzing and predicting, and deicing again.
The invention provides a real-time deicing system for blades of a wind generating set based on machine learning, which comprises the following components:
and a data acquisition module: acquiring real-time data information of different parts of each fan blade in the target wind generating set based on a preset sensor;
and a model building module: processing the acquired real-time data information based on machine learning, and establishing a first ice block coverage model of each fan blade based on a preset algorithm;
analysis and prediction module: determining key ice cube attributes of corresponding fan blades based on the first ice cube covering model, and transmitting the key ice cube attributes to an intelligent management terminal for analysis;
deicing processing module: based on the analysis result, predicting the ice cube index, determining a deicing scheme, and realizing real-time deicing of the target fan blade.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims thereof as well as the appended drawings.
The technical scheme of the invention is further described in detail through the drawings and the embodiments.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention. In the drawings:
FIG. 1 is a flow chart of a real-time deicing method for wind turbine generator set blades based on machine learning in an embodiment of the invention;
FIG. 2 is a flowchart of a method for predicting ice cube index and determining deicing scheme based on analysis results in a machine learning-based real-time deicing method for wind turbine generator set blades according to an embodiment of the present invention;
FIG. 3 is a block diagram of a real-time deicing system for wind turbine generator set blades based on machine learning in an embodiment of the present invention;
FIG. 4 is another flow chart of a real-time deicing method for wind turbine generator set blades based on machine learning in an embodiment of the invention.
Detailed Description
The preferred embodiments of the present invention will be described below with reference to the accompanying drawings, it being understood that the preferred embodiments described herein are for illustration and explanation of the present invention only, and are not intended to limit the present invention.
Example 1:
the embodiment of the invention provides a real-time deicing method for a wind generating set blade based on machine learning, which comprises the following steps as shown in fig. 1:
step 1: acquiring real-time data information of different parts of each fan blade in the target wind generating set based on a preset sensor;
step 2: processing the acquired real-time data information based on machine learning, and establishing a first ice block coverage model of each fan blade based on a preset algorithm;
step 3: determining key ice cube attributes of corresponding fan blades based on the first ice cube covering model, and transmitting the key ice cube attributes to an intelligent management terminal for analysis;
step 4: based on the analysis result, predicting the ice cube index, determining a deicing scheme, and realizing real-time deicing of the target fan blade.
In this embodiment, the preset sensor is a sensor device for collecting data information such as blade acceleration, temperature, humidity, air pressure, wind speed, load, etc. of the fan blade.
In this embodiment, the target fan blade refers to a fan blade in a wind generating set that needs to generate wind power.
In this embodiment, the real-time data information refers to data information of the fan blade collected based on a preset sensor, for example, including blade acceleration, temperature, humidity, air pressure, wind speed, load, and the like.
In this embodiment, machine learning refers to a process of selecting a proper mathematical model, formulating parameters, inputting data information of fan blades, training the model by applying a proper learning algorithm according to a certain strategy, and finally analyzing and predicting the data by applying the trained model based on preliminary knowledge of the data information of the fan blades and analysis of learning purposes.
In this embodiment, the preset algorithm is a data algorithm of the same type that can be matched in the information type based on real-time information and the information analysis result screening algorithm database, the analysis result is substituted into the data algorithm of the same type, the matching degree of the analysis result and the data algorithm of the same type is determined, the algorithm with the highest matching value is screened based on the obtained matching degree, the preset algorithm is used as the preset algorithm of the current fan blade, and the machine learning model required to be used for different scenes is different, so that the matching screening of the related algorithm can be performed, the algorithm with the highest matching degree, namely the machine learning model, is obtained, and then after various information acquired by the sensor is acquired, various learning of the machine learning model is performed, and the establishment of the ice block coverage model is realized.
In this embodiment, the first ice covering model is an original ice covering model of a fan blade based on a preset algorithm, and based on historical data information of a current fan blade, the original ice covering model is adjusted, and an obtained ice covering model is, for example, each set of historical data information of the fan blade is input into the original ice covering model, an output result of an ice covering condition is obtained based on the current original ice covering model, and based on the output result and the ice covering result obtained by the historical data information, comparison is performed, if a deviation exists in the comparison result, a corresponding algorithm is obtained based on the current historical data information, a historical ice covering model is generated, the difference between the historical ice covering model and the original ice model is compared, and corresponding adjustment is performed.
In this embodiment, the key ice properties refer to measurable properties of ice on the fan blade, such as ice size, ice location.
In this embodiment, the ice cube index refers to obtaining a corresponding ice cube grade based on key ice cube attributes, and obtaining a corresponding ice cube index based on an analysis result corresponding to the current ice cube grade.
In this embodiment, the deicing scheme refers to a deicing scheme for obtaining corresponding ice cubes based on the ice cube index, and different deicing schemes can be determined based on different ice cube indexes, for example, according to different ice cube indexes, the blade rotation inertia force of the fan blade, the corresponding laser module, the ultrasonic module and the like are used as the deicing scheme of the fan blade.
In this embodiment, the specific procedure for steps 1-4 is as shown in FIG. 4.
The beneficial effects of the technical scheme are as follows: through processing and analyzing the real-time data of the fan blades, an ice block coverage model is built, so that ice blocks are judged, a corresponding accurate deicing scheme is obtained, deicing reliability is improved, meanwhile, power consumption in a deicing process is reduced, and damage to equipment is reduced.
Example 2:
based on embodiment 1, acquiring real-time data information of different parts of each fan blade in the target wind turbine generator system based on a preset sensor includes:
step 11: acquiring real-time sensor data of different parts of a target fan blade in a wind generating set based on a plurality of preset different types of sensors;
step 12: and filling the sensor data into a preset sensor data table, and carrying out standardized processing on the sensor data based on a standard data format of a corresponding data type to obtain real-time data information.
In this embodiment, the preset sensor is a sensor device for collecting data information such as blade acceleration, temperature, humidity, air pressure, wind speed, load, etc. of the fan blade.
In this embodiment, the target fan blade refers to a fan blade that performs wind power generation in a wind turbine generator set.
In this embodiment, the real-time sensor data of different parts refers to real-time sensor data obtained by corresponding sensors at the front, middle and rear parts of the fan blade.
In this embodiment, the preset sensor data table refers to a sensor data table set based on a sensor acquisition data type of a fan blade, where a first column is a sensor data type, for example, a blade acceleration, a temperature, a humidity, an air pressure, a wind speed, a load, and a second column corresponds to data information of the sensor.
In this embodiment, the standard data format refers to a standard data format for the corresponding data type in the sensor table, for example, a standard data format for blade acceleration of 10 meters per square second.
In this embodiment, the normalization processing refers to processing for performing inverse index normalization, such as reciprocal normalization and subtractive normalization, on the real-time sensor data.
In this embodiment, the real-time data information refers to data information of the fan blade collected based on a preset sensor, for example, including blade acceleration, temperature, humidity, air pressure, wind speed, load, and the like.
The technical scheme has the beneficial effects that: the sensor data are processed to obtain real-time data of the fan blades, and the real-time data are analyzed, so that an ice block coverage model is built, ice blocks are judged, a corresponding accurate deicing scheme is obtained, and deicing reliability can be improved.
Example 3:
based on embodiment 2, processing the acquired real-time data information based on machine learning, and establishing a first ice covering model of each fan blade based on a preset algorithm, including:
step 21: carrying out information analysis on the acquired real-time data information;
step 22: screening the data algorithms of the same type which can be matched in the algorithm database based on the information type and the information analysis result of the real-time data information;
substituting the analysis result into the same type of data algorithm, and determining the matching degree of the analysis result and the same type of data algorithm;
step 23: screening an algorithm with the highest matching value based on the obtained matching degree, and taking the algorithm as a preset algorithm of the current fan blade;
step 24: based on the preset algorithm, an original ice block coverage model of the current fan blade is established;
step 25: and based on the historical data information of the current fan blade, testing and adjusting the original ice block coverage model to obtain an adjusted first ice block coverage model.
In this embodiment, the information analysis refers to classifying the real-time data information, and analyzing and sorting the classified data information.
In this embodiment, the algorithm database refers to a database containing all types of algorithms.
In this embodiment, the data algorithm of the same type refers to a data algorithm which is screened out from the algorithm database based on the information type of the real-time data information and is consistent with the current information type.
In this embodiment, the matching degree refers to substituting the information analysis result into the same type of data algorithm, and judging the matching degree of the information analysis result and the same type of data algorithm.
In this embodiment, the matching value refers to a value corresponding to the matching degree when the information analysis result is matched with the data algorithm of the same type.
In this embodiment, the preset algorithm is based on the matching value, and the algorithm with the highest matching value is selected as the preset algorithm.
In this embodiment, the original ice overlay model is an ice model constructed based on a predetermined algorithm and real-time data information.
In this embodiment, the historical data information refers to historical data information collected by the sensor of the target fan blade.
In this embodiment, the first ice covering model is an original ice covering model of a fan blade based on a preset algorithm, and the original ice covering model is adjusted based on historical data information of the current fan blade, so as to obtain an ice covering model.
In this embodiment, the first ice overlay model communicates with the fan SCADA background to obtain the fan and ice key attributes.
The technical scheme has the beneficial effects that: through the real-time data of fan blade and analysis to establish ice-cube cover model, can judge ice-cube, thereby obtain corresponding accurate deicing scheme, can make deicing reliability increase.
Example 4:
based on embodiment 3, determining key ice attributes of corresponding fan blades based on the first ice coverage model, and transmitting the key ice attributes to an intelligent management terminal for analysis, wherein the key ice attributes comprise:
step 31: determining key ice cube attributes corresponding to the current fan blade based on the first ice cube covering model;
step 32: transmitting the key ice block attributes to an intelligent management terminal based on the key ice block attributes, and matching the key ice block attributes with preset ice block attributes;
step 33: determining a standard ice cube key attribute with the highest matching degree with the ice cube key attribute of the current fan blade and an ice cube grade corresponding to the standard ice cube key attribute based on the matching result;
step 34: and performing a first analysis on the ice of the current fan blade based on the ice grade and the ice attribute of the same ice grade corresponding to the standard ice key attribute.
In this embodiment, the first ice covering model is an original ice covering model of a fan blade based on a preset algorithm, and the original ice covering model is adjusted based on historical data information of the current fan blade, so as to obtain an ice covering model.
In this embodiment, the key ice properties refer to measurable properties of ice on the fan blade, such as ice size, ice location.
In this embodiment, the preset ice cube attributes refer to all ice cube attributes of ice cubes that may be present on the preset fan blade.
In this embodiment, the standard ice key attribute refers to an ice key attribute with the highest matching degree with the ice key attribute of the current fan blade in the preset ice attributes.
In this embodiment, the ice level refers to a corresponding ice level obtained based on the key attribute of the current standard ice, for example, ice size of 10 cubic centimeters, and the ice level corresponding to the ice position in the middle of the fan blade is 2.
In this embodiment, the ice cube attributes refer to preset ice cube attributes that include standard ice cube key attributes.
In this embodiment, the first analysis refers to analyzing ice cube attributes of the same ice cube level corresponding to standard ice cube key attributes.
The technical scheme has the beneficial effects that: the key attribute of the ice is determined by establishing the ice covering model, so that the ice is judged, a corresponding accurate deicing scheme can be obtained, and the deicing reliability is improved.
Example 5:
based on the embodiment 4, based on the analysis result, the ice cube index is predicted, and the deicing scheme is determined, so that the target fan blade is deicing in real time, as shown in fig. 2, and the method comprises the following steps:
step 41: the first analysis result is corresponding to a preset analysis table, and the index of the matched ice cubes is used as a first prediction index of the ice cubes corresponding to the fan blades;
step 42: determining a corresponding first deicing scheme based on a first prediction index, and simulating the first deicing scheme on the basis of the first deicing scheme in an intelligent management terminal, so that the first deicing scheme is adjusted, and a second deicing scheme is obtained;
step 43: and carrying out real-time deicing on the corresponding positions of the fan blades based on the second deicing scheme.
In this embodiment, the first analysis refers to analyzing ice cube attributes of the same ice cube level corresponding to standard ice cube key attributes.
In this embodiment, the preset analysis table is a preset ice cube analysis table based on historical data information of the current fan blade, and the preset ice cube index corresponding to the analysis result corresponding to the first analysis result in the preset analysis table is the ice cube index of the ice cube of the current fan blade.
In this embodiment, the ice cube index refers to obtaining a corresponding ice cube grade based on key ice cube attributes, and obtaining a corresponding ice cube index based on an analysis result corresponding to the current ice cube grade.
In this embodiment, the first predictive index is determined based on the ice cube index.
In this embodiment, the first deicing scheme refers to a deicing scheme for obtaining corresponding ice cubes based on the first prediction index, and different deicing schemes may be determined based on different first prediction indexes, for example, according to different first prediction indexes, a blade rotational inertia force of the fan blade, a corresponding laser module, an ultrasonic module, and the like are used as deicing schemes for ice cubes of the fan blade.
In this embodiment, the second deicing scheme is obtained by adjusting the first deicing scheme when performing simulation based on the first deicing scheme.
The beneficial effects of the technical scheme are as follows: by predicting the ice cube index, determining the deicing scheme based on the ice cube index and adjusting the deicing scheme, the deicing reliability can be improved, meanwhile, the power consumption in the deicing process is reduced, and the damage to equipment is reduced.
Example 6:
based on embodiment 5, determining a corresponding first deicing scheme based on the first prediction index, and simulating at the intelligent management terminal based on the scheme, thereby adjusting the first deicing scheme to obtain a second deicing scheme, including:
step 421: comparing the index range based on the first predictive index to an index scheme table;
if the first prediction index is in a first preset range of the index scheme table, working based on the blade rotation inertia force of the fan blade and the corresponding ultrasonic module to serve as a first deicing scheme of the current fan blade;
if the first prediction index is in a second preset range of the index scheme table, working based on the blade rotation inertia force of the fan blade and the corresponding laser module to serve as a first deicing scheme of the current fan blade;
otherwise, based on the blade rotation inertia force of the fan blade, the corresponding laser module and the ultrasonic module, the fan blade is used as a first deicing scheme of the current fan blade;
step 422: simulating in the intelligent management terminal based on the predicted first deicing scheme, and predicting deicing time of the corresponding fan blade;
wherein T is0 The influence time of the common influence factors; t (T)1 The deicing time of the fan blade when the first prediction index is in a first preset range of the index scheme table is set; t (T)2 The deicing time of the fan blade when the first prediction index is in a second preset range of the index scheme table is set; t (T)3 Deicing time of the fan blade when the first prediction index is out of a second preset range of the index scheme table; t (T)s The deicing time of the fan blade in the standard state is set; beta1 、β2 、β3 、β4 Time conversion coefficients for each of the common influencing factors; h is a1 The maximum ice coating thickness of the current fan blade is set; h is a2 Is the minimum ice thickness; h is a0 The thickness of the ice coating is used for influencing the change point of the deicing time; alpha1 A first impact index of ice coating thickness on deicing time; alpha2 A second impact index of ice coating thickness on deicing time; t is t1 The maximum environmental temperature of the current fan blade is the environmental; t is t2 Is the minimum ambient temperature; t is t0 The environmental temperature is the environmental temperature which influences the change point of the deicing time; mu (mu)1 A first impact index of ambient temperature on deicing time; mu (mu)2 A second impact index of ambient temperature on deicing time; v is the real-time working wind speed of the fan blade; vs The wind speed is the standard wind speed for the normal operation of the fan blade; l is the real-time liquid water content of the environment where the fan blade is located; l (L)s The standard liquid water content is the standard liquid water content of the environment where the fan blade is positioned; p (P)1 The heat release power of the fan blade under the influence of ultrasonic waves is provided; c (C)1 The heating time of the fan blade under the influence of ultrasonic waves is set; c is the specific heat capacity of ice; m is the mass of ice cubes on the target fan blade; omega is a temperature time conversion coefficient; p (P)2 The heat release power of the fan blade under the influence of laser is provided; c (C)2 The heating time of the fan blade under the influence of laser is set; epsilon1 、ε2 The influence weight of the current heating power on the current deicing time is given;
if the deicing time is within a first preset time range, determining that the current first deicing scheme is a second deicing scheme;
otherwise, data acquisition, algorithm matching and model establishment are carried out again, and a second deicing scheme is obtained.
In this embodiment, the index scheme table refers to a scheme table including all index ranges and solutions corresponding to each range.
In this embodiment, the first preset range and the second preset range are predetermined based on the attribute of the fan blade and the environmental factor, and the first preset range and the second preset range can be adjusted based on the difference of the attribute of the fan blade and the difference of the environment, where the first preset range is smaller than the second preset range.
In this embodiment, the first deicing scheme may be adjusted according to different ice cube indices, where the first prediction index is within a first preset range of the index scheme table, and the first deicing scheme is used as the first deicing scheme of the current fan blade based on the rotational inertial force of the fan blade and the corresponding ultrasonic module; if the first prediction index is in a second preset range of the index scheme table, working based on the blade rotation inertia force of the fan blade and the corresponding laser module to serve as a first deicing scheme of the current fan blade; otherwise, based on the blade rotation inertia force of the fan blade, the corresponding laser module and the ultrasonic module, the fan blade is used as a first deicing scheme of the current fan blade.
In this embodiment, the laser module and the ultrasonic module are both located on the outer wall of the tower of the fan blade.
In this embodiment, the de-icing time refers to the time based on the removal of the corresponding ice cubes when under the first de-icing regimen.
In this embodiment, the ice coating thickness refers to the thickness of ice cubes on the fan blade, the impact index of ice cubes on ice cubes when the ice coating thickness is higher than the ice coating thickness of the ice cube time impact change point is different from the impact index of ice cubes on ice cubes when the ice coating thickness is lower than the ice coating thickness of the ice cube time impact change point, and the first impact index is greater than the second impact index.
In the embodiment, the standard wind speed of the normal working of the fan blade is not less than the real-time wind speed of the fan blade, and the standard liquid water content of the environment where the fan blade is positioned is not less than the real-time liquid water content of the current environment.
In this embodiment, the first preset time range is determined based on the operating efficiency and the operating time of the fan blade.
In this embodiment, the second deicing scheme is obtained by determining a corresponding deicing time based on the first deicing scheme, and adjusting the first deicing scheme based on the deicing time.
The beneficial effects of the technical scheme are as follows: through the simulation to the first deicing scheme to judge deicing time, and adjust first deicing scheme based on deicing time, can obtain corresponding accurate deicing scheme, make deicing reliability increase, make the power consumption in the deicing process reduce simultaneously, reduce the harm to equipment.
Example 7:
based on the basis of embodiment 5, based on the analysis result, the ice cube index is predicted, and the deicing scheme is determined, and after the real-time deicing of the target fan blade is realized, the method further comprises the following steps: the method for verifying the deicing result specifically comprises the following steps:
step 01: when the real-time deicing task is displayed, acquiring real-time data information of a preset sensor again, and performing standardized processing;
step 02: comparing the real-time data information after standardized processing with standard state data corresponding to the current fan blade to realize result verification;
if the checking result is within the preset normal range, judging that the current deicing task is completed, and transmitting the completed result to the intelligent management terminal;
otherwise, judging that the current deicing task is not completed, reestablishing a first ice block coverage model based on real-time data information of the fan blades with not completed deicing, analyzing and predicting, and deicing again.
In this embodiment, the standard state data refers to state data collected by all sensors when the current fan blade normally works in the current working environment, for example, corresponding state data in the states of 101kpa,15 degrees celsius, primary wind, no ice and the like are standard state data.
In this embodiment, the result verification means that the real-time data information is compared with the standard state data corresponding to the current fan blade, so as to implement the verification of the deicing result.
The beneficial effects of the technical scheme are as follows: the real-time data information after deicing is compared with the standard state information, so that the deicing result of the fan blade is verified, the deicing reliability can be increased, the precision of the deicing result is increased, and the damage to equipment is reduced.
Example 8:
the embodiment of the invention provides a real-time deicing system for blades of a wind generating set based on machine learning, which is shown in fig. 3 and comprises the following components:
and a data acquisition module: acquiring real-time data information of different parts of each fan blade in the target wind generating set based on a preset sensor;
and a model building module: processing the acquired real-time data information based on machine learning, and establishing a first ice block coverage model of each fan blade based on a preset algorithm;
analysis and prediction module: determining key ice cube attributes of corresponding fan blades based on the first ice cube covering model, and transmitting the key ice cube attributes to an intelligent management terminal for analysis;
deicing processing module: based on the analysis result, predicting the ice cube index, determining a deicing scheme, and realizing real-time deicing of the target fan blade.
The beneficial effects of the technical scheme are as follows: through processing and analyzing the real-time data of the fan blades, an ice block coverage model is built, so that ice blocks are judged, a corresponding accurate deicing scheme is obtained, deicing reliability is improved, meanwhile, power consumption in a deicing process is reduced, and damage to equipment is reduced.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (8)
1. The real-time deicing method for the blades of the wind generating set based on machine learning is characterized by comprising the following steps of:
step 1: acquiring real-time data information of different parts of each fan blade in the target wind generating set based on a preset sensor;
step 2: processing the acquired real-time data information based on machine learning, and establishing a first ice block coverage model of each fan blade based on a preset algorithm;
step 3: determining key ice cube attributes of corresponding fan blades based on the first ice cube covering model, and transmitting the key ice cube attributes to an intelligent management terminal for analysis;
step 4: based on the analysis result, predicting the ice cube index, determining a deicing scheme, and realizing real-time deicing of the target fan blade.
2. The machine learning-based real-time deicing method for wind turbine blades of claim 1, wherein acquiring real-time data information of different parts of each fan blade in the target wind turbine based on the preset sensor comprises:
step 11: acquiring real-time sensor data of different parts of a target fan blade in a wind generating set based on a plurality of preset different types of sensors;
step 12: and filling the sensor data into a preset sensor data table, and carrying out standardized processing on the sensor data based on a standard data format of a corresponding data type to obtain real-time data information.
3. The machine learning based real-time deicing method for wind turbine generator set blades of claim 2, wherein processing the acquired real-time data information based on machine learning and establishing a first ice overlay model for each fan blade based on a preset algorithm comprises:
step 21: carrying out information analysis on the acquired real-time data information;
step 22: screening the data algorithms of the same type which can be matched in the algorithm database based on the information type and the information analysis result of the real-time data information;
substituting the analysis result into the same type of data algorithm, and determining the matching degree of the analysis result and the same type of data algorithm;
step 23: screening an algorithm with the highest matching value based on the obtained matching degree, and taking the algorithm as a preset algorithm of the current fan blade;
step 24: based on the said preset algorithm(s), establishing an original ice block coverage model of the current fan blade;
step 25: and based on the historical data information of the current fan blade, testing and adjusting the original ice block coverage model to obtain an adjusted first ice block coverage model.
4. The machine learning based real-time deicing method for wind turbine blades of claim 3, wherein determining key ice attributes of corresponding fan blades based on the first ice overlay model and transmitting to an intelligent management terminal for analysis comprises:
step 31: determining key ice cube attributes corresponding to the current fan blade based on the first ice cube covering model;
step 32: transmitting the key ice block attributes to an intelligent management terminal based on the key ice block attributes, and matching the key ice block attributes with preset ice block attributes;
step 33: determining a standard ice cube key attribute with the highest matching degree with the ice cube key attribute of the current fan blade and an ice cube grade corresponding to the standard ice cube key attribute based on the matching result;
step 34: and performing a first analysis on the ice of the current fan blade based on the ice grade and the ice attribute of the same ice grade corresponding to the standard ice key attribute.
5. The machine learning based real-time deicing method for wind turbine generator set blades of claim 4, wherein predicting ice cube indices based on analysis results, determining a deicing scheme to achieve real-time deicing for target fan blades comprises:
step 41: the first analysis result is corresponding to a preset analysis table, and the index of the matched ice cubes is used as a first prediction index of the ice cubes corresponding to the fan blades;
step 42: determining a corresponding first deicing scheme based on a first prediction index, and simulating the first deicing scheme on the basis of the first deicing scheme in an intelligent management terminal, so that the first deicing scheme is adjusted, and a second deicing scheme is obtained;
step 43: and carrying out real-time deicing on the corresponding positions of the fan blades based on the second deicing scheme.
6. The machine learning based real-time deicing method for wind turbine generator set blades of claim 5, wherein determining a corresponding first deicing scheme based on the first predictive index, and simulating at the intelligent management terminal based on the scheme, thereby adjusting the first deicing scheme to obtain a second deicing scheme, comprises:
step 421: comparing the index range based on the first predictive index to an index scheme table;
if the first prediction index is in a first preset range of the index scheme table, working based on the blade rotation inertia force of the fan blade and the corresponding ultrasonic module to serve as a first deicing scheme of the current fan blade;
if the first prediction index is in a second preset range of the index scheme table, working based on the blade rotation inertia force of the fan blade and the corresponding laser module to serve as a first deicing scheme of the current fan blade;
otherwise, based on the blade rotation inertia force of the fan blade, the corresponding laser module and the ultrasonic module, the fan blade is used as a first deicing scheme of the current fan blade;
step 422: simulating in the intelligent management terminal based on the predicted first deicing scheme, and predicting deicing time of the corresponding fan blade;
wherein T is0 The influence time of the common influence factors; t (T)1 The deicing time of the fan blade when the first prediction index is in a first preset range of the index scheme table is set; t (T)2 The deicing time of the fan blade when the first prediction index is in a second preset range of the index scheme table is set; t (T)3 Deicing time of the fan blade when the first prediction index is out of a second preset range of the index scheme table; t (T)s The deicing time of the fan blade in the standard state is set; beta1 、β2 、β3 、β4 Time conversion coefficients for each of the common influencing factors; h is a1 The maximum ice coating thickness of the current fan blade is set; h is a2 Is the minimum ice thickness; h is a0 The thickness of the ice coating is used for influencing the change point of the deicing time; alpha1 A first impact index of ice coating thickness on deicing time; alpha2 A second impact index of ice coating thickness on deicing time; t is t1 The maximum environmental temperature of the current fan blade is the environmental; t is t2 Is the minimum ambient temperature; t is t0 The environmental temperature is the environmental temperature which influences the change point of the deicing time; mu (mu)1 A first impact index of ambient temperature on deicing time; mu (mu)2 A second impact index of ambient temperature on deicing time; v is the real-time working wind speed of the fan blade; vs The wind speed is the standard wind speed for the normal operation of the fan blade; l is the real-time liquid water content of the environment where the fan blade is located; l (L)s The standard liquid water content is the standard liquid water content of the environment where the fan blade is positioned; p (P)1 The heat release power of the fan blade under the influence of ultrasonic waves is provided; c (C)1 The heating time of the fan blade under the influence of ultrasonic waves is set; c is the specific heat capacity of ice; m is the mass of ice cubes on the target fan blade; omega is a temperature time conversion coefficient; p (P)2 The heat release power of the fan blade under the influence of laser is provided; c (C)2 The heating time of the fan blade under the influence of laser is set; epsilon1 、ε2 The influence weight of the current heating power on the current deicing time is given;
if the deicing time is within a first preset time range, determining that the current first deicing scheme is a second deicing scheme;
otherwise, data acquisition, algorithm matching and model establishment are carried out again, and a second deicing scheme is obtained.
7. The machine learning based real-time deicing method for wind turbine generator set blades of claim 5, wherein, based on the analysis result, predicting the ice cube index, determining the deicing scheme, and after implementing the real-time deicing for the target fan blade, further comprises: the method for verifying the deicing result specifically comprises the following steps:
step 01: when the real-time deicing task is displayed, acquiring real-time data information of a preset sensor again, and performing standardized processing;
step 02: comparing the real-time data information after standardized processing with standard state data corresponding to the current fan blade to realize result verification;
if the checking result is within the preset normal range, judging that the current deicing task is completed, and transmitting the completed result to the intelligent management terminal;
otherwise, judging that the current deicing task is not completed, reestablishing a first ice block coverage model based on real-time data information of the fan blades with not completed deicing, analyzing and predicting, and deicing again.
8. Real-time deicing system of wind generating set blade based on machine learning, its characterized in that includes:
and a data acquisition module: acquiring real-time data information of different parts of each fan blade in the target wind generating set based on a preset sensor;
and a model building module: processing the acquired real-time data information based on machine learning, and establishing a first ice block coverage model of each fan blade based on a preset algorithm;
analysis and prediction module: determining key ice cube attributes of corresponding fan blades based on the first ice cube covering model, and transmitting the key ice cube attributes to an intelligent management terminal for analysis;
deicing processing module: based on the analysis result, predicting the ice cube index, determining a deicing scheme, and realizing real-time deicing of the target fan blade.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310228069.9ACN116412091A (en) | 2023-03-09 | 2023-03-09 | Real-time deicing method and system for wind generating set blades based on machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310228069.9ACN116412091A (en) | 2023-03-09 | 2023-03-09 | Real-time deicing method and system for wind generating set blades based on machine learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116412091Atrue CN116412091A (en) | 2023-07-11 |
Family
ID=87054074
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310228069.9AWithdrawnCN116412091A (en) | 2023-03-09 | 2023-03-09 | Real-time deicing method and system for wind generating set blades based on machine learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116412091A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117404262A (en)* | 2023-11-24 | 2024-01-16 | 湖南防灾科技有限公司 | Control method and controller of fan air-heat deicing system based on fuzzy control |
| CN117910335A (en)* | 2023-12-08 | 2024-04-19 | 国家电投集团湖北电力有限公司风电分公司 | Anti-icing performance test method and system for anti-icing paint of wind power blade |
- 2023
- 2023-03-09CNCN202310228069.9Apatent/CN116412091A/ennot_activeWithdrawn
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117404262A (en)* | 2023-11-24 | 2024-01-16 | 湖南防灾科技有限公司 | Control method and controller of fan air-heat deicing system based on fuzzy control |
| CN117404262B (en)* | 2023-11-24 | 2024-06-04 | 湖南防灾科技有限公司 | Control method and controller of fan air-heat deicing system based on fuzzy control |
| CN117910335A (en)* | 2023-12-08 | 2024-04-19 | 国家电投集团湖北电力有限公司风电分公司 | Anti-icing performance test method and system for anti-icing paint of wind power blade |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109522627B (en) | Fan blade icing prediction method based on SCADA (Supervisory control and data acquisition) data | |
| CN109751206B (en) | Fan blade icing fault prediction method and device and storage medium | |
| CN116412091A (en) | Real-time deicing method and system for wind generating set blades based on machine learning | |
| CN106951997B (en) | Method and device for predicting fault of fan | |
| CN105508152B (en) | Method for constructing blade icing model, monitoring method and device for icing state | |
| CN109209790A (en) | A kind of wind electricity blade icing estimation method based on data modeling | |
| CN113994087B (en) | Methods and systems for controlling quantities of a wind turbine by selecting a controller via machine learning | |
| CN117407773B (en) | Digital twinning-based method, system and equipment for predicting icing state of fan blade | |
| CN117421926A (en) | Intelligent wind power equipment management and control method and system based on digital twin technology | |
| CN110991701A (en) | Wind power plant fan wind speed prediction method and system based on data fusion | |
| CN109494716A (en) | Wind power output power confidence interval prediction technique based on Bootstrap | |
| CN114909267A (en) | A kind of wind turbine blade icing detection method | |
| WO2024041409A1 (en) | Method and apparatus for determining representative wind generating set, and control method and apparatus | |
| CN119358428B (en) | Method and system for predicting power consumption of medium-temperature water chilling unit based on data assimilation | |
| CN107656091B (en) | A kind of wind measurement method and its system based on air-blower control sensor | |
| CN119825659A (en) | Icing detection method, device, storage medium and system for fan | |
| GB2555010B (en) | Determining loads on a wind turbine | |
| CN112211794B (en) | Cabin temperature abnormity early warning method, device, equipment and medium of wind turbine generator | |
| CN118886743A (en) | Offshore wind farm power calculation method, system and device | |
| TWI482041B (en) | Method and apparatus for evaluating efficiency of wind generator | |
| CN117633650A (en) | Marine wind turbine generator system fault prediction method and system based on digital twin | |
| CN115329639A (en) | Wind power station maintenance method, platform and electronic equipment | |
| CN112348700B (en) | A Line Capacity Prediction Method Combining SOM Clustering and IFOU Equation | |
| Bai et al. | Comparative differences of physical mechanism model and data-driven model on overhead transmission line icing thickness prediction | |
| Dang | Analysis of Factors Influencing Mountain Wind Power Generation Based on Grey Relational Analysis |
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 | ||
| WW01 | Invention patent application withdrawn after publication | Application publication date:20230711 | |
| WW01 | Invention patent application withdrawn after publication |