WO2023063887A2

WO2023063887A2 - Method and apparatus for predicting state of wind turbine blade, and device and storage medium therefor

Info

Publication number: WO2023063887A2
Application number: PCT/SG2022/050725
Authority: WO
Inventors: Weiyu CUI
Original assignee: Envision Digital International Pte. Ltd.; Shanghai Envision Digital Co., Ltd.
Priority date: 2021-10-14
Filing date: 2022-10-11
Publication date: 2023-04-20
Also published as: CN113847216B; CN113847216A; WO2023063887A3

Abstract

The present disclosure provides a method and apparatus for predicting a state of a wind turbine blade, a device and a storage medium, relating to the technical field of wind power generation. The method includes: acquiring an actual ambient parameter and an actual power value of a target wind turbine blade; acquiring an expected power value of the target wind turbine blade based on the actual ambient parameter; acquiring a power droop index based on the actual power value and the expected power value; acquiring an icing risk index of the target wind turbine blade based on a real-time ambient temperature in the actual ambient parameter; and predicting, based on the power droop index and the icing risk index, whether the target wind turbine blade is in an icing state or not. According to this method, whether the wind turbine blade is in the icing state or not is predicted based on the operation data and the ambient parameter of the wind turbine blade, and the accuracy of predicting whether the wind turbine blade is in the icing state or not is improved, and meanwhile, an additional sensor for the prediction of the icing state is not needed, such that application of the method for predicting the icing state is not limited but the method is widely applied.

Description

METHOD AND APPARATUS FOR PREDICTING STATE OF WIND TURBINE BLADE, AND DEVICE AND STORAGE MEDIUM THEREFOR

TECHNICAL FIELD

[0001] Embodiments of the present disclosure relates to the technical field of wind power generation, and in particular to a method and apparatus for predicting a state of a wind turbine blade, and a device and a storage medium therefor.

BACKGROUND

[0002] In a wind power generation scene, in a cold region, in order to improve the wind power generation efficiency and reduce the influence on the service life of wind power devices, whether a wind turbine blade is in an icing state or not needs to be detected and the fault of blade icing needs to be eliminated in time.

[0003] In some practices, it is generally determined whether a wind turbine blade is icy according to a preset icing temperature threshold in combination with temperature data of the wind turbine blade collected by various sensors or ultrasonic detection technologies, wherein the sensors are infrared sensors, optical fiber sensors, or the like.

[0004] However, in the above practices, whether the wind turbine blade is in an icing state or not is predicted only depending on the temperature data collected by the sensors arranged on each wind turbine blade, such that the prediction on the icing state of the wind turbine blade is poor, and the prediction accuracy is low.

SUMMARY

[0005] Embodiments of the present disclosure provides a method and apparatus for predicting a state of a wind turbine blade, and a device and a storage medium therefor. The application range of the detection technology is expanded while the accuracy of predicting the icing state of the wind turbine blade is ensured. The technical solution is as follows.

[0006] In one aspect of the embodiments of the present disclosure, a method for predicting a state of a wind turbine blade is provided. The method includes:

[0007] acquiring an actual ambient parameter and an actual power value of a taiget wind turbine blade; [0008] acquiring an expected power value of the target wind turbine blade based on the actual ambient parameter;

[0009] acquiring a power droop index based on the actual power value and the expected power value, the power droop index being indicative of a difference between the actual power value and the expected power value;

[0010] acquiring an icing risk index of the target wind turbine blade based on a real-time ambient temperature in the actual ambient parameter; and

[0011] predicting, based on the power droop index and the icing risk index, whether the target wind turbine blade is in an icing state or not.

[0012] In another aspect of the embodiments of the present disclosure, an apparatus for predicting a state of a wind turbine blade is provided. The apparatus includes:

[0013] an actual data acquiring module, configured to acquire an actual ambient parameter and an actual power value of a target wind turbine blade;

[0014] an expected power value acquiring module, configured to acquire an expected power value of the target wind turbine blade based on the actual ambient parameter;

[0015] a power droop index acquiring module, configured to acquire a power droop index based on the actual power value and the expected power value, the power droop index being indicative of a difference between the actual power value and the expected power value;

[0016] an icing risk index acquiring module, configured to acquire an icing risk index of the target wind turbine blade based on a real-time ambient temperature in the actual ambient parameter; and [0017] an icing state predicting module, configured to predict, based on the power droop index and the icing risk index, whether the target wind turbine blade is in an icing state or not.

[0018] In some embodiments, the icing state predicting module is configured to predict that the target wind turbine blade is in the icing state in response to the power droop index indicating that a power generation performance of the target wind turbine blade is lower than a performance threshold and the icing risk index indicating that an icing risk of the target wind turbine blade is higher than a risk threshold.

[0019] In some embodiments, the apparatus further includes:

[0020] a duration acquiring module, configured to acquire a duration of the target wind turbine blade in the icing state in response to predicting that the target wind turbine blade is in the icing state; and

[0021] an icing alarming module, configured to issue an icing warning in response to the duration being greater than a duration threshold. [0022] In some embodiments, the expected power value acquiring module is configured to acquire the expected power value output by a power prediction model by inputting the actual ambient parameter into the power prediction model; and

[0023] wherein the power prediction model is acquired by training based on an ambient parameter sample and a power value label corresponding to the ambient parameter sample.

[0024] In some embodiments, the ambient parameter sample and the power value label corresponding to the ambient parameter sample are acquired by pre-processing history ambient parameters and history power values corresponding to the history ambient parameters; and [0025] the pre-processing includes at least one of:

[0026] cleaning a null value of the history ambient parameters and the history power values corresponding to the history ambient parameters;

[0027] cleaning a dead value and interpolation data of the history ambient parameters and the histoiy power values corresponding to the history ambient parameters; and

[0028] cleaning low temperature data of the histoiy ambient parameters and the histoiy power values corresponding to the history ambient parameters, the low temperature data referring to history ambient parameters having a corresponding history ambient temperature lower than a temperature threshold and history power values corresponding to the history ambient parameters. [0029] In some embodiments, the power droop index is equal to a ratio of the actual power value to the expected power value.

[0030] In some embodiments, the icing risk index acquiring module is configured to acquire the icing risk index of the target wind turbine blade based on an actual temperature in the actual ambient parameter in combination with a temperature-icing risk index model; and

[0031] wherein the temperature-icing risk index model is a mathematical model constructed based on a sigmoid function.

[0032] In still another aspect of the embodiments of the present disclosure, a computer device is provided. The computer device includes a processor and a memoiy, wherein the memory stores at least one instruction, at least one program, a code set, or an instruction set stored thereon. The at least one instruction, the at least one program, the code set, or the instruction set, when loaded and executed by the processor, causes the processor to perform the method for predicting the state of the wind turbine blade according the above aspect.

[0033] In still yet another aspect of the embodiments of the present disclosure, a non-transitory computer-readable storage medium is provided. The storage medium stores one or more instructions thereon. The one or more instructions, when loaded and executed by a processor of a computer device, cause the computer device to perform the method for predicting the state of the wind turbine blade according the above aspect. [0034] In still yet another aspect of the embodiments of the present disclosure, a computer program product or a computer program is provided. The computer program product or the computer program includes one or more computer instructions stored in a non-transitory computer-readable storage medium. The one or more computer instructions, when loaded and executed by a processor of a computer device, cause the computer device to perform the method for predicting the state of the wind turbine blade according to the above aspect.

[0035] The technical solutions according to the present disclosure achieve the following beneficial effects:

[0036] Whether the wind turbine blade is in an icing state or not is comprehensively predicted according to the acquired power droop index and the icing risk index, wherein the power droop index is determined based on the actual power value and the expected power value of the wind turbine blade, and the icing risk index is determined by an ambient temperature, such that whether the wind turbine blade is in the icing state or not is predicted based on the operation data and the ambient parameter of the wind turbine blade, and the accuracy of predicting whether the wind turbine blade is in the icing state or not is improved, and meanwhile, an additional sensor for the prediction of the icing state is not needed, such that application of the method for predicting the icing state is not limited but the method is widely applied.

BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure and, together with the specification, serve to explain the principles of the present disclosure.

[0038] FIG. 1 illustrates a wind speed power scatter diagram according to an exemplary embodiment of the present disclosure;

[0039] FIG. 2 illustrates a schematic structural diagram of a system used in a method for predicting a state of a wind turbine blade according to an exemplary embodiment of the present disclosure;

[0040] FIG. 3 illustrates a flowchart of a method for predicting a state of a wind turbine blade according to an exemplary embodiment of the present disclosure;

[0041] FIG. 4 illustrates a flowchart of a method for predicting a state of a wind turbine blade according to an exemplary embodiment of the present disclosure;

[0042] FIG. 5 illustrates a timing diagram according to an exemplary embodiment of the present disclosure;

[0043] FIG. 6 illustrates a schematic diagram of a temperature-icing risk index model according to an exemplary embodiment of the present disclosure; [0044] FIG. 7 illustrates a schematic diagram of an icing warning flow according to an exemplary embodiment of the present disclosure;

[0045] FIG. 8 illustrates a block diagram of an apparatus for predicting a state of a wind turbine blade according to an exemplaiy embodiment of the present disclosure; and

[0046] FIG. 9 illustrates a structural block diagram of a computer device according to an example embodiment.

DETAILED DESCRIPTION

[0047] Reference is not made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different accompanying drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.

[0048] It is to be understood that the term "several" refers to one or more, and the term "a plurality of' refers to two or more. The term "and/or" describes the association relationship of the associated objects, and indicates that three relationships may be present. For example, A and/or B may indicate that: only A is present, both A and B are present, and only B is present. The symbol "/" usually indicates an "or" relationship between the associated objects.

[0049] As the wind power generator set operates in cold regions where wind turbines are influenced by meteorological conditions such as frost ice, silver thaw and wet snow, blades are easily prone to icing, and thus a series of consequences may be triggered, and the following risks may be probably caused:

[0050] 1) The icy wind turbine blade profile is changed, such that the wind energy capture capacity is reduced, and in addition, an ice layer is attached to the blade, such that the energy required for the rotation of the blade is increased, and finally, the power loss of wind power generation is caused.

[0051] FIG. 1 illustrates a wind speed power scatter diagram according to an exemplary embodiment of the present disclosure. As shown in FIG. 1, in a case that a wind turbine blade is in a normal operating state, a relationship between a wind speed and a wind turbine power conforms to a speed-power curve 110; in a case that the wind turbine blade is in an icing state, as the wind turbine blade profile is changed, the wind energy capture capacity of the wind turbine blade is reduced, such that the wind turbine power is reduced. As shown in FIG. 1, a point in a region 120 is the wind turbine power corresponding to the wind speed in the icing state; wherein the wind turbine power is an active power.

[0052] 2) After the wind turbine blade is icy, part of structural parameters of the blade are directly changed, such that inherent modal parameters of the blade are influenced, and the blade is induced to break.

[0053] 3) After the wind turbine blade is icy with ice buildups for some time, the ice layer breaks and flies out due to its weight, and may hit inspection personnel in the wind field, thereby causing casualty.

[0054] Therefore, the in-time detection and elimination on the fault of blade icing has an important significance for prolonging the service life of wind power devices and preventing major safety accidents.

[0055] In the operation and maintenance process of wind power generation stations in cold regions, whether the wind turbine blades of the wind power generator set are in an icing state or not needs to be detected. The present disclosure provides a method for predicting a state of a wind turbine blade. The application range of the detection technology is expanded while the accuracy of predicting the icing state of the wind turbine blade is ensured. For the sake of easy understanding, the terms referred to in the embodiments of the present disclosure are explained below.

[0056] A supervisory control and data acquisition (SCADA) system is provided.

[0057] The SCADA system is a distributed control system (DCS) and an automatic power supervisory control system based on a computer, and is applicable to the fields of data collection, supervisory control, process control, and the like in the fields of electric power, metallurgy, petroleum, chemical industry, gas, railways, and the like.

[0058] FIG. 2 illustrates a schematic structural diagram of a system corresponding to a method for predicting a state of a wind turbine blade according to an exemplary embodiment of the present disclosure. As shown in FIG. 2, the system includes: a wind power generation station 201 and a supervisory control platform 202.

[0059] The wind power generation station 201 includes a plurality of wind power generator sets, wherein each of the wind power generator sets includes wind turbine blades and a cabin. In the embodiment of the present disclosure, the wind power generation station 201 may be provided with a plurality of sensors configured to collect data of classified power generation stations. Illustratively, the sensors may include a temperature sensor, a wind speed sensor, and the like, and are configured to collect parameters such as an ambient temperature and a wind speed in the wind power generation station 201, and send the collected data to the supervisory control platform 202. [0060] The wind power generation station 201 is connected to the supervisory control platform 202 over a wired or wireless network.

[0061] The supervisory control platform 202 is a computer device having functions of storing data sent by the wind power generation station 201, processing the data, generating an alarming record, and the like, wherein the computer device may be a server or a server cluster formed by a plurality of servers or a cloud server, or the computer device may also be implemented as a terminal. The present disclosure is not limited to the embodiments of the computer device.

[0062] For convenience of description, in the method embodiments described below, the supervisory control platform 202 is described as a computer device is equivalent to illustrating the method for predicting the state of the wind turbine blade according to the present disclosure.

[0063] FIG. 3 illustrates a flowchart of a method for predicting a state of a wind turbine blade according to an exemplary embodiment of the present disclosure, wherein the method may be applicable to a computer device, and the computer device may be implemented as the supervisory control platform shown in FIG. 2. As shown in FIG. 3, the method includes the following steps.

[0064] In step 310, an actual ambient parameter and an actual power value of a target wind turbine blade are acquired.

[0065] In some embodiments, the target wind turbine blade is any one of a plurality of wind turbine blades in a wind field station, and the actual ambient parameter and the actual power value of the target wind turbine blade are acquired based on data collected by a SCADA system of the wind turbine, the SDCADA system being configured to implement functions such as data collection, device control, measurement, and parameter adjustment.

[0066] In step 320, an expected power value of the target wind turbine blade is acquired based on the actual ambient parameter.

[0067] The expected power value is indicative of a power value that is generated by the target wind turbine blade working normally under the ambient parameter in a non-icing state.

[0068] In step 330, a power droop index is acquired based on the actual power value and the expected power value, wherein the power droop index is indicative of a difference between the actual power value and the expected power value.

[0069] Generally, due to the ambient factors, a difference is often present between the actual power value and the expected power value, and thus a power droop index is set in the embodiment of the present disclosure to measure the magnitude of the difference between the actual power value and the expected power value.

[0070] In step 340, an icing risk index of the target wind turbine blade is acquired based on a realtime ambient temperature in the actual ambient parameter. [0071] The icing risk index is indicative of a probability that the target wind turbine blade is icy at the ambient temperature, and in general, the lower the ambient temperature is, the higher the icing risk index of the target wind turbine blade is, and the higher the ambient temperature is, the lower the icing risk index of the target wind turbine blade is.

[0072] In step 350, whether the target wind turbine blade is in an icing state or not is predicted based on the power droop index and the icing risk index.

[0073] In summary, in the method for predicting the state of the wind turbine blade according to the embodiments of the present disclosure, whether the wind turbine blade is in an icing state or not is comprehensively predicted according to the acquired power droop index and the icing risk index, wherein the power droop index is determined based on the actual power value and the expected power value of the wind turbine blade, and the icing risk index is determined by an ambient temperature, such that whether the wind turbine blade is in the icing state or not is predicted based on the operation data and the ambient parameter of the wind turbine blade, and the accuracy of predicting whether the wind turbine blade is in the icing state or not is improved, and meanwhile, an additional sensor for the prediction of the icing state is not needed, such that application of the method for predicting the icing state is not limited but the method is widely applied.

[0074] FIG. 4 illustrates a flowchart of a method for predicting a state of a wind turbine blade according to an exemplary embodiment of the present disclosure, wherein the method may be applicable to a computer device, and the computer device may be implemented as the supervisory control platform shown in FIG. 2. As shown in FIG. 4, the method includes the following steps. [0075] In step 410, an actual ambient parameter and an actual power value of a target wind turbine blade are acquired.

[0076] In step 420, an expected power value of the target wind turbine blade is acquired based on the actual ambient parameter.

[0077] In some embodiments, the expected power value output by a power prediction model is acquired by inputting the actual ambient parameter into the power prediction model.

[0078] The power prediction model is acquired by training based on an ambient parameter sample and a power value label corresponding to the ambient parameter sample.

[0079] In some embodiments, the power prediction model is a model built based on a regression model. Illustratively, the regression model is a LightGBM model.

[0080] The ambient parameter sample and the power value label corresponding to the ambient parameter sample are acquired by pre-processing history ambient parameters and history power values corresponding to the history ambient parameters. [0081] The history ambient parameters and the history power values corresponding to the history ambient parameters are acquired based on data collected by an SC ADA system of the wind turbine. A time period corresponding to the history data (the history ambient parameters and the history power values corresponding to the history ambient parameters) is set based on actual requirements. For example, the time period is a month, a quarter, a year, or the like. The present disclosure is not limited to the time period for acquiring the history data.

[0082] The power value label is the history power value corresponding to the ambient parameter sample.

[0083] Pre-processing the history ambient parameters and the history power values corresponding to the history ambient parameters includes at least one of:

[0084] cleaning a null value of the history ambient parameters and the history power values corresponding to the history ambient parameters;

[0085] cleaning a dead value and interpolation data of the history ambient parameters and the history power values corresponding to the history ambient parameters; and

[0086] cleaning low temperature data of the history ambient parameters and the history power values corresponding to the history ambient parameters, the low temperature data referring to history ambient parameters having a corresponding history ambient temperature lower than a temperature threshold and history power values corresponding to the history ambient parameters. [0087] The low temperature data are cleaned such that the power prediction model is capable of learning the normal operation rule of the wind turbine blade in a normal state (namely, in a nonicing state).

[0088] In some embodiments, the process of acquiring the ambient parameter includes:

[0089] sequencing at least two ambient attributes by using a tree model based on the influence degree regarding whether the wind turbine blade is in an icing state or not to acquire a sequencing result;

[0090] acquiring a target ambient attribute from the at least two ambient attributes based on the sequencing result; and

[0091] acquiring a parameter corresponding to the target ambient attribute as an ambient parameter.

[0092] The number of acquired target ambient attributes is acquired according to actual requirements, that is, a more important ambient attribute of the ambient attributes is acquired according to actual requirements and is taken as the target ambient attribute, and a parameter corresponding to the target ambient attribute is acquired and is taken as the ambient parameter. In some embodiments, the target ambient attribute includes a pitch angle, a turbulence value, a temperature, and a cabin position; and accordingly, the ambient parameter includes parameters such as a wind speed value, a pitch angle value, a temperature value, and cabin position coordinates.

[0093] In some embodiments, training the power prediction model includes:

[0094] acquiring a prediction result corresponding to the ambient parameter sample by inputting the ambient parameter sample into the power prediction model, wherein the prediction result includes a predicted power value;

[0095] calculating a loss function value based on the power value label corresponding to the ambient parameter sample and the prediction result corresponding to the ambient parameter sample; and

[0096] performing a parameter update on the power prediction model based on the loss function value.

[0097] Since the accuracy of the power value predicted by the power prediction model during application is ensured only in a case that the prediction result (namely, the predicted power value) acquired by the power prediction model based on the ambient parameter sample is the same as or similar to the power value label corresponding to the ambient parameter sample, training needs to be performed multiple times in the process for training the power prediction model, and each of the parameters in the power prediction model is updated iteratively until the power prediction model converges.

[0098] In step 430, a power droop index is acquired based on the actual power value and the expected power value, wherein the power droop index is indicative of a difference between the actual power value and the expected power value.

[0099] The power droop index is equal to the ratio of the actual power value to the expected power value, which is expressed as follows:

. , Actual power value Power droop

¹ index = . j - ; — Predicted power value

[00100] In step 440, an icing risk index of the target wind turbine blade is acquired based on a real-time ambient temperature in the actual ambient parameter.

[00101] To be specific, the ambient temperature reflects the influence of the icing state on the wind turbine power of the target wind turbine blade. FIG. 5 illustrates a timing diagram according to an exemplary embodiment of the present disclosure, wherein portion A of FIG. 5 illustrates a timeambient temperature timing diagram, and portion B of FIG. 5 illustrates a time-wind turbine power timing diagram. As seen from the timing diagram of portion A and the timing diagram of portion B in FIG. 5, in a case that the ambient temperature is lower than a specified temperature threshold, the wind turbine blade is in an icing state, and in the icing state, the wind turbine blade is in an abnormal operating state due to the influence of icing, such that the power of the wind turbine corresponding to the wind turbine blade is reduced. As shown in FIG. 5, during a first time period 510, the wind turbine blade is in a normal operating state, during a second time period 520, the wind turbine blade is in an icing state, and during a third time period 530, the wind turbine blade is reverted to be in a normal operating state. Therefore, the icing risk index of a wind turbine blade is determined based on changes in the ambient temperature.

[00102] In some embodiments, the icing risk index of the target wind turbine blade is acquired based on a real-time ambient temperature in the actual ambient parameter according to history experience.

[00103] The history experience is working experience or expert experience, for example, in a case that the ambient temperature is 0 degrees or below 0 degrees, the icing risk is high, and as the ambient temperature decreases, the icing risk index gradually increases.

[00104] In some embodiments, in order to improve the accuracy of acquiring the icing risk index, the icing risk index of the target wind turbine blade is acquired based on the actual temperature in the actual ambient parameter in combination with a temperature-icing risk index model.

[00105] The temperature-icing risk index model is a mathematical model constructed based on a sigmoid function.

[00106] FIG. 6 illustrates a schematic diagram of a temperature-icing risk index model according to an exemplary embodiment of the present disclosure. As shown in FIG. 6, the temperature-icing risk index model (a curve 610) is configured to determine the icing risk index (ordinate) according to the ambient temperature (abscissa), and a mathematical formula of the temperature-icing risk index model is expressed as follows: y = sigmoid(-x)

[00107] In the formula, x represents the ambient temperature, and y represents the icing risk index. [00108] In step 450, the target wind turbine blade is predicted to be in the icing state in response to the power droop index indicating that a power generation performance of the target wind turbine blade is lower than a performance threshold and the icing risk index indicating that an icing risk of the target wind turbine blade is higher than a risk threshold.

[00109] Except the scenario where the target wind turbine blade is predicted to be in an icing state under the case that the power droop index indicates that a power generation performance of the target wind turbine blade is lower than a performance threshold and the icing risk index indicates that an icing risk of the target wind turbine blade is higher than a risk threshold, the target wind turbine blade is predicted to be in a non-icing state under other cases.

[00110] To be specific, the target wind turbine blade is predicted to be in the non-icing state in response to the power droop index indicating that a power generation performance of the target wind turbine blade is lower than a performance threshold and the icing risk index indicating that an icing risk of the target wind turbine blade is lower than a risk threshold.

[00111] In some embodiments, the target wind turbine blade is predicted to be in the non-icing state in response to the power droop index indicating that a power generation performance of the target wind turbine blade is higher than a performance threshold and the icing risk index indicating that an icing risk of the target wind turbine blade is higher than a risk threshold.

[00112] In some embodiments, the target wind turbine blade is predicted to be in the non-icing state in response to the power droop index indicating that a power generation performance of the target wind turbine blade is lower than a performance threshold and the icing risk index indicating that an icing risk of the target wind turbine blade is lower than a risk threshold.

[00113] Since various power reduction factors for wind turbine blades are present (namely, factors causing the power of wind turbine blades to be reduced), in order to prevent the case that whether the target wind turbine blade is in an icing state or not is wrongly determined because the target wind turbine blade is in a power reducing state due to non-icing power reduction factors, in some embodiments, in the case that the target wind turbine blade is predicted to be in the icing state, a factor causing the target wind turbine blade to be in the power reducing state is acquired, and in the case that the factor causing the target wind turbine blade to be in the power reducing state is determined to be the non-icing power reduction, the target wind turbine blade is predicted to be in the icing state.

[00114] The non-icing power reduction factors include turbulence, over-temperature of the generator, over-temperature of the gearbox, and the like; and different power reduction factors correspond to different power-reducing determination mechanisms. For example, whether the factor causing the wind turbine blade to be in the power reducing state is turbulence or not is determined by detecting whether the wind speed fluctuation exceeds a wind speed fluctuation threshold or not; whether the factor causing the wind turbine blade to be in the power reducing state is over-temperature of the generator or not is determined by detecting whether the temperature of the generator exceeds a temperature threshold of the generator; and whether the factor causing the wind turbine blade to be in the power reducing state is over-temperature of the gearbox is determined by detecting whether the temperature of the gearbox exceeds a temperature threshold of the gearbox.

[00115] In some embodiments, a duration of the target wind turbine blade in the icing state is acquired in response to predicting that the target wind turbine blade is in the icing state; and [00116] an icing warning is issued in response to the duration being greater than a duration threshold. [00117] Since the wind turbine blade is continuously in the icing state, in some embodiments, a duration of the wind turbine blade in the icing state is acquired by a sliding window approach, and the process includes: taking a marked icing point as a starting point, acquiring a length of a time period including the icing point by a sliding window, and determining the time period including the icing point as the duration of the wind turbine blade in the icing state, wherein the marked icing point is the starting point of the icing state determined based on the above method for predicting the icing state; and in the case that the duration of the icing state is greater than the duration threshold, an icing warning is issued so as to indicate related personnel to inspect and maintain the wind turbine blade.

[00118] In summary, in the method for predicting the state of the wind turbine blade according to the embodiments of the present disclosure, whether the wind turbine blade is in an icing state or not is comprehensively predicted according to the acquired power droop index and the icing risk index, wherein the power droop index is determined based on the actual power value and the expected power value of the wind turbine blade, and the icing risk index is determined by an ambient temperature, such that whether the wind turbine blade is in the icing state or not is predicted based on the operation data and the ambient parameter of the wind turbine blade, and the accuracy of predicting whether the wind turbine blade is in the icing state or not is improved, and meanwhile, an additional sensor for the prediction of the icing state is not needed, such that application of the method for predicting the icing state is not limited but the method is widely applied.

[00119] FIG. 7 illustrates a schematic diagram of an icing warning flow according to an exemplary embodiment of the present disclosure. As shown in FIG. 6, the icing warning process includes the following steps.

[00120] In step 710, operation data of the wind turbine blade are acquired, wherein the operation data includes an actual ambient parameter and an actual power value.

[00121] In step 720, the power generation performance of the wind turbine blade is evaluated.

[00122] The evaluation on the power generation performance corresponds to the process of calculating the power droop index in the embodiments of FIGS. 3 and 4, and is not described herein in detail.

[00123] In step 730, icing risk assessment is performed on the wind turbine blade.

[00124] The icing risk assessment corresponds to the process of acquiring the icing risk index in the embodiments of FIGS. 3 and 4, which is not described herein in detail.

[00125] In step 740, whether the wind turbine blade is in a state with low power generation performance and high icing risk or not is determined; in the case that the wind turbine blkade in in the state with low power generation performance and high icing risk, step 750 is performed; and otherwise, the wind turbine blade is determined to be in a non-icing state.

[00126] In step 750, whether other power reduction factors are excluded or not is determined; in the case that the other power reduction factors are excluded, step 760 is performed; and otherwise, the process ends.

[00127] In step 760, whether the duration of icing is greater than the time threshold for icing or not is determined; in the case that the duration of icing is greater than the time threshold, step 770 is performed; and otherwise, the process ends.

[00128] In step 770, an icing warning is issued.

[00129] In summary, in the method for predicting the state of the wind turbine blade according to the embodiments of the present disclosure, whether the wind turbine blade is in an icing state or not is comprehensively predicted according to the acquired power droop index and the icing risk index, wherein the power droop index is determined based on the actual power value and the expected power value of the wind turbine blade, and the icing risk index is determined by an ambient temperature, such that whether the wind turbine blade is in the icing state or not is determined based on the operation data and the ambient parameter of the wind turbine blade, and the accuracy of predicting whether the wind turbine blade is in the icing state or not is improved, and meanwhile, an additional sensor for the prediction of the icing state is not needed, such that application of the method for predicting the icing state is not limited but the method is widely applied.

[00130] FIG. 8 illustrates a block diagram of an apparatus for predicting a state of a wind turbine blade according to an exemplary embodiment of the present disclosure. As shown in FIG. 8, the apparatus includes:

[00131] an actual data acquiring module 810, configured to acquire an actual ambient parameter and an actual power value of a target wind turbine blade;

[00132] an expected power value acquiring module 820, configured to acquire an expected power value of the target wind turbine blade based on the actual ambient parameter;

[00133] a power droop index acquiring module 830, configured to acquire a power droop index based on the actual power value and the expected power value, the power droop index being indicative of a difference between the actual power value and the expected power value;

[00134] an icing risk index acquiring module 840, configured to acquire an icing risk index of the target wind turbine blade based on a real-time ambient temperature in the actual ambient parameter; and

[00135] an icing state predicting module 850, configured to predict, based on the power droop index and the icing risk index, whether the target wind turbine blade is in an icing state or not. [00136] In some embodiments, the icing state predicting module 850 is configured to predict that the target wind turbine blade is in the icing state in response to the power droop index indicating that a power generation performance of the target wind turbine blade is lower than a performance threshold and the icing risk index indicating that an icing risk of the target wind turbine blade is higher than a risk threshold.

[00137] In some embodiments, the apparatus further includes:

[00138] a duration acquiring module, configured to acquire a duration of the target wind turbine blade in the icing state in response to predicting that the target wind turbine blade is in the icing state; and

[00139] an icing alarming module, configured to issue an icing warning in response to the duration being greater than a duration threshold.

[00140] In some embodiments, the expected power value acquiring module 820 is configured to acquire the expected power value output by a power prediction model by inputting the actual ambient parameter into the power prediction model,

[00141] wherein the power prediction model is acquired by training based on an ambient parameter sample and a power value label corresponding to the ambient parameter sample.

[00142] In some embodiments, the ambient parameter sample and the power value label corresponding to the ambient parameter sample are acquired by pre-processing history ambient parameters and histoiy power values corresponding to the history ambient parameters; and [00143] the pre-processing includes at least one of:

[00144] cleaning a null value of the history ambient parameters and the history power values corresponding to the history ambient parameters;

[00145] cleaning a dead value and interpolation data of the history ambient parameters and the history power values corresponding to the histoiy ambient parameters; and

[00146] cleaning low temperature data of the history ambient parameters and the history power values corresponding to the history ambient parameters, the low temperature data referring to history ambient parameters having a corresponding history ambient temperature lower than a temperature threshold and history power values corresponding to the history ambient parameters. [00147] In some embodiments, the interpolation data is artificial difference data.

[00148] In some embodiments, the power droop index is equal to a ratio of the actual power value to the expected power value.

[00149] In some embodiments, the icing risk index acquiring module 840 is configured to acquire the icing risk index of the target wind turbine blade based on an actual temperature in the actual ambient parameter in combination with a temperature-icing risk index model. [00150] In summary, in the method for predicting the state of the wind turbine blade according to the embodiments of the present disclosure, whether the wind turbine blade is in an icing state or not is comprehensively predicted according to the acquired power droop index and the icing risk index, wherein the power droop index is determined based on the actual power value and the expected power value of the wind turbine blade, and the icing risk index is determined by an ambient temperature, such that whether the wind turbine blade is in the icing state or not is predicted based on the operation data and the ambient parameter of the wind turbine blade, and the accuracy of predicting whether the wind turbine blade is in the icing state or not is improved, and meanwhile, an additional sensor for the prediction of the icing state is not needed, such that application of the method for predicting the icing state is not limited but the method is widely applied.

[00151] FIG. 9 illustrates a structural block diagram of a computer device 900 according to an exemplary embodiment. The computer device may be implemented as the supervisory control platform in the above solution of the present disclosure. The computer device 900 includes a central processing unit (CPU) 901, a system memory 904 including a random access memory (RAM) 902 and a read-only memory (ROM) 903, and a system bus 905 for connecting the system memory 904 to the CPU 901. The computer device 900 further includes a basic input/output system (I/O system) 906 configured to facilitate information transfer between components within the computer, and a mass storage device 907 configured to store an operating system 913, application programs 914, and other program modules 915.

[00152] The basic input/output system 906 includes a display 908 configured to display information and an input device 909 configured to input information by a user, such as a mouse and a keyboard. The display 908 and the input device 909 are connected to the CPU 901 by the input/output controller 910 connected to the system bus 905. The basic input/output system 906 may further include an input/output controller 910 configured to receive and process inputs from a plurality of other devices, such as a keyboard, a mouse, or electronic stylus. Similarly, the input/output controller 910 further provides output devices configured to output data or signals onto a display screen and a printer, or other type of output devices.

[00153] The mass storage device 907 is connected to the CPU 901 by a mass storage controller (not shown) connected to the system bus 905. The mass storage device 907 and computer readable media associated therewith provide non-volatile storage for the computer device 900. That is, the mass storage device 907 may include a computer-readable medium (not shown) such as a hard disk or a compact disc read-only memory (CD-ROM) drive.

[00154] Generally, the computer-readable medium may include a computer storage medium and a communication medium. The computer storage media include a volatile and non-volatile, removable and non-removable medium implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules or other data. The computer storage media include a RAM, a ROM, an erasable programmable read only memory (EPROM), an electrically-erasable programmable read-only memory (EEPROM), a flash memory or other solid state storage techniques, a CD-ROM, a digital versatile disc (DVD), or other optical storage, magnetic cassette, magnetic tape, magnetic disc storage or magnetic storage devices. It is appreciated by those skilled in the art that the computer storage medium is not limited to the foregoing. The system memory 904 and the mass storage device 907 described above may be collectively referred to as memories.

[00155] According to various embodiments of the present disclosure, the computer device 900 may be further connected to a remote computer on a network over the network, such as the Internet, for running. That is, the computer device 900 may be connected to the network 912 by a network interface unit 911 connected to the system bus 905, or may be connected to other types of networks or remote computer systems (not shown) using the network interface unit 911.

[00156] The memory further includes one or more programs, the one or more programs being stored in the memory, and the CPU 901 implements all or part of the steps of the method shown in FIG. 3, FIG. 4 or FIG. 7 by loading and running the one or more programs.

[00157] Those skilled in the art appreciate that in one or more of the above embodiments, the functions described in the embodiments of the present disclosure may be implemented in hardware, software, firmware, or any combination thereof. The functions, when implemented in software, may be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. The computer-readable medium includes a computer storage medium and a communication medium, wherein the communication medium includes any medium that facilitates transfer of a computer program. The storage medium is any available medium that is accessible by a general purpose or special purpose computer.

[00158] An embodiment of the present disclosure further provides a non-transitory computer- readable storage medium storing at least one instruction, at least one program, a code set, or an instruction set stored thereon. The at least one instruction, the at least one program, the code set, or the instruction set, when loaded and executed by a processor of a computer device, causes the computer device to perform the method for predicting the state of the wind turbine blade. For example, the computer-readable storage medium may be a ROM, a RAM, a CD-ROM, a magnetic tape, a soft disk, an optical data storage device, and the like.

[00159] An embodiment of the present disclosure further provides a computer program product or a computer program including one or more computer instructions. The one or more computer instructions are stored in a non-transitory computer-readable storage medium. The one or more computer instructions, when loaded and executed by a processor of a computer device, cause the computer device to perform the method for predicting the state of the wind turbine blade according to the above various alternative embodiments.

[00160] Other embodiments of the present disclosure are apparent to those skilled in the art from consideration of the specification and practice of the present invention disclosed herein. The present disclosure is intended to cover any variations, uses, or adaptations of the present disclosure following the general principles of the present disclosure and including known common knowledge or customary technical means undisclosed in the art of the present disclosure. The specification and embodiments are only considered as exemplary, and a true scope and spirit of the present disclosure are indicated in the following claims.

[00161] It is understood that the present disclosure is not limited to the precise arrangements that have been described above and shown in the accompanying drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

CLAIMS What is claimed is:

1. A method for predicting a state of a wind turbine blade, comprising: acquiring an actual ambient parameter and an actual power value of a target wind turbine blade; acquiring an expected power value of the target wind turbine blade based on the actual ambient parameter; acquiring a power droop index based on the actual power value and the expected power value, the power droop index being indicative of a difference between the actual power value and the expected power value; acquiring an icing risk index of the target wind turbine blade based on a real-time ambient temperature in the actual ambient parameter; and predicting, based on the power droop index and the icing risk index, whether the target wind turbine blade is in an icing state or not.

2. The method according to claim 1, wherein predicting, based on the power droop index and the icing risk index, whether the target wind turbine blade is in the icing state or not comprises: predicting that the target wind turbine blade is in the icing state in response to the power droop index indicating that a power generation performance of the target wind turbine blade is lower than a performance threshold and the icing risk index indicating that an icing risk of the target wind turbine blade is higher than a risk threshold.

3. The method according to claim 2, further comprising: acquiring a duration of the target wind turbine blade in the icing state in response to predicting that the target wind turbine blade is in the icing state; and issuing an icing warning in response to the duration being greater than a duration threshold.

4. The method according to claim 1, wherein acquiring the expected power value of the target wind turbine blade based on the actual ambient parameter comprises: acquiring the expected power value output by a power prediction model by inputting the actual ambient parameter into the power prediction model, wherein the power prediction model is acquired by training based on an ambient parameter sample and a power value label corresponding to the ambient parameter sample.

5. The method according to claim 4, wherein the ambient parameter sample and the power value label corresponding to the ambient parameter sample are acquired by pre-processing history ambient parameters and history power values corresponding to the history ambient parameters; and the pre-processing comprises at least one of: cleaning a null value of the history ambient parameters and the history power values corresponding to the history ambient parameters; cleaning a dead value and interpolation data of the history ambient parameters and the history power values corresponding to the history ambient parameters; and cleaning low temperature data of the history ambient parameters and the history power values corresponding to the history ambient parameters, the low temperature data referring to history ambient parameters having a corresponding history ambient temperature lower than a temperature threshold and history power values corresponding to the histoiy ambient parameters.

6. The method according to claim 1, wherein the power droop index is equal to a ratio of the actual power value to the expected power value.

7. The method according to claim 1, wherein acquiring the icing risk index of the target wind turbine blade based on the real-time ambient temperature in the actual ambient parameter comprises: acquiring the icing risk index of the target wind turbine blade based on an actual temperature in the actual ambient parameter in combination with a temperature-icing risk index model; and wherein the temperature-icing risk index model is a mathematical model constructed based on a sigmoid function.

8. An apparatus for predicting a state of a wind turbine blade, comprising: an actual data acquiring module, configured to acquire an actual ambient parameter and an actual power value of a target wind turbine blade; an expected power value acquiring module, configured to acquire an expected power value of the target wind turbine blade based on the actual ambient parameter; a power droop index acquiring module, configured to acquire a power droop index based on the actual power value and the expected power value, the power droop index being indicative of a difference between the actual power value and the expected power value; an icing risk index acquiring module, configured to acquire an icing risk index of the target wind turbine blade based on a real-time ambient temperature in the actual ambient parameter; and an icing state predicting module, configured to predict, based on the power droop index and the icing risk index, whether the target wind turbine blade is in an icing state or not.

9. A computer device, comprising a processor, and a memory configured to store at least one instruction, at least one program, a code set, or an instruction set stored thereon, wherein the at least one instruction, the at least one program, the code set, or the instruction set, when loaded and executed by the processor, causes the processor to perform the method for predicting the state of the wind turbine blade according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium, storing at least one instruction, at least one program, a code set, or an instruction set stored thereon, wherein the at least one instruction, the at least one program, the code set, or the instruction set, when loaded and executed by a processor of a computer device, causes the computer device to perform the method for predicting the state of the wind turbine blade according to any one of claims 1 to 7.