Method and system for inhibiting icing of blades of wind generating set

Abstract

The invention discloses a method and a system for inhibiting icing of blades of a wind generating set, and relates to the technical field of wind power generation, and the method comprises the following steps: obtaining weather data and environment state data of an area where the wind generating set is located, a current icing state and icing thickness of each sensor in an ultrasonic sensor array; performing inverse distance weighting processing according to the icing thickness at each sensor in the ultrasonic sensor array to obtain an icing area; performing icing prediction according to the weather data and the environment state data to obtain an icing prediction probability; determining a corresponding deicing mode according to at least two of the icing prediction probability, the current icing state and the icing area; and according to at least two of the deicing mode, the icing thickness and the icing area, controlling the ultrasonic device and / or the heating device to perform icing inhibition on the blade.

Description

Technical Field

[0001] This application relates to the technical field of wind power generation, and in particular to a method and system for suppressing icing on the blades of a wind turbine generator. Background Technology

[0002] As an important form of clean energy, wind power faces the severe challenge of blade icing in cold or high-altitude regions. In low-temperature and high-humidity environments, icing easily forms on blade surfaces, leading to deterioration of aerodynamic performance, imbalance in mass distribution, and increased vibration. In severe cases, this can cause blade structural damage, reduced power generation efficiency, or even turbine shutdown, directly impacting the economics and safety of wind power generation. Currently, countermeasures against blade icing mainly include regular manual inspections, microwave de-icing, and temperature-triggered electric heating de-icing. These methods often overlook the coupled effects of multiple meteorological parameters, have limited ability to assess different blade icing conditions, and struggle to achieve precise, zoned, and graded de-icing control. Furthermore, existing methods primarily focus on post-icing, lacking early prediction and prevention capabilities for icing risks, and are unable to take suppressive measures before or after icing occurs. The overall system's adaptability and energy efficiency still have significant room for improvement.

[0003] Therefore, improving the adaptability and energy efficiency ratio of wind turbine blade de-icing is an urgent problem to be solved. Summary of the Invention

[0004] The main objective of this application is to provide a method and system for suppressing icing on wind turbine blades, aiming to solve the technical problem of how to improve the adaptability and energy efficiency ratio of wind turbine blade de-icing.

[0005] To achieve the above objectives, this application proposes a method for suppressing icing on wind turbine blades. This method is applied to a wind turbine blade icing suppression system, which includes an ultrasonic sensor array, an ultrasonic device, and a heating device. The method comprises: Acquire weather and environmental data of the area where the wind turbine is located, the current icing status, and the icing thickness at each sensor in the ultrasonic sensor array; The icing region is obtained by inverse distance weighting processing based on the icing thickness at each sensor in the ultrasonic sensor array. Based on the weather data and the environmental condition data, icing prediction is performed to obtain the icing prediction probability; Based on the icing prediction probability, the current icing state, and at least two of the icing regions, a corresponding de-icing mode is determined. The de-icing mode includes an icing suppression mode, a rapid de-icing mode, a zoned de-icing mode, and an ultrasonic de-icing mode. The blades are icing suppressed according to the de-icing mode, the icing thickness, and at least two of the ultrasonic devices and / or heating devices in the icing area.

[0006] In one embodiment, the step of obtaining the icing region by performing inverse distance weighting processing based on the icing thickness at each sensor in the ultrasonic sensor array includes: Obtain the position coordinates of each sensor in the ultrasonic sensor array; Grid points are generated on the surface of the wind turbine blades according to a preset spatial resolution, and the spatial distance between each grid point and the position coordinates of each sensor is calculated. The ice thickness at each sensor in the ultrasonic sensor array is weighted and interpolated based on the spatial distance to obtain the estimated ice thickness at the grid points. The icing area is determined based on the estimated icing thickness.

[0007] In one embodiment, the step of performing weighted interpolation processing on the ice thickness at each sensor in the ultrasonic sensor array based on the spatial distance to obtain the estimated ice thickness at the grid points includes: Obtain the measurement confidence parameters of each sensor in the ultrasonic sensor array; For each grid point, calculate the Euclidean distance between the grid point and the sensor, and calculate the initial weighting factor based on the Euclidean distance; The initial weighting factor is corrected based on the measured confidence parameter to obtain the target weighting factor; The ice thickness values ​​at each sensor are weighted and summed according to the target weighting factor to obtain the initial estimated ice thickness at the grid points. Spatially smooth the initial estimated icing thickness at adjacent grid points to obtain the estimated icing thickness.

[0008] In one embodiment, the step of obtaining the current icing state and the icing thickness at each sensor in the ultrasonic sensor array includes: Acquire echo signals and blade surface temperatures; The propagation time difference of the ultrasonic wave from transmission to reception is calculated based on the echo signal and the blade surface temperature. The echo signal is collected by an ultrasonic sensor array installed inside the wind turbine blade. The current icing state and the icing thickness at each sensor in the ultrasonic sensor array are determined based on the propagation time difference and the thickness of the wind turbine blade skin.

[0009] In one embodiment, the step of calculating the propagation time difference of the ultrasonic wave from transmission to reception based on the echo signal and the blade surface temperature includes: The propagation speed of sound on the blade skin at the current temperature is obtained by querying a preset sound speed temperature mapping table based on the blade surface temperature. Based on the propagation speed of sound and the preset blade skin thickness, calculate the theoretical echo time in the ice-free state; Based on the theoretical echo time and the preset time window length, a sliding window peak search is performed on the echo signal to obtain the arrival time of the target echo pulse. The target echo pulse is a pulse signal with the maximum amplitude that exceeds a preset threshold. Calculate the time offset between the arrival time and the theoretical echo time, and use the time offset as the propagation time difference of the ultrasonic wave from transmission to reception.

[0010] In one embodiment, the step of determining the corresponding de-icing mode based on the icing prediction probability, the current icing state, and at least two factors within the icing region includes: When the current icing state is an uniced state and the predicted probability of icing in a future preset time period is greater than a preset first probability threshold, the de-icing mode is determined to be an icing suppression mode. When the current icing state is an already iced state, the predicted probability of icing in a future preset time period is greater than a preset first probability threshold, and the proportion of the icing area is greater than a preset icing proportion threshold, the de-icing mode is determined to be the fast de-icing mode. When the current icing state is already icy, the predicted probability of icing in a future preset time period is greater than a preset second probability threshold, and the proportion of the icing area is less than or equal to a preset area proportion, the de-icing mode is determined to be the partition de-icing mode. When the current icing state is already icy and the predicted icing probability for a future preset time period is less than or equal to a preset second probability threshold, the de-icing mode is determined to be ultrasonic de-icing mode.

[0011] In one embodiment, the step of controlling the ultrasonic device and / or heating device to suppress icing on the blades based on the de-icing mode, the icing thickness, and at least two of the icing regions includes: When the de-icing mode is the icing suppression mode, the heating device is controlled to heat the preset area of ​​the blade with a first preset power. When the de-icing mode is the rapid de-icing mode, the corresponding de-icing frequency and heating power are determined according to the ice thickness; The ultrasonic device is controlled to de-ice the blades according to the de-icing frequency, and the heating device is controlled to heat the blades according to the heating power. When the de-icing mode is the zoned de-icing mode, the ultrasonic device is controlled to de-ic the blades at a preset frequency according to the icing area, and the heating device is controlled to heat the blades at a second preset power, wherein the second preset power is greater than the first preset power. When the de-icing mode is ultrasonic de-icing mode, the corresponding ultrasonic device is controlled to de-ic the blades at a preset frequency according to the icing area.

[0012] In one embodiment, the step of predicting icing based on the weather data and the environmental state data to obtain the icing prediction probability includes: The weather data and environmental state data from the past preset time period are preprocessed in chronological order to obtain multidimensional features, which include relative humidity, temperature, wind speed and liquid water content. The preset network model is trained based on the multidimensional features to obtain the trained icing probability prediction model. The weather data and environmental state data for a future preset time period are input into the icing probability prediction model to obtain the icing prediction probability for different future time periods.

[0013] In one embodiment, after the step of inputting the weather data and environmental state data for a future preset time period into the icing probability prediction model to obtain the icing prediction probability for different future time periods, the method further includes: The current icing state is used as the actual label, and a sample pair is formed with the icing prediction probability output by the model at the same time. Count a preset number of sample pairs and calculate the deviation between the predicted icing probability and the measured label. When the deviation value exceeds a preset error threshold, the calibration function is refitted according to a preset scaling method, and the icing prediction probability output by the icing probability prediction model is mapped according to the fitted calibration function to obtain the calibrated icing prediction probability.

[0014] Furthermore, to achieve the above objectives, this application also proposes a device for suppressing icing on wind turbine blades, the device comprising: The data acquisition module is used to acquire weather data and environmental status data of the area where the wind turbine is located, the current icing status, and the icing thickness at each sensor in the ultrasonic sensor array. An icing detection module is used to obtain the icing area by performing inverse distance weighting processing based on the icing thickness at each sensor in the ultrasonic sensor array. The icing prediction module is used to predict icing based on the weather data and the environmental state data, and to obtain the icing prediction probability. The de-icing mode confirmation module is used to determine the corresponding de-icing mode based on the icing prediction probability, the current icing state, and at least two of the icing areas. The de-icing modes include icing suppression mode, fast de-icing mode, zoned de-icing mode, and ultrasonic de-icing mode. An icing suppression module is used to suppress icing of the blades based on the de-icing mode, the icing thickness, and at least two controls of the ultrasonic device and / or heating device in the icing area.

[0015] In addition, to achieve the above objectives, this application also proposes a device for suppressing icing on wind turbine blades, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the method for suppressing icing on wind turbine blades as described above.

[0016] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the method for suppressing icing of wind turbine blades as described above.

[0017] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the method for suppressing icing on wind turbine blades as described above.

[0018] In addition, to achieve the above objectives, this application also proposes a system for suppressing icing on wind turbine blades, which includes an ultrasonic sensor array, an ultrasonic device, and a heating device installed inside the wind turbine blades.

[0019] This application provides a method for suppressing icing on wind turbine blades. The method includes: acquiring weather data and environmental state data of the area where the wind turbine is located, the current icing state, and the icing thickness at each sensor in the ultrasonic sensor array; performing inverse distance weighting processing on the icing thickness at each sensor in the ultrasonic sensor array to obtain an icing region; performing icing prediction based on the weather data and the environmental state data to obtain an icing prediction probability; determining a corresponding de-icing mode based on the icing prediction probability, the current icing state, and at least two of the icing regions, the de-icing modes including an icing suppression mode, a rapid de-icing mode, a zoned de-icing mode, and an ultrasonic de-icing mode; and controlling the ultrasonic device and / or heating device to suppress icing on the blades based on the de-icing mode, the icing thickness, and at least two of the icing regions. In summary, this application constructs an icing probability prediction model by integrating multi-source data such as weather, environment, and ultrasound, and adaptively matches the corresponding de-icing mode based on probability, state, and regional information. This completes the closed-loop control of the entire process from icing early warning and precise detection to zoned suppression and removal, solving the problems of slow response, high energy consumption, and incomplete coverage of traditional de-icing methods, and improving the adaptability and energy efficiency ratio of wind turbine blade de-icing. Attached Figure Description

[0020] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0021] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 A flowchart illustrating the first embodiment of the method for suppressing icing on wind turbine blades according to this application; Figure 2 A flowchart illustrating the second embodiment of the method for suppressing icing on wind turbine blades according to this application; Figure 3 A flowchart illustrating the third embodiment of the method for suppressing icing on wind turbine blades in this application; Figure 4 This is a schematic diagram of the module structure of the wind turbine blade icing suppression device according to an embodiment of this application; Figure 5 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the method for suppressing icing on wind turbine blades in the embodiments of this application.

[0023] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0024] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0025] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0026] The main solution of this application embodiment is as follows: Acquire weather data and environmental state data of the area where the wind turbine is located, the current icing state, and the icing thickness at each sensor in the ultrasonic sensor array; perform inverse distance weighting processing based on the icing thickness at each sensor in the ultrasonic sensor array to obtain the icing area; perform icing prediction based on the weather data and the environmental state data to obtain the icing prediction probability; determine the corresponding de-icing mode based on the icing prediction probability, the current icing state, and at least two of the icing areas, the de-icing mode including icing suppression mode, rapid de-icing mode, zoned de-icing mode, and ultrasonic de-icing mode; control the ultrasonic device and / or heating device to suppress icing on the blades based on the de-icing mode, the icing thickness, and at least two of the icing areas.

[0027] As an important form of clean energy, wind power faces the severe challenge of blade icing in cold or high-altitude regions. In low-temperature and high-humidity environments, icing easily forms on blade surfaces, leading to deterioration of aerodynamic performance, imbalance in mass distribution, and increased vibration. In severe cases, this can cause blade structural damage, reduced power generation efficiency, or even turbine shutdown, directly impacting the economics and safety of wind power generation. Currently, countermeasures against blade icing mainly include regular manual inspections, microwave de-icing, and temperature-triggered electric heating de-icing. These methods often overlook the coupled effects of multiple meteorological parameters, have limited ability to assess different blade icing conditions, and struggle to achieve precise, zoned, and graded de-icing control. Furthermore, existing methods primarily focus on post-icing, lacking early prediction and prevention capabilities for icing risks, and are unable to implement suppression measures before or after icing occurs. The overall system's adaptability and energy efficiency still have significant room for improvement. Therefore, improving the adaptability and energy efficiency of wind turbine blade de-icing is a pressing issue that needs to be addressed.

[0028] It should be noted that the implementing entity in this embodiment can be a wind turbine blade icing suppression system, a computing service device with data processing, network communication, and program execution functions, or an electronic device capable of achieving the aforementioned wind turbine blade icing suppression function, etc. This embodiment is not specifically limited in this regard. The following uses a wind turbine blade icing suppression system as an example to describe this embodiment and the following embodiments.

[0029] Based on this, embodiments of this application provide a method for suppressing icing on wind turbine blades, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the method for suppressing icing on wind turbine blades according to this application.

[0030] In this embodiment, the method for suppressing icing on wind turbine blades is applied to a system for suppressing icing on wind turbine blades. The system includes an ultrasonic sensor array, an ultrasonic device, and a heating device. The method for suppressing icing on wind turbine blades includes steps S10-S50: Step S10: Obtain weather data and environmental status data of the area where the wind turbine is located, the current icing status, and the icing thickness at each sensor in the ultrasonic sensor array.

[0031] It should be noted that weather data refers to weather forecast data for the geographical area where the wind turbine is located for a future period (e.g., 6-72 hours), obtained from meteorological service agencies or local numerical prediction models via internet APIs. This data includes future temperature, humidity, wind speed, and precipitation type. Environmental condition data refers to environmental physical parameters collected in real time by sensor arrays installed on the nacelle, hub, and blades of the wind turbine, including atmospheric temperature, blade surface temperature, relative humidity, wind speed, air pressure, precipitation amount and type, and liquid water content. Current icing status refers to the immediate icing condition of the blade surface as directly detected by the ultrasonic sensor array; it is a Boolean value (yes / no) determined by the characteristics of the echo signal. Icing thickness refers to the average thickness of the icing layer at each ultrasonic sensor installation location, calculated by analyzing the ultrasonic echo signal.

[0032] In one feasible implementation, the step of obtaining the current icing state and the icing thickness at each sensor in the ultrasonic sensor array specifically includes: Step S101: Acquire the echo signal and blade surface temperature.

[0033] It should be noted that in this step, the system acquires two key physical quantities: one is the echo signal collected by the paired ultrasonic sensors (transmitter Tx and receiver Rx), and the other is the blade surface temperature (Ts) measured or calculated by a temperature sensor. The echo signal carries all the information about the interaction between the ultrasonic waves and the blade skin and surface medium (air or ice layer), and is the direct basis for calculating the icing state and thickness; while the blade surface temperature is used to calibrate the material's sound velocity to improve the accuracy of subsequent time difference calculations.

[0034] Additionally, it should be noted that the echo signal refers to the timing voltage signal captured by a receiving sensor at the same location after the pulse emitted by the ultrasonic transmitting sensor penetrates the blade skin and is reflected at its outer surface (skin-medium interface). The waveform, amplitude, and arrival time of this signal directly reflect the physical properties of the interface medium (air or ice). The blade surface temperature (Ts) is measured by an array of thermocouples installed on the leeward side of the blade root, and the temperature at various points on the surface is calculated by combining the internal heat conduction model of the blade; or it is obtained by scanning the blade with a non-contact infrared thermal imager installed in the nacelle or hub.

[0035] Step S102: Calculate the propagation time difference of the ultrasonic wave from transmission to reception based on the echo signal and the blade surface temperature. The echo signal is collected by an ultrasonic sensor array installed inside the wind turbine blade.

[0036] It should be noted that in this step, the system calculates the actual round-trip time of the ultrasonic wave within the skin and surface medium, and compares it with the theoretical propagation time in the ice-free state. This yields the time increment (propagation time difference) caused by the presence of ice. The propagation time difference (Δt) refers to the difference between the measured round-trip time of the ultrasonic wave at the same sensor location and the theoretically calculated time in the ice-free state. When the surface is ice-free, this difference should be zero or within a very small noise range; when the surface is covered with ice, because the ultrasonic wave needs to propagate additionally within the ice layer, this difference will be positive and proportional to the ice thickness. This time difference is a direct input for calculating the ice thickness.

[0037] In one feasible implementation, step S102 specifically includes: Step A10: Based on the blade surface temperature, query the preset sound velocity temperature mapping table to obtain the propagation speed of sound on the blade skin at the current temperature.

[0038] It should be noted that since the ultrasonic propagation speed of the blade composite material changes with temperature, directly using a fixed sound velocity value would introduce measurement errors. In this step, the system first queries a pre-calibrated mapping table of sound velocity and temperature to obtain the sound propagation speed of the skin material corresponding to the current measured blade surface temperature (Ts). (Ts).

[0039] Step A20: Calculate the theoretical echo time in the ice-free state based on the propagation speed of sound and the preset blade skin thickness.

[0040] It should be noted that in this step, the system will adjust the speed of sound based on temperature compensation. (Ts) and the known skin thickness at the sensor location The theoretical time for an ultrasonic pulse to travel from emission, through the skin, reflect at the skin-air interface, and back through the skin to the receiving sensor in an ice-free state is calculated and converted into the theoretical echo arrival time from the zero moment of emission. The calculation formula is: (Formula 1) Step A30: Based on the theoretical echo time and the preset time window length, perform a sliding window peak search in the echo signal to obtain the arrival time of the target echo pulse. The target echo pulse is a pulse signal with the maximum amplitude that exceeds the preset threshold.

[0041] It should be noted that, due to noise in the actual echo signal and the potential distortion of the echo waveform caused by ice, a robust method is needed to accurately identify the true interface reflection echo. Specifically, the system will use the theoretical echo time... Centered on a preset time window, a range is extended forward and backward. Within this window, the amplitude of the acquired echo signal is scanned to find the pulse peak point with the largest amplitude exceeding a preset threshold. The time corresponding to this peak point is determined as the arrival time of the actual target echo. The preset threshold is used to filter out stray reflections caused by circuit noise or minor defects inside the material.

[0042] Step A40: Calculate the time offset between the arrival time and the theoretical echo time, and use the time offset as the propagation time difference of the ultrasonic wave from transmission to reception.

[0043] It should be noted that in this step, the system will identify the actual arrival time. Compared with the calculated theoretical ice-free time By subtracting the values, we obtain the time offset Δt = - This Δt is the ultrasonic wave propagation time difference that the system needs to calculate. If Δt < 0 (the actual echo time is earlier than the theoretical time, possibly due to noise interference), it is considered an invalid value, and the system will use the average propagation time difference of the first 3 cycles instead.

[0044] Step S103: Determine the current icing state and the icing thickness at each sensor in the ultrasonic sensor array based on the propagation time difference and the thickness of the wind turbine blade skin.

[0045] It should be noted that the thickness of the wind turbine blade skin refers to the actual thickness of the blade skin at the location where the ultrasonic sensor is installed. This thickness is calibrated using an ultrasonic thickness gauge during blade manufacturing and stored in the system parameter database as a fixed value (denoted as...). In this step, the system will call the preset icing state determination threshold (such as...). The value is 0.2 μs, determined experimentally, corresponding to a minimum detectable ice thickness of 0.1 mm. The calculated propagation time difference is used to... Compare with this threshold. When ≤ When the current sensor corresponds to a blade position, it is determined that the blade is in an un-iced state, and the icing thickness is recorded as 0 mm; when > When the current sensor detects a blade location that is already icy, the ice thickness at that location is calculated using the ice thickness calculation formula, based on the ultrasonic wave propagation velocity in the calibrated ice. The detection results from all sensors are then combined to form icing state distribution matrices and ice thickness distribution matrices along the blade spanwise and chordal directions.

[0046] Step S20: Obtain the icing region by performing inverse distance weighting processing based on the icing thickness at each sensor in the ultrasonic sensor array.

[0047] It should be noted that in this step, the system reconstructs the thickness measurements at discrete points (each sensor location) into a continuous, visualized icing region distribution map using an inverse distance-weighted spatial interpolation method, thereby quantifying the area and spatial characteristics of the icing. The icing region is the continuous spatial range of the blade surface covered by ice, determined based on the interpolation results. The system uses a set thickness threshold to determine whether a grid point belongs to the icing region, and then calculates the size of the icing region, i.e., the icing area and its proportion of the total blade area.

[0048] Step S30: Based on the weather data and the environmental state data, perform icing prediction to obtain the icing prediction probability.

[0049] It's important to note that in this step, the system utilizes a data-driven model to integrate historical and real-time meteorological and environmental information to quantitatively assess the probability of icing events occurring within a specific future timeframe, thus providing a proactive early warning. Specifically, the icing prediction probability is a value between 0 and 1, representing the likelihood of leaf icing at a future time (e.g., one hour). This probability is output by a pre-trained machine learning model (such as a gradient boosting decision tree or an LSTM network). The model learns the complex nonlinear relationship between meteorological and environmental characteristics and icing events using historical data. To maintain prediction accuracy, the system also incorporates an online calibration mechanism to correct the model's output probability using recent ultrasonic detection results.

[0050] Step S40: Determine the corresponding de-icing mode based on the icing prediction probability, the current icing state, and at least two of the icing regions. The de-icing mode includes icing suppression mode, rapid de-icing mode, zoned de-icing mode, and ultrasonic de-icing mode.

[0051] It should be noted that in this step, the system adaptively selects the most economical and effective icing suppression mode based on three dimensions of information: future risk (i.e., icing prediction probability), current facts (i.e., current icing state), and severity (i.e., icing area). This is achieved through a pre-defined multi-mode matching logic. The icing suppression modes include: Icing Suppression Mode: Preventative low-power heating before icing occurs; Rapid De-icing Mode: For severe and persistent icing, powerful removal using a combination of ultrasonic waves and high-power heating; Zoned De-icing Mode: For localized icing when the weather is easing, de-icing is only performed on the iced area; Ultrasonic De-icing Mode: When there is no new icing risk, de-icing is performed using only ultrasonic waves. The triggering of each mode is determined by at least two of the following: icing prediction probability, current icing state, and the proportion of icing area.

[0052] In one feasible implementation, step S40 specifically includes: Step S401: When the current icing state is an uniced state and the predicted probability of icing in the future preset time period is greater than a preset first probability threshold, the de-icing mode is determined to be the icing suppression mode.

[0053] It should be noted that in this step, the system checks the current icing state S_ice obtained from ultrasonic detection. If it is not icing, it indicates that the blade surface is currently safe. It then further detects the future risk output by the prediction model, namely the predicted probability of icing P_ice (t=1) over a preset time period (e.g., 1 hour). When this probability exceeds a high preset first probability threshold θ1 (e.g., θ1=0.7), it indicates that although there is currently no ice, icing is highly likely to occur in the near future. At this point, the system enters icing suppression mode. The core logic of this mode is to proactively intervene before icing occurs by raising the blade surface temperature through low-intensity heating, thereby changing the environment around the blade and preventing the freezing process of supercooled water droplets.

[0054] Step S402: When the current icing state is an icing state, the predicted probability of icing in the future preset time period is greater than a preset first probability threshold, and the proportion of the icing area is greater than a preset icing proportion threshold, the de-icing mode is determined to be the fast de-icing mode.

[0055] It should be noted that when the current icing state S_ice is "iced," it indicates that a problem has already occurred. The system assesses the severity in two dimensions: first, the persistence of future risks (P_ice(t=1)>θ1), indicating that severe weather conditions are continuing and the ice layer may continue to thicken; second, the spatial severity of the current ice condition, i.e., whether the icing area ratio R_ice (the ratio of iced area to the total blade area) exceeds a preset icing ratio threshold θR (e.g., θR=0.3). When all three conditions are met simultaneously, the system enters rapid de-icing mode. The goal of this mode is to remove the ice as quickly as possible using high-frequency ultrasound and high-power heating to restore aerodynamic performance and prevent accidents.

[0056] Step S403: When the current icing state is an icing state, the predicted probability of icing in the future preset time period is greater than a preset second probability threshold, and the proportion of the icing area is less than or equal to a preset area proportion, the de-icing mode is determined to be the partition de-icing mode.

[0057] It should be noted that when the current state is icy, but the future risk is at a moderate level (P_ice(t=1)>θ2, where the preset second probability threshold θ2 is 0.3), and the proportion of iced area R_ice does not exceed the severe threshold (R_ice ≤θR), this indicates that there is localized icing, but weather conditions are easing, and the risk of future icing is reduced. In this situation, the system will enter a zoned de-icing mode. The strategy of this mode is to apply moderate-intensity ultrasonic waves and heating only to the specific iced areas detected, avoiding unnecessary energy consumption in ice-free areas, and achieving an optimal balance between efficiency and energy consumption.

[0058] Step S404: When the current icing state is already iced and the predicted icing probability for a future preset time period is less than or equal to a preset second probability threshold, determine the de-icing mode as ultrasonic de-icing mode.

[0059] It should be noted that the current state is that the blades are already icy, but the future risk is extremely low (P_ice(t=1) ≤ θ2). This means that although there is ice on the blades, the weather forecast indicates favorable future conditions, with virtually no risk of new supercooled water impacting the blades. In this situation, the system will enter the most energy-efficient ultrasonic de-icing mode. This mode does not activate the heating device; it only uses ultrasonic transducers to generate mechanical vibrations in the icy area, causing the ice to break and fall off due to fatigue, thus maximizing energy savings while ensuring there is no risk of new ice formation.

[0060] Step S50: Based on the de-icing mode, the ice thickness, and at least two control devices in the ice region, the ultrasonic device and / or heating device are used to suppress icing on the blades.

[0061] It should be noted that in this step, the system dynamically adjusts parameters such as the operating frequency of the ultrasonic device, the power of the heating device (e.g., through PWM duty cycle adjustment), and the area of ​​action, according to the specific requirements of the selected de-icing mode. Understandably, this control strategy aims to achieve efficient and energy-saving de-icing or ice-suppressing effects.

[0062] In one feasible implementation, step S50 specifically includes: Step S501: When the de-icing mode is the icing suppression mode, control the heating device to heat the preset area of ​​the blade with a first preset power.

[0063] It should be noted that this step is performed when the de-icing mode is the rapid de-icing mode (i.e., the system predicts a high risk of icing but there is currently no icing). Its purpose is to prevent the freezing and adhesion of supercooled water droplets by pre-applying a low, continuous heating power to critical aerodynamic areas of the blade (such as the leading edge). The first preset power is a relatively low power density setting, for example, 200-400 W / m². This power is sufficient to maintain the blade surface temperature above zero degrees Celsius, but far below the power required for melting ice, achieving minimal preventative performance consumption. The preset area refers to the area on the blade most prone to icing and with the greatest impact on aerodynamic performance, such as a specific section of the blade's leading edge.

[0064] Step S502: When the de-icing mode is the fast de-icing mode, determine the corresponding de-icing frequency and heating power according to the ice thickness.

[0065] It should be noted that when the de-icing mode is the rapid de-icing mode (i.e., when facing a severe and continuous risk of icing), the ice layer needs to be removed quickly. The system dynamically calculates the ultrasonic vibration frequency and heating melting power required to remove the ice layer of that thickness based on real-time monitoring of the ice thickness, achieving a match between the force and the demand. Specifically, the de-icing frequency and heating power are positively correlated with the ice thickness, as shown in Formula 2: (Formula 2) in, For ice thickness, Here, k is the preset base de-icing frequency, and k is the preset coefficient relating the de-icing frequency to the ice thickness. The preset maximum heating power, The preset icing thickness threshold (e.g., 10mm).

[0066] Step S503: Control the ultrasonic device to de-ice the blades according to the de-icing frequency, and control the heating device to heat the blades according to the heating power.

[0067] It should be noted that in this step, the system drives an ultrasonic device (such as an ultrasonic transducer) to generate high-intensity vibrations based on the calculated de-icing frequency, breaking the adhesion between the ice layer and the skin. Simultaneously, based on the calculated heating power, a heating device (such as carbon fiber heating cloth / foil or alloy resistance wire / foil integrated inside the wind turbine blades with zoned control) is activated to quickly melt any remaining ice at the interface and evaporate moisture. Both processes cycle according to a pre-set coordinated sequence (e.g., ultrasonic vibration followed by heating) until the ice layer is removed.

[0068] Step S504: When the de-icing mode is the zoned de-icing mode, the ultrasonic device is controlled to de-ic the blades at a preset frequency according to the icing area, and the heating device is controlled to heat the blades at a second preset power, wherein the second preset power is greater than the first preset power.

[0069] It should be noted that this step is performed when the de-icing mode is the zoned de-icing mode (i.e., when the icing is localized). Specifically, the system establishes a mapping relationship between the actuator (ultrasonic transducer and heating zone) positions and the positions of the icing areas on the blade surface, and generates an actuator sequence that only excites the actuators covering the icing areas based on the distribution map of the icing areas. Then, the system activates only the ultrasonic transducers and heating zones covering the identified icing areas according to the actuator sequence to perform de-icing and heating respectively, avoiding unnecessary energy consumption. Additionally, it should be noted that the second preset power is a medium power density setting value, for example, approximately 600 W / m², higher than the maximum power of the icing suppression mode but lower than the maximum power of the rapid de-icing mode, suitable for localized de-icing.

[0070] Step S505: When the de-icing mode is ultrasonic de-icing mode, control the corresponding ultrasonic device to de-ic the blades at a preset frequency according to the icing area.

[0071] It should be noted that this step is performed when the de-icing mode is ultrasonic de-icing mode (i.e., icing has occurred but the risk of future icing is low). The system will only activate the ultrasonic transducers covering the identified icing area to generate a mechanical vibration effect at a preset frequency, causing the ice layer to fatigue, break, and detach, without activating any heating devices to reduce unnecessary energy consumption. The preset frequency is a preset frequency that can cause the ice layer to resonate or generate effective shear stress (e.g., 20kHz-100kJz, depending on the type and material of the wind turbine blades).

[0072] This embodiment provides a method for suppressing icing on wind turbine blades. The method includes: acquiring weather data and environmental state data of the area where the wind turbine is located, the current icing state, and the icing thickness at each sensor in the ultrasonic sensor array; performing inverse distance weighting processing on the icing thickness at each sensor in the ultrasonic sensor array to obtain an icing region; performing icing prediction based on the weather data and the environmental state data to obtain an icing prediction probability; determining a corresponding de-icing mode based on the icing prediction probability, the current icing state, and at least two of the icing regions, the de-icing modes including an icing suppression mode, a rapid de-icing mode, a zoned de-icing mode, and an ultrasonic de-icing mode; and controlling the ultrasonic device and / or heating device to suppress icing on the blades based on the de-icing mode, the icing thickness, and at least two of the icing regions. In summary, this embodiment constructs an icing probability prediction model by integrating multi-source data such as weather, environment, and ultrasound, and adaptively matches the corresponding de-icing mode based on probability, state, and regional information. It completes the closed-loop control of the entire process from icing early warning and precise detection to zoned suppression and removal, solving the problems of slow response, high energy consumption, and incomplete coverage of traditional de-icing methods, and improving the adaptability and energy efficiency ratio of wind turbine blade de-icing.

[0073] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in Embodiment 1 above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 2 , Figure 2 This is a flowchart illustrating the second embodiment of the method for suppressing icing on wind turbine blades according to this application. Step S20 specifically includes: Step S201: Obtain the position coordinates of each sensor in the ultrasonic sensor array.

[0074] It should be noted that position coordinates refer to the three-dimensional coordinates of each ultrasonic sensor (usually referring to its acoustic center) within a certain agreed-upon blade coordinate system (e.g., a coordinate system with the center of the blade root as the origin, the spanwise direction as the X-axis, the chordal direction as the Y-axis, and the thickness direction as the Z-axis). , , These coordinates represent the specific location of the sensor within the blade. In this step, the system reads the installation position coordinates of each ultrasonic sensor in the three-dimensional space of the blade from the configuration information database. These coordinates are determined based on the blade design drawings and actual sensor installation records, and are entered during system initialization.

[0075] Step S202: Generate grid points on the surface of the wind turbine blades according to the preset spatial resolution, and calculate the spatial distance between each grid point and the position coordinates of each sensor.

[0076] It should be noted that in this step, the system discretizes the continuous blade surface and assigns a value to each discrete point (grid point, ( , , )) and known measurement points (sensor locations, ( , , Establish spatial relationships. Specifically, the system generates a series of grid points to be valued on the 3D surface model of the blade according to a preset grid resolution (e.g., one point every 0.5 meters along the spanwise direction and one point every 0.1 meters along the chordwise direction). Then, for each grid point, the spatial distance between it and all sensor locations is calculated. As shown in Formula 3: (Formula 3) Additionally, it should be noted that the preset spatial resolution is a parameter pre-set based on the blade size, sensor density, and accuracy requirements for identifying icing areas. Higher resolution generates more grid points and reconstructs richer details, but also increases computational complexity.

[0077] Step S203: Perform weighted interpolation processing on the ice thickness at each sensor in the ultrasonic sensor array based on the spatial distance to obtain the estimated ice thickness at the grid points.

[0078] It should be noted that weighted interpolation is an estimation method based on spatial proximity. The estimated ice thickness at each grid point is a weighted average of the thickness values ​​measured by all sensors, with the weights being a function of the distance from the sensor to the grid point (i.e., the closer a sensor is to a grid point, the greater its measurement influences the estimation result for that grid point). Understandably, the purpose of this step is to use the known ice thickness at sensor points, and through a weighted averaging method that considers distance factors, to calculate the estimated ice thickness at each unknown grid point.

[0079] In one feasible implementation, step S203 specifically includes: Step B10: Obtain the measurement confidence parameters of each sensor in the ultrasonic sensor array.

[0080] It should be noted that, to improve the reliability of the interpolation results, this step introduces the confidence difference of the sensor measurements. The system dynamically assigns a measurement confidence parameter to each sensor. This parameter is calculated based on pre-determined signal quality (such as echo signal-to-noise ratio), the sensor's recent self-test status, or the consistency of historical data. Sensor data with high confidence will have a higher weight in the interpolation.

[0081] Step B20: For each grid point, calculate the Euclidean distance between the grid point and the sensor, and calculate the initial weighting factor based on the Euclidean distance.

[0082] It should be noted that each grid point refers to a node corresponding to a group of sensors within a preset area (e.g., arranging a group of sensors at 10% spans along the blade's span, with each group consisting of 4*4 nodes). In this step, the system calculates initial weights based on spatial distance. For grid point g corresponding to sensor i in a group, its Euclidean distance is calculated. Then, the initial weighting factors are calculated according to the inverse distance weighting formula. Specifically, as shown in Formula 4: (Formula 4) Where p is the distance attenuation coefficient (usually taken as 2). The closer the distance, the greater the attenuation. The larger the value.

[0083] Step B30: Correct the initial weighting factor according to the measurement confidence parameter to obtain the target weighting factor.

[0084] It should be noted that in this step, the system will measure the confidence parameters. With initial weighting factor Combined, calculate the weighting factor ( Specifically, the system will use a weighted fusion method, with the following formula: (Formula 5) Understandably, this correction process enables high-confidence sensors ( The initial weights of sensors with a confidence level close to 1 remain basically unchanged, while those with a low confidence level ( The weights (close to 0) are significantly weakened. Secondly, for a set of grid points g, their weights are compared with the weight factors of a set of sensors in the corresponding region. Normalization is performed to obtain the target weight factor. This involves normalizing the sum of the weight factors of the nearest set of sensors within the area where the grid node is located, so that the sum of the target weight factors of all sensors for that grid point is 1, thus ensuring the rationality of the weighted summation.

[0085] Step B40: The ice thickness values ​​at each sensor are weighted and summed according to the target weighting factor to obtain the initial estimated ice thickness at the grid points.

[0086] It should be noted that in this step, for grid point g, the system will obtain the ice thickness measured by the nearest group (e.g., N) of sensors. and the corresponding target weight factor Specifically, the initial estimated ice thickness at grid point g ( Calculated using the weighted average formula: (Formula 6) Step B50: Perform spatial smoothing on the initial estimated icing thickness of adjacent grid points to obtain the estimated icing thickness.

[0087] It should be noted that inverse distance weighted interpolation may produce local abrupt changes or unnatural bullseye effects in sparse sensor regions or when the weight distribution is uneven. Therefore, in this step, the system performs post-processing on the initial estimation results using spatial smoothing filtering. For example, mean filtering or Gaussian filtering is used to calculate the estimated icing thickness at each grid point using the initial estimates of each grid point and its neighboring grid points. This eliminates unreasonable, drastic local variations. Understandably, spatial smoothing refers to suppressing random noise and outliers through neighborhood averaging. For example, a 4x4 mean filter replaces the value of a grid point with the average of its own value and the values ​​of its 16 neighboring points.

[0088] Step S204: Determine the icing area based on the estimated icing thickness.

[0089] It should be noted that after obtaining the estimated icing thickness of all grid points on the blade surface, the system will distinguish whether a grid point belongs to an icing area based on a preset thickness judgment threshold (e.g., 0.1 mm). A continuous area formed by all grid points with an estimated thickness greater than this threshold is determined to be an icing area. The system can also further calculate the percentage of the total area of ​​the icing area to the blade surface area.

[0090] In this embodiment, by obtaining the ice thickness from discrete sensor points and performing spatial interpolation and smoothing processing based on their position coordinates and confidence levels, a continuous and accurate spatial distribution reconstruction of the ice area on the blade surface is achieved. This solves the problem that traditional point-based measurements cannot fully reflect the ice range and morphology, and improves the spatial integrity of ice area identification and the accuracy of subsequent zonal de-icing decisions.

[0091] Based on the first and second embodiments of this application, in the third embodiment of this application, the content that is the same as or similar to that in embodiments one and two above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 3 , Figure 3 This is a flowchart illustrating the third embodiment of the method for suppressing icing on wind turbine blades according to this application. Step S30 specifically includes: Step S301: Preprocess the weather data and environmental state data from the past preset time period in chronological order to obtain multidimensional features, including relative humidity, temperature, wind speed and liquid water content.

[0092] It should be noted that in this step, the system extracts weather data (such as NWP forecast data) and environmental condition data (such as measured data) from a continuous period of time (e.g., the past 24 hours) from the historical database and aligns them according to timestamps. Subsequently, these raw time-series data undergo preprocessing, including missing value imputation, data standardization / normalization, and the construction of a sliding window feature for time series forecasting. The processed data is organized into a multi-dimensional feature vector, which includes key parameters strongly correlated with the icing physical process, such as relative humidity (RH), temperature (Ta), wind speed (WS), and liquid water content (LWC), as well as other features such as wind direction, air pressure, and precipitation type.

[0093] Step S302: Train the preset network model based on the multidimensional features to obtain the trained icing probability prediction model.

[0094] It should be noted that in this step, the system uses historical feature vectors corresponding to the multidimensional features as input (feature X), and correspondingly uses the icing status recorded by ultrasonic detection and manual confirmation within the same historical period (yes = 1, no = 0) as labels (Y). A suitable machine learning model architecture (such as Gradient Boosting Decision Tree (XGBoost), LightGBM, or Temporal Recurrent Neural Network (LSTM)) is selected as the preset network model. By inputting historical features and labels into the model for training, the model automatically learns the complex nonlinear mapping relationship between multidimensional meteorological environmental features and the probability of icing events. The trained model is the icing probability prediction model.

[0095] Step S303: Input the weather data and environmental state data for a future preset time period into the icing probability prediction model to obtain the icing prediction probability for different future time periods.

[0096] It should be noted that during online system operation, the latest weather forecast data (NWP) for a future time period (e.g., the next 6 hours) and current environmental status data are preprocessed using the same process as in the training phase to form feature vectors representing the future. These feature vectors are then input into the icing probability prediction model trained in step S302. The model outputs a corresponding icing prediction probability P_ice(t) for each future target time period (e.g., the 1st hour, the 2nd hour, etc.), thus forming a quantitative prediction sequence for future icing risk.

[0097] In one feasible implementation, after step S303, the method further includes: Step S304: Take the current icing state as the actual label and form a sample pair with the icing prediction probability output by the model at the same time.

[0098] It should be noted that, in order to maintain the long-term accuracy of the prediction model, the system establishes a closed loop for continuous learning. Whenever the system obtains the current icing state S_ice of the blade through ultrasonic detection (as the real situation, i.e., the actual label), it pairs this label with the icing prediction probability P_ice made by the model for the future period at the corresponding time, forming a sample pair (P_ice, S_ice) used to evaluate and calibrate the model performance.

[0099] Understandably, a sample pair is a data unit used for model calibration, containing the model's predicted output (probability value) for a given time and the actual observation (binary label) at that time. Accumulating a large number of such sample pairs can reflect the model's prediction bias.

[0100] Step S305: Count a preset number of sample pairs and calculate the deviation between the predicted icing probability and the measured label.

[0101] It should be noted that in this step, the system continuously collects the generated sample pairs. Once a certain number has been accumulated (e.g., the past 72 hours or 100 samples), statistical analysis is performed on these samples. The calibration degree of the current model prediction is quantified by calculating the overall deviation between the predicted probability and the actual binary label (e.g., using the Brier score, the area under the calibration curve, or the difference in the distribution of the predicted probability on positive and negative samples as the deviation value). In other words, it is determined whether the probability value output by the model truly reflects the actual frequency of icing.

[0102] Step S306: When the deviation value exceeds the preset error threshold, the calibration function is refitted according to the preset scaling method, and the icing prediction probability output by the icing probability prediction model is mapped according to the fitted calibration function to obtain the calibrated icing prediction probability.

[0103] It should be noted that when the evaluation finds an excessively large deviation, indicating probability drift in the model, the system will trigger a calibration process. A pre-defined probability calibration scaling method (such as Platt Scaling or Isotonic Regression) is employed. This method uses recently collected sample pairs to refit a simple calibration function, which adjusts the model's original predicted probabilities. Mapped to a frequency closer to the actual observation frequency For Platt Scaling, the calibration function is a logistic regression function: (Formula 7) in, It is a logistic function, specifically the sigmoid function. A and B are parameters fitted based on the new samples. Subsequently, the system transforms the original predicted probabilities of the model using this function, outputting calibrated icing prediction probabilities. This allows for rapid correction of prediction bias without retraining the complex main model, maintaining the long-term reliability of the system's predictions.

[0104] In this embodiment, a machine learning model is trained using historical meteorological and environmental data to predict the probability of icing. An online calibration mechanism based on measured results is introduced to achieve a dynamic and accurate quantitative assessment of future icing risks. This solves the problems of inaccurate early warning using traditional threshold methods and performance drift of models over long-term applications, and improves the reliability, foresight, and adaptability of icing early warning systems.

[0105] This application also provides a device for suppressing icing on wind turbine blades; please refer to [reference needed]. Figure 4 The device for suppressing icing on the wind turbine blades includes: Data acquisition module 10 is used to acquire weather data and environmental status data of the area where the wind turbine is located, the current icing status, and the icing thickness at each sensor in the ultrasonic sensor array. The icing detection module 20 is used to obtain the icing area by performing inverse distance weighting processing based on the icing thickness at each sensor in the ultrasonic sensor array. The icing prediction module 30 is used to predict icing based on the weather data and the environmental state data, and obtain the icing prediction probability. The de-icing mode confirmation module 40 is used to determine the corresponding de-icing mode based on the icing prediction probability, the current icing state, and at least two of the icing areas. The de-icing mode includes icing suppression mode, fast de-icing mode, zoned de-icing mode, and ultrasonic de-icing mode. The icing suppression module 50 is used to suppress icing of the blades based on the de-icing mode, the icing thickness, and at least two of the ultrasonic devices and / or heating devices in the icing area.

[0106] The wind turbine blade icing suppression device provided in this application, employing the wind turbine blade icing suppression method described in the above embodiments, can solve the technical problem of how to improve the adaptability and energy efficiency ratio of wind turbine blade de-icing. Compared with the prior art, the beneficial effects of the wind turbine blade icing suppression device provided in this application are the same as those of the wind turbine blade icing suppression method provided in the above embodiments, and other technical features in the wind turbine blade icing suppression device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0107] This application provides a device for suppressing icing on wind turbine blades. The device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the method for suppressing icing on wind turbine blades as described in Embodiment 1 above.

[0108] The following is for reference. Figure 5This document illustrates a structural schematic diagram of a device suitable for suppressing icing on wind turbine blades in implementing embodiments of this application. The device for suppressing icing on wind turbine blades in embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 5 The wind turbine blade icing suppression device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0109] like Figure 5 As shown, the wind turbine blade icing suppression device may include a processing unit 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in ROM (Read Only Memory) 1002 or a program loaded from storage device 1003 into RAM (Random Access Memory) 1004. RAM 1004 also stores various programs and data required for the operation of the wind turbine blade icing suppression device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via bus 1005. Input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, LCDs (Liquid Crystal Displays), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the wind turbine blade icing suppression device to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows wind turbine blade icing suppression devices with various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0110] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0111] The wind turbine blade icing suppression device provided in this application, employing the wind turbine blade icing suppression method described in the above embodiments, can solve the technical problem of how to improve the adaptability and energy efficiency ratio of wind turbine blade de-icing. Compared with the prior art, the beneficial effects of the wind turbine blade icing suppression device provided in this application are the same as those of the wind turbine blade icing suppression method provided in the above embodiments, and other technical features of this wind turbine blade icing suppression device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0112] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0113] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0114] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the method for suppressing icing on wind turbine blades in the above embodiments.

[0115] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, RAM (Random Access Memory), ROM (Read Only Memory), EPROM (Erasable Programmable Read Only Memory or Flash Memory), optical fibers, CD-ROM (CD-Read Only Memory), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0116] The aforementioned computer-readable storage medium may be included in a device for suppressing icing of wind turbine blades; or it may exist independently and not be assembled into a device for suppressing icing of wind turbine blades.

[0117] The aforementioned computer-readable storage medium carries one or more programs that, when executed by a wind turbine blade icing suppression device, cause the wind turbine blade icing suppression device to: acquire weather data and environmental state data of the area where the wind turbine is located, the current icing state, and the icing thickness at each sensor in the ultrasonic sensor array; perform inverse distance weighting processing on the icing thickness at each sensor in the ultrasonic sensor array to obtain an icing area; perform icing prediction based on the weather data and the environmental state data to obtain an icing prediction probability; determine a corresponding de-icing mode based on the icing prediction probability, the current icing state, and at least two of the icing areas, the de-icing mode including an icing suppression mode, a rapid de-icing mode, a zoned de-icing mode, and an ultrasonic de-icing mode; and control the ultrasonic device and / or heating device to suppress icing on the blades based on the de-icing mode, the icing thickness, and at least two of the icing areas.

[0118] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including LAN (Local Area Network) or WAN (Wide Area Network)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0119] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0120] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0121] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described method for suppressing icing on wind turbine blades. This method can solve the technical problem of how to improve the adaptability and energy efficiency ratio of wind turbine blade de-icing. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the wind turbine blade icing suppression method provided in the above embodiments, and will not be repeated here.

[0122] This application also provides a system for suppressing icing on wind turbine blades, characterized in that the system includes an ultrasonic sensor array, an ultrasonic device, and a heating device installed inside the wind turbine blade. The wind turbine blade icing suppression system provided in this application, employing the icing suppression method described in the above embodiments, can solve the technical problem of how to improve the adaptability and energy efficiency ratio of wind turbine blade de-icing. Compared with the prior art, the beneficial effects of the wind turbine blade icing suppression system provided in this application are the same as those of the icing suppression method provided in the above embodiments, and other technical features of this system are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0123] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method for suppressing icing on wind turbine blades as described above.

[0124] The computer program product provided in this application can solve the technical problem of how to improve the adaptability and energy efficiency ratio of wind turbine blade de-icing. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as the beneficial effects of the wind turbine blade icing suppression method provided in the above embodiments, and will not be repeated here.

[0125] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for suppressing icing on wind turbine blades, characterized in that, The method for suppressing icing on wind turbine blades is applied to a system for suppressing icing on wind turbine blades. The system includes an ultrasonic sensor array, an ultrasonic device, and a heating device. The method includes: Acquire weather and environmental data of the area where the wind turbine is located, the current icing status, and the icing thickness at each sensor in the ultrasonic sensor array; The icing region is obtained by inverse distance weighting processing based on the icing thickness at each sensor in the ultrasonic sensor array. Based on the weather data and the environmental condition data, icing prediction is performed to obtain the icing prediction probability; Based on the icing prediction probability, the current icing state, and at least two of the icing regions, a corresponding de-icing mode is determined. The de-icing mode includes an icing suppression mode, a rapid de-icing mode, a zoned de-icing mode, and an ultrasonic de-icing mode. The blades are icing suppressed according to the de-icing mode, the icing thickness, and at least two of the ultrasonic devices and / or heating devices in the icing area.

2. The method as described in claim 1, characterized in that, The step of obtaining the icing region by performing inverse distance-weighted processing based on the icing thickness at each sensor in the ultrasonic sensor array includes: Obtain the position coordinates of each sensor in the ultrasonic sensor array; Grid points are generated on the surface of the wind turbine blades according to a preset spatial resolution, and the spatial distance between each grid point and the position coordinates of each sensor is calculated. The ice thickness at each sensor in the ultrasonic sensor array is weighted and interpolated based on the spatial distance to obtain the estimated ice thickness at the grid points. The icing area is determined based on the estimated icing thickness.

3. The method as described in claim 2, characterized in that, The step of performing weighted interpolation processing on the ice thickness at each sensor in the ultrasonic sensor array based on the spatial distance to obtain the estimated ice thickness at the grid points includes: Obtain the measurement confidence parameters of each sensor in the ultrasonic sensor array; For each grid point, calculate the Euclidean distance between the grid point and the sensor, and calculate the initial weighting factor based on the Euclidean distance; The initial weighting factor is corrected based on the measured confidence parameter to obtain the target weighting factor; The ice thickness values ​​at each sensor are weighted and summed according to the target weighting factor to obtain the initial estimated ice thickness at the grid points. Spatially smooth the initial estimated icing thickness at adjacent grid points to obtain the estimated icing thickness.

4. The method as described in claim 1, characterized in that, The steps for obtaining the current icing state and the icing thickness at each sensor in the ultrasonic sensor array include: Acquire echo signals and blade surface temperatures; The propagation time difference of the ultrasonic wave from transmission to reception is calculated based on the echo signal and the blade surface temperature. The echo signal is collected by an ultrasonic sensor array installed inside the wind turbine blade. The current icing state and the icing thickness at each sensor in the ultrasonic sensor array are determined based on the propagation time difference and the thickness of the wind turbine blade skin.

5. The method as described in claim 4, characterized in that, The step of calculating the propagation time difference of the ultrasonic wave from transmission to reception based on the echo signal and the blade surface temperature includes: The propagation speed of sound on the blade skin at the current temperature is obtained by querying a preset sound speed temperature mapping table based on the blade surface temperature. Based on the propagation speed of sound and the preset blade skin thickness, calculate the theoretical echo time in the ice-free state; Based on the theoretical echo time and the preset time window length, a sliding window peak search is performed on the echo signal to obtain the arrival time of the target echo pulse. The target echo pulse is a pulse signal with the maximum amplitude that exceeds a preset threshold. Calculate the time offset between the arrival time and the theoretical echo time, and use the time offset as the propagation time difference of the ultrasonic wave from transmission to reception.

6. The method as described in claim 1, characterized in that, The step of determining the corresponding de-icing mode based on the icing prediction probability, the current icing state, and at least two factors within the icing region includes: When the current icing state is an uniced state and the predicted probability of icing in a future preset time period is greater than a preset first probability threshold, the de-icing mode is determined to be an icing suppression mode. When the current icing state is an already iced state, the predicted probability of icing in a future preset time period is greater than a preset first probability threshold, and the proportion of the icing area is greater than a preset icing proportion threshold, the de-icing mode is determined to be the fast de-icing mode. When the current icing state is already icy, the predicted probability of icing in a future preset time period is greater than a preset second probability threshold, and the proportion of the icing area is less than or equal to a preset area proportion, the de-icing mode is determined to be the partition de-icing mode. When the current icing state is already icy and the predicted icing probability for a future preset time period is less than or equal to a preset second probability threshold, the de-icing mode is determined to be ultrasonic de-icing mode.

7. The method as described in claim 1, characterized in that, The step of controlling the ultrasonic device and / or heating device to suppress icing on the blades according to the de-icing mode, the icing thickness, and the icing region includes: When the de-icing mode is the icing suppression mode, the heating device is controlled to heat the preset area of ​​the blade with a first preset power. When the de-icing mode is the rapid de-icing mode, the corresponding de-icing frequency and heating power are determined according to the ice thickness; The ultrasonic device is controlled to de-ice the blades according to the de-icing frequency, and the heating device is controlled to heat the blades according to the heating power. When the de-icing mode is the zoned de-icing mode, the ultrasonic device is controlled to de-ic the blades at a preset frequency according to the icing area, and the heating device is controlled to heat the blades at a second preset power, wherein the second preset power is greater than the first preset power. When the de-icing mode is ultrasonic de-icing mode, the corresponding ultrasonic device is controlled to de-ic the blades at a preset frequency according to the icing area.

8. The method as described in claim 1, characterized in that, The step of predicting icing based on the weather data and the environmental state data to obtain the icing prediction probability includes: The weather data and environmental state data from the past preset time period are preprocessed in chronological order to obtain multidimensional features, which include relative humidity, temperature, wind speed and liquid water content. The preset network model is trained based on the multidimensional features to obtain the trained icing probability prediction model. The weather data and environmental state data for a future preset time period are input into the icing probability prediction model to obtain the icing prediction probability for different future time periods.

9. The method as described in claim 8, characterized in that, After the step of inputting the weather data and environmental state data for a future preset time period into the icing probability prediction model to obtain the icing prediction probability for different future time periods, the method further includes: The current icing state is used as the actual label, and a sample pair is formed with the icing prediction probability output by the model at the same time. Count a preset number of sample pairs and calculate the deviation between the predicted icing probability and the measured label. When the deviation value exceeds a preset error threshold, the calibration function is refitted according to a preset scaling method, and the icing prediction probability output by the icing probability prediction model is mapped according to the fitted calibration function to obtain the calibrated icing prediction probability.

10. A system for suppressing icing on wind turbine blades, characterized in that, The wind turbine blade icing suppression system includes an ultrasonic sensor array, an ultrasonic device, and a heating device installed inside the wind turbine blade.

Images