Methods, apparatus, storage media and systems for detecting icing in wind turbines

Abstract

This application provides a method, apparatus, storage medium, and system for detecting icing on wind turbines. The method includes: acquiring current image data of the surface position of any blade of the wind turbine, the top of the nacelle of the wind turbine, and current meteorological data; determining the current motion data of the wind turbine; preprocessing the surface position of any blade, the current meteorological data, the current image data, and the current motion data; inputting the preprocessed surface position of any blade, the current motion data, the current image data, and the current meteorological data into an icing detection model to obtain a current icing correction value at any blade surface position; acquiring the current static icing value at the top of the nacelle of the wind turbine; and determining the icing monitoring value at any blade surface position based on the current icing correction value and the current static icing value, which can comprehensively perceive the icing status of the entire wind turbine blade.

Description

Technical Field

[0001] This application relates to the field of icing detection technology, and specifically to an icing detection method, apparatus, storage medium and system for wind turbines. Background Technology

[0002] Wind turbine blades (hereinafter referred to as wind turbine blades) are key moving components of wind turbines, primarily converting the kinetic energy of wind into mechanical energy, and then into electrical energy. During winter cold waves, ambient temperatures are typically below freezing, and supercooled water droplets may exist in the air. When these supercooled water droplets come into contact with the surface of the wind turbine blades, because the blade surface temperature is lower than the water temperature, the water droplets quickly absorb heat and freeze on the blade surface. Icing on wind turbine blades alters their aerodynamic performance, leading to power loss. It also changes the mass distribution of the blades, causing unstable wind turbine operation, and in severe cases, even blade damage or shutdown. Effective wind turbine blade icing monitoring methods help analyze the operating status of wind turbines and improve the accuracy of wind power prediction during freezing weather.

[0003] Current wind turbine icing monitoring methods mostly involve placing sensors on top of the nacelle. However, as the blades rotate, ice ridges of varying shapes form at the blade tips, leading edges, or trailing edges. A single, static sensor cannot reflect the true overall icing status of the blade. Localized, point-based icing monitoring on the blade surface can only detect specific locations and is also not universally applicable. Therefore, existing single-point static icing detection technologies have significant limitations and cannot comprehensively assess the icing status of the entire blade. Summary of the Invention

[0004] The purpose of this application is to provide a method, device, storage medium and system for detecting icing in wind turbines, in order to solve the problem that the existing single-point static icing detection has great limitations and cannot fully perceive the icing situation of the entire blade.

[0005] To achieve the above objectives, the first aspect of this application provides a method for detecting icing on wind turbines, comprising:

[0006] Acquire the current image data of the position of any blade surface of the wind turbine, the top of the wind turbine nacelle, and the current meteorological data;

[0007] Determine the current motion data of the wind turbine, which includes the rotational speed at any blade surface position, the current yaw angle, and the current pitch angle of any blade.

[0008] Preprocess the location of any blade surface, current meteorological data, current image data, and current motion data;

[0009] Input the preprocessed blade surface position, current motion data, current image data, and current meteorological data into the icing detection model to obtain the current icing correction value at any blade surface position;

[0010] Get the current static value of icing on the top of the wind turbine nacelle;

[0011] The icing monitoring value at any blade surface location is determined based on the current icing correction value and the current static icing value.

[0012] In this embodiment, at least one patch-type capacitive icing sensor is installed on the wind turbine. The icing detection model is trained through the following steps: determining training data, which includes the installation position of each patch-type capacitive icing sensor on the wind turbine, historical image data of the top surface of the wind turbine nacelle, historical meteorological data of the wind turbine, and historical motion data of the wind turbine; inputting the training data into a preset neural network so that the preset neural network outputs historical icing correction values; acquiring historical static icing values ​​of the top of the wind turbine nacelle and blade surface icing monitoring values ​​collected by each patch-type capacitive icing sensor; determining the loss value of the preset neural network based on the historical icing correction value, the historical static icing value, and all blade surface icing monitoring values; and determining that the preset neural network training is complete when the loss value meets preset conditions, thus obtaining the icing detection model.

[0013] In this embodiment of the application, the loss value of the preset neural network is determined by the following formula (1):

[0014] Formula (1)

[0015] in, This refers to the preset loss value of the neural network. This refers to the number of surface-mount capacitive icing sensors. This refers to the historical icing correction value. This refers to historical static values ​​of icing. This refers to the leaf surface icing monitoring value collected by the i-th patch-type capacitive icing sensor.

[0016] In this embodiment, the wind turbine is equipped with meteorological sensors and image acquisition equipment. Determining the training data includes: acquiring the installation position of each patch-type capacitive icing sensor on the wind turbine; receiving historical image data of the top surface of the nacelle acquired by the image acquisition equipment; receiving historical meteorological data acquired by the meteorological sensors; acquiring the historical rotational speed, yaw angle, and pitch angle of the blade where each patch-type capacitive icing sensor is located; determining the rotational speed of the blade where each patch-type capacitive icing sensor is located based on the historical rotational speed and the installation position of each patch-type capacitive icing sensor; determining the historical motion data of the wind turbine based on the rotational speed, yaw angle, and pitch angle; and preprocessing the installation position, historical image data, historical meteorological data, and historical motion data to obtain the training data.

[0017] In this embodiment of the application, determining the rotational speed of the blade where each patch-type capacitive icing sensor is located based on the historical rotational speed and the installation position of each patch-type capacitive icing sensor includes: determining the distance between the installation position of each patch-type capacitive icing sensor and the center point of the wind turbine hub; and determining the rotational speed of each patch-type capacitive icing sensor based on the historical rotational speed of the wind turbine and the distance.

[0018] In this embodiment, the rotational speed of each patch-type capacitive icing sensor is determined by the following formula:

[0019] Formula (2)

[0020] in, Let be the rotational speed of the i-th patch-type capacitive icing sensor. This refers to the distance between the i-th surface-mount capacitive icing sensor and the center point of the wind turbine hub. This refers to the historical rotational speed of the wind turbine where the i-th patch-type capacitive icing sensor is located.

[0021] In this embodiment of the application, determining the icing monitoring value at any blade surface position based on the current icing correction value and the current icing static value includes: determining the total value between the current icing correction value and the current icing static value as the icing monitoring value.

[0022] A second aspect of this application provides an icing detection device for wind turbines, comprising:

[0023] The memory is configured to store instructions;

[0024] The processor is configured to retrieve instructions from memory and, when executing the instructions, to implement the aforementioned icing detection method for wind turbines.

[0025] A third aspect of this application provides a machine-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform the aforementioned icing detection method for a wind turbine.

[0026] The fourth aspect of this application provides an icing detection system for wind turbines, comprising:

[0027] A microwave icing monitoring sensor is installed on the top of the wind turbine nacelle to collect static icing values ​​on the top of the wind turbine nacelle.

[0028] A weather sensor is installed on the top of the wind turbine nacelle to collect current weather data on the top surface of the wind turbine nacelle.

[0029] At least one patch capacitive sensor is installed on the outer surface of the wind turbine blade to collect the blade surface icing monitoring value.

[0030] Image acquisition equipment, installed on the top of the wind turbine nacelle, is used to acquire current image data of the top surface of the wind turbine nacelle;

[0031] The data acquisition unit is used to receive and forward static icing values, current meteorological data, leaf surface icing monitoring values, and current image data.

[0032] The aforementioned icing detection device for wind turbines.

[0033] The above technical solution effectively integrates meteorological data, image data, and motion data, fully considers the impact of wind turbine blade location and blade movement on icing monitoring, and effectively generalizes monitoring data from different points on the blades or even different wind turbines to locations not directly monitored for icing. This avoids the limitation of using a single icing thickness data to represent the icing situation of the wind turbine. By predicting the icing situation on the blade surface, the icing situation of the entire wind turbine blade can be comprehensively perceived.

[0034] Other features and advantages of the embodiments of this application will be described in detail in the following detailed description section. Attached Figure Description

[0035] The accompanying drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the following detailed description to explain the embodiments of this application, but do not constitute a limitation on the embodiments of this application. In the drawings:

[0036] Figure 1 The schematic diagram illustrates a flow chart of an icing detection method for a wind turbine according to an embodiment of this application;

[0037] Figure 2 The schematic diagram illustrates a flow chart of a wind turbine icing detection method according to another embodiment of this application;

[0038] Figure 3 A schematic diagram of an icing detection system for a wind turbine according to an embodiment of this application is shown.

[0039] Figure 4 Internal structure diagram of a computer device according to an embodiment of this application.

[0040] Explanation of reference numerals in the attached figures

[0041] Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for illustration and explanation of the embodiments of this application and are not intended to limit the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0043] It should be noted that if the embodiments of this application involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0044] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0045] Figure 1 A schematic flowchart illustrating an embodiment of the icing detection method for a wind turbine according to this application is shown. Figure 1 As shown in one embodiment of this application, a method for detecting icing on a wind turbine is provided, comprising the following steps:

[0046] Step 101: Acquire the current image data of the position of any blade surface of the wind turbine, the top of the wind turbine nacelle, and the current meteorological data.

[0047] During icing detection of a wind turbine, the processor can acquire current image data of any blade surface position, the top of the turbine nacelle, and current meteorological data. Specifically, an image acquisition device can be installed on the wind turbine, positioned to easily capture images of the top surface of the turbine nacelle. Meteorological sensors can be installed on the wind turbine to collect current meteorological data of the top of the turbine nacelle, including but not limited to temperature, atmospheric pressure, relative humidity, wind speed, and wind direction.

[0048] Step 102: Determine the current motion data of the wind turbine, which includes the rotational speed of any blade surface position, the current yaw angle, and the current pitch angle of any blade.

[0049] The processor can determine the current motion data of the wind turbine, including the rotational speed of any blade surface position, the current yaw angle, and the current pitch angle of any blade. Specifically, the processor can acquire the current rotational speed of the wind turbine, determine the distance between any blade surface position and the center point of the wind turbine hub, and determine the rotational speed of any blade surface position based on the current rotational speed and this distance. The current yaw angle and the current pitch angle of any blade can be obtained through, but are not limited to, sensor monitoring.

[0050] Step 103: Preprocess any blade surface position, current meteorological data, current image data, and current motion data.

[0051] Because the acquired or determined data comes from different acquisition devices, it has significant differences in data structure and characteristics. Therefore, the processor can preprocess any blade surface position, current meteorological data, current image data, and current motion data. Preprocessing methods can include repeated expansion and combined processing.

[0052] Step 104: Input the preprocessed blade surface position, current motion data, current image data, and current meteorological data into the icing detection model to obtain the current icing correction value at any blade surface position.

[0053] The processor can input preprocessed data on any blade surface location, current motion data, current image data, and current meteorological data into the icing detection model to obtain the current icing correction value at any blade surface location. The icing detection model is trained based on a preset neural network, which may include a multi-head attention layer, a forward propagation layer, and an output layer.

[0054] In this embodiment, at least one patch-type capacitive icing sensor is installed on the wind turbine. The icing detection model is trained through the following steps: determining training data, which includes the installation position of each patch-type capacitive icing sensor on the wind turbine, historical image data of the top surface of the wind turbine nacelle, historical meteorological data of the wind turbine, and historical motion data of the wind turbine; inputting the training data into a preset neural network so that the preset neural network outputs historical icing correction values; acquiring historical static icing values ​​of the top of the wind turbine nacelle and blade surface icing monitoring values ​​collected by each patch-type capacitive icing sensor; determining the loss value of the preset neural network based on the historical icing correction value, the historical static icing value, and all blade surface icing monitoring values; and determining that the preset neural network training is complete when the loss value meets preset conditions, thus obtaining the icing detection model.

[0055] At least one patch-type capacitive icing sensor is installed on the wind turbine. This patch-type capacitive sensor uses a silicon diaphragm capacitive sensor as its core, measuring the ice thickness on the outer surface of the blades based on the equivalent capacitance change caused by icing. A flexible base plate is used to firmly and reliably attach to the blade surface without affecting the aerodynamic shape of the blades. The device is powered by solar panels and a power management system. Data transmission with the E-data acquisition cabinet is achieved using a ZigBee wireless transmission component, offering advantages of low power consumption and low latency over distances of up to 100 meters.

[0056] The processor can determine the training data, which includes the installation location of each patch capacitive icing sensor on the wind turbine, historical image data of the top surface of the wind turbine nacelle, historical meteorological data of the wind turbine, and historical motion data of the wind turbine.

[0057] In this embodiment, the wind turbine is equipped with meteorological sensors and image acquisition equipment. Determining the training data includes: acquiring the installation position of each patch-type capacitive icing sensor on the wind turbine; receiving historical image data of the top surface of the nacelle acquired by the image acquisition equipment; receiving historical meteorological data acquired by the meteorological sensors; acquiring the historical rotational speed, yaw angle, and pitch angle of the blade where each patch-type capacitive icing sensor is located; determining the rotational speed of the blade where each patch-type capacitive icing sensor is located based on the historical rotational speed and the installation position of each patch-type capacitive icing sensor; determining the historical motion data of the wind turbine based on the rotational speed, yaw angle, and pitch angle; and preprocessing the installation position, historical image data, historical meteorological data, and historical motion data to obtain the training data.

[0058] The processor can acquire the installation location of each patch-type capacitive icing sensor on the wind turbine. These locations can be the blade tip, root, leading edge, and trailing edge. The processor can receive historical image data of the nacelle's top surface from image acquisition equipment and historical meteorological data from meteorological sensors. This historical meteorological data may include, but is not limited to, temperature, atmospheric pressure, relative humidity, wind speed, and wind direction. The processor can also acquire historical wind turbine speed, yaw angle, and pitch angle of each blade containing a patch-type capacitive icing sensor from the wind turbine's active system.

[0059] The processor can determine the rotational speed of the blade where each patch-type capacitive icing sensor is located based on historical rotational speed and the installation position of each patch-type capacitive icing sensor. In this embodiment, determining the rotational speed of the blade where each patch-type capacitive icing sensor is located based on historical rotational speed and the installation position of each patch-type capacitive icing sensor includes: determining the distance between the installation position of each patch-type capacitive icing sensor and the center point of the wind turbine hub; and determining the rotational speed of each patch-type capacitive icing sensor based on the historical rotational speed of the wind turbine and the distance.

[0060] The processor can determine the distance between the mounting position of each patch capacitive icing sensor and the center point of the wind turbine hub, and can determine the rotational speed of each patch capacitive icing sensor based on the historical rotational speed of the wind turbine and the distance. Specifically, in the embodiments of this application, the rotational speed of each patch capacitive icing sensor is determined by the following formula:

[0061] Formula (2)

[0062] in, Let be the rotational speed of the i-th patch-type capacitive icing sensor. This refers to the distance between the i-th surface-mount capacitive icing sensor and the center point of the wind turbine hub. This refers to the historical rotational speed of the wind turbine where the i-th patch-type capacitive icing sensor is located.

[0063] The processor can determine the wind turbine's historical motion data based on rotational speed, yaw angle, and pitch angle. It can also preprocess installation location, historical image data, historical meteorological data, and historical motion data to obtain training data. Specifically, the location, meteorological, and motion data features are essentially floating-point numbers f, which are repeatedly expanded into n-dimensional vectors. Then, within each category, vector addition, Hadmar product, and other methods are used to perform preliminary fusion and expansion of the feature representations within each category, resulting in a set of... Image data is typically represented as a multi-channel matrix with high dimensionality. Feature extraction is performed using pre-trained image models (such as VGG and ResNet) to obtain an n-dimensional vector representation, Img_ext. Subsequently, various features are preprocessed and concatenated to obtain k n-dimensional vector representations. The multi-head self-attention mechanism within the Transformer structure is used to model the interactions and relationships between different categories of features, effectively fusing heterogeneous modal features while ensuring model stability. Finally, a fully connected layer is used as the final output layer, outputting a floating-point number as the icing monitoring correction value. .

[0064] The processor can input training data into a preset neural network, causing the network to output historical icing correction values. The processor can acquire historical static icing values ​​from the top of the wind turbine nacelle, as well as blade surface icing monitoring values ​​collected by each patch-type capacitive icing sensor. The historical static icing values ​​can be obtained from microwave icing monitoring sensors on the wind turbine. The processor can determine the loss value of the preset neural network based on the historical icing correction values, the historical static icing values, and all blade surface icing monitoring values.

[0065] In this embodiment of the application, the loss value of the preset neural network is determined by the following formula (1):

[0066] Formula (1)

[0067] in, This refers to the preset loss value of the neural network. This refers to the number of surface-mount capacitive icing sensors. This refers to the historical icing correction value. This refers to historical static values ​​of icing. This refers to the leaf surface icing monitoring value collected by the i-th patch-type capacitive icing sensor.

[0068] If the loss value meets the preset conditions, the processor can determine that the preset neural network training is complete and obtain the icing detection model. The preset conditions can be that the loss value reaches a certain range or that the loss value is less than a preset value. They can be set according to actual needs so that the model output results are similar to the blade surface monitoring results after model correction.

[0069] The processor can input the preprocessed blade surface position, current motion data, current image data, and current meteorological data into the icing detection model to obtain the current icing correction value at any blade surface position.

[0070] Step 105: Obtain the current static value of icing on the top of the wind turbine nacelle.

[0071] The processor can acquire the current static value of icing on the top of the wind turbine nacelle. Specifically, a microwave icing monitoring sensor can be installed on the wind turbine to collect the current static value of icing on the top of the wind turbine nacelle.

[0072] Step 106: Determine the icing monitoring value at any blade surface location based on the current icing correction value and the current static icing value.

[0073] The processor can determine the icing monitoring value at any blade surface location based on the current icing correction value and the current static icing value.

[0074] In this embodiment of the application, determining the icing monitoring value at any blade surface position based on the current icing correction value and the current icing static value includes: determining the total value between the current icing correction value and the current icing static value as the icing monitoring value.

[0075] The processor can determine the total value between the current icing correction value and the current icing static value as the icing monitoring value.

[0076] like Figure 2 As shown, a flowchart of another method for detecting icing in wind turbines is provided.

[0077] During icing monitoring, location data, meteorological data, motion data, and image data can be acquired. Meteorological data may include temperature, relative humidity, atmospheric pressure, wind speed, and wind direction. Motion data may include engine speed, yaw angle, and pitch angle. Subsequently, feature combinations can be performed on the location, meteorological, and motion data, and image data can be pre-trained and extracted to obtain a fused, complete monitoring dataset.

[0078] The monitoring data can then be sequentially input into the multi-head attention layer, forward propagation layer, and output layer to obtain the icing thickness correction value. Static icing data can be acquired, referring to the icing monitoring values ​​of the nacelle roof from the microwave icing monitoring sensor. Finally, the sum of the static icing data and the icing thickness correction value can be determined as the icing value.

[0079] In one embodiment, a method for detecting icing on a wind turbine is provided, comprising the following steps:

[0080] Step 1: Based on the 3D model of the wind turbine blades and actual monitoring needs, select several monitoring points on the wind turbine blades (such as the blade tip, blade root, leading edge, and trailing edge of each blade), and install multiple patch-type capacitive icing sensors C, numbered sequentially as follows: Record the distance from the blade to the center point of the hub. (i.e., the distance from the installation location of each surface-mount capacitive sensor to the center point of the wind turbine hub) and the three-dimensional coordinates of the blade surface. (i.e., the three-dimensional mounting position of each patch capacitive sensor).

[0081] Step 2: Install microwave icing monitoring sensor A, meteorological sensor and image acquisition equipment on the top of the wind turbine nacelle.

[0082] Step 3: The data acquisition cabinet periodically collects the icing monitoring values ​​of the nacelle roof from microwave icing monitoring sensor A. Leaf surface icing monitoring values ​​of each patch-type capacitive icing sensor C Meteorological data from meteorological sensor B (temperature Te, atmospheric pressure Pr, relative humidity Rh, wind speed Ws, wind direction Wd) and icing conditions on the top surface of the cabin (Img).

[0083] Step 4: The data acquisition cabinet transmits the collected data back to the host computer, which then obtains the wind turbine's rotational speed Rs (revolutions per minute, rpm), yaw angle Ay, and corresponding pitch angle Ap from the wind turbine's main control system in real time, and integrates them into monitoring data. ,in Indicates the rotational speed of the sensor. This indicates the opening and closing status of the blade on the blade where the sensor is located.

[0084]

[0085] Step 5: Construct a multi-sensor collaborative sensing model to correct the static icing monitoring values ​​on the top of the cabin by comprehensively considering sensor location, meteorological, motion, and image data. This data can serve as a label for the icing thickness at this monitoring point, and the collaborative sensing model can be used to fit it with other data to the static icing monitoring value of the cabin top. The difference.

[0086] Because the data comes from different acquisition devices, there are significant differences in the structure and characteristics of the data. Preprocessing and internal expansion are carried out before fusion.

[0087] Location, meteorological, and motion data features are primarily floating-point numbers f, which are repeatedly expanded into n-dimensional vectors. Then, within each category, vector addition, Hadmar product, and other methods are used to perform preliminary fusion and expansion of the feature representations within each category, resulting in a set of... .

[0088] Image data is typically represented as a multi-channel matrix with high dimensionality. Image pre-trained models (such as VGG and ResNet) are used for feature extraction to obtain an n-dimensional vector representation, Img_ext. After preprocessing and concatenation of various features, k n-dimensional vector representations are obtained. The multi-head self-attention mechanism within the Transformer structure is used to model the interactions and relationships between different categories of features, effectively fusing heterogeneous modal features while ensuring model stability. A fully connected layer is used as the final output layer, outputting a floating-point number as the icing monitoring correction value. .

[0089] Step 6: Repeat steps 3-5, and use the data collected from each monitoring point to train the multi-sensor collaborative sensing model constructed in step 5. Use mean square error (MSE) as the loss function so that the model output results are similar to the blade surface monitoring results after the model is corrected.

[0090] Step 7: Given any three-dimensional coordinates of the blade surface By combining current meteorological, motion, and image data with the model trained in step 6, the corresponding icing monitoring results can be output, thus generalizing the monitoring data to any location on the blade surface.

[0091] The above scheme effectively integrates meteorological data, image data, and motion data, fully considers the impact of wind turbine blade location and blade movement on icing monitoring, and effectively generalizes monitoring data from different points on the blades and even different wind turbines to locations not directly monitored for icing. This avoids the limitation of using a single icing thickness data to represent the icing situation of the wind turbine. By predicting the icing situation on the blade surface, the icing situation of the entire wind turbine blade can be comprehensively perceived.

[0092] Figure 1 and Figure 2 This is a flowchart illustrating a method for detecting icing on a wind turbine in one embodiment. It should be understood that, although... Figure 1 and Figure 2 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise explicitly stated herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 and Figure 2 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0093] In one embodiment, an icing detection device for a wind turbine is provided, comprising:

[0094] The memory is configured to store instructions;

[0095] The processor is configured to retrieve instructions from memory and, when executing the instructions, to implement the aforementioned icing detection method for wind turbines.

[0096] In one embodiment, a storage medium is provided on which a program is stored, which, when executed by a processor, implements the above-described method for detecting icing in a wind turbine.

[0097] In one embodiment, a processor is provided for running a program, wherein the program executes the above-described icing detection method for wind turbines.

[0098] In this application embodiment, an icing detection system for wind turbines is provided, comprising:

[0099] A microwave icing monitoring sensor is installed on the top of the wind turbine nacelle to collect static icing values ​​on the top of the wind turbine nacelle.

[0100] A weather sensor is installed on the top of the wind turbine nacelle to collect current weather data on the top surface of the wind turbine nacelle.

[0101] At least one patch capacitive sensor is installed on the outer surface of the wind turbine blade to collect the blade surface icing monitoring value.

[0102] Image acquisition equipment, installed on the top of the wind turbine nacelle, is used to acquire current image data of the top surface of the wind turbine nacelle;

[0103] The data acquisition unit is used to receive and forward static icing values, current meteorological data, leaf surface icing monitoring values, and current image data.

[0104] The aforementioned icing detection device for wind turbines.

[0105] The microwave icing monitoring sensor includes a microwave antenna, a microwave transceiver module, and a detection circuit. It estimates the dielectric constant of the medium by utilizing the attenuation of microwaves during transmission through a specific medium, thus determining the icing thickness on the microwave antenna. The meteorological sensor collects current meteorological data from the top surface of the wind turbine nacelle, including temperature, atmospheric pressure, relative humidity, wind speed, and wind direction. The patch-type capacitive sensor, based on a silicon diaphragm capacitive sensor, measures the icing thickness on the blade's outer surface based on the equivalent capacitance change caused by icing. It uses a flexible base plate to firmly and reliably attach to the blade surface without affecting the blade's aerodynamic shape. Power is harvested from the device via solar panels and a power management system. Data is transmitted via ZigBee wireless transmission to the data acquisition cabinet, offering advantages of low power consumption and low latency over distances of hundreds of meters. Image acquisition equipment can be cameras, camcorders, recorders, or other devices with image acquisition capabilities.

[0106] like Figure 3 The diagram shows a schematic of an icing detection system for wind turbines.

[0107] In this system, A is a microwave icing monitoring sensor, B is a meteorological sensor, C is a patch-type capacitive sensor, D is an image acquisition device, E is a data acquisition cabinet, and F is a host computer. The microwave icing monitoring sensor A, meteorological sensor B, and image acquisition device D are located on a flat area on top of the wind turbine nacelle. The patch-type capacitive sensor C is attached to the outer surface of the blades. The data acquisition cabinet E is installed in an empty space within the nacelle. The host computer F is deployed in the relay protection room of the wind farm's booster station.

[0108] The microwave icing monitoring sensor A includes a microwave antenna, a microwave transceiver module, and a detection circuit. It estimates the dielectric constant of the medium by utilizing the degree of microwave attenuation during transmission through a specific medium, and then inversely calculates the icing thickness on the microwave antenna.

[0109] Meteorological sensor B is responsible for collecting data on temperature, atmospheric pressure, relative humidity, wind speed, and wind direction at the top of the nacelle. The patch-type capacitive sensor C, with a silicon diaphragm capacitive sensor at its core, measures the ice thickness on the outer surface of the blades based on the equivalent capacitance change caused by icing. It is firmly and reliably attached to the blade surface using a flexible base plate without affecting the aerodynamic shape of the blades. The device is powered by solar panels and a power management system. Data transmission with the data acquisition cabinet E is achieved using a ZigBee wireless transmission component, offering advantages of low power consumption and low latency over distances of up to 100 meters.

[0110] Image acquisition device D is used to photograph the icing condition of the top surface of the cabin. Acquisition cabinet E is responsible for receiving data from microwave icing monitoring sensor A, meteorological sensor B, patch capacitive sensor C, and image acquisition device D. It is fixed in an empty space in the cabin using a dedicated bracket, can be powered from inside the cabin, and is connected to the tower-based switch via fiber optic cable.

[0111] The host computer F is responsible for summarizing the icing-related data collected from each wind turbine in the entire wind farm and building a multi-sensor collaborative perception model. The monitoring data is mainly stored in a structured database, and a high-performance GPU is used for deep learning model training and inference.

[0112] Data acquisition cabinet E is responsible for data acquisition. The microwave icing monitoring sensor A, meteorological sensor B, and image acquisition device D, located on the top of the nacelle, are connected to data acquisition cabinet E via wired communication. The patch capacitive sensor C transmits data to data acquisition cabinet E wirelessly. Each wind turbine's data acquisition cabinet E is connected to the wind turbine tower base switch via fiber optic cables laid inside the tower. The data is then transmitted to the host computer F in the wind farm's booster station relay protection room via the wind farm's fiber optic ring network for aggregation. The host computer F stores the various monitoring data, performs data preprocessing and labeling for the regression problem of icing thickness monitoring, uses GPU to perform deep learning heterogeneous modal feature fusion training, and is responsible for wind turbine blade icing prediction.

[0113] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 4 As shown. The computer device includes a processor A01, a network interface A02, memory (not shown), and a database (not shown) connected via a system bus. The processor A01 provides computing and control capabilities. The memory includes internal memory A03 and a non-volatile storage medium A04. The non-volatile storage medium A04 stores an operating system B01, a computer program B02, and a database (not shown). The internal memory A03 provides an environment for the operation of the operating system B01 and the computer program B02 stored in the non-volatile storage medium A04. The database stores icing detection data. The network interface A02 communicates with external terminals via a network connection. When the processor A01 executes the computer program B02, it implements an icing detection method for wind turbines.

[0114] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0115] This application provides an apparatus including a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it performs the following steps: acquiring the position of any blade surface of a wind turbine, current image data of the top of the wind turbine nacelle, and current meteorological data; determining the current motion data of the wind turbine, including the rotational speed, current yaw angle, and current pitch angle of any blade surface; preprocessing the position of any blade surface, current meteorological data, current image data, and current motion data; inputting the preprocessed position of any blade surface, current motion data, current image data, and current meteorological data into an icing detection model to obtain a current icing correction value at any blade surface position; acquiring the current static icing value of the top of the wind turbine nacelle; and determining the icing monitoring value at any blade surface position based on the current icing correction value and the current static icing value.

[0116] In one embodiment, at least one patch-type capacitive icing sensor is installed on the wind turbine. The icing detection model is trained through the following steps: determining training data, which includes the installation location of each patch-type capacitive icing sensor on the wind turbine, historical image data of the top surface of the wind turbine nacelle, historical meteorological data of the wind turbine location, and historical motion data of the wind turbine; inputting the training data into a preset neural network to make the preset neural network output historical icing correction values; acquiring historical static icing values ​​of the top surface of the wind turbine nacelle and blade surface icing monitoring values ​​collected by each patch-type capacitive icing sensor; determining the loss value of the preset neural network based on the historical icing correction value, the historical static icing value, and all blade surface icing monitoring values; and determining that the preset neural network training is complete when the loss value meets preset conditions, thus obtaining the icing detection model.

[0117] In one embodiment, the loss value of the preset neural network is determined by the following formula (1):

[0118] Formula (1)

[0119] in, This refers to the preset loss value of the neural network. This refers to the number of surface-mount capacitive icing sensors. This refers to the historical icing correction value. This refers to historical static values ​​of icing. This refers to the leaf surface icing monitoring value collected by the i-th patch-type capacitive icing sensor.

[0120] In one embodiment, the wind turbine is equipped with meteorological sensors and image acquisition equipment. Determining the training data includes: acquiring the installation location of each patch-type capacitive icing sensor on the wind turbine; receiving historical image data of the nacelle top surface acquired by the image acquisition equipment; receiving historical meteorological data acquired by the meteorological sensors; acquiring the wind turbine's historical rotational speed, yaw angle, and pitch angle of the blade where each patch-type capacitive icing sensor is located; determining the rotational speed of the blade where each patch-type capacitive icing sensor is located based on the historical rotational speed and the installation location of each patch-type capacitive icing sensor; determining the wind turbine's historical motion data based on the rotational speed, yaw angle, and pitch angle; and preprocessing the installation location, historical image data, historical meteorological data, and historical motion data to obtain training data.

[0121] In one embodiment, determining the rotational speed of the blade where each patch-type capacitive icing sensor is located, based on historical rotational speed and the installation position of each patch-type capacitive icing sensor, includes: determining the distance between the installation position of each patch-type capacitive icing sensor and the center point of the wind turbine hub; and determining the rotational speed of each patch-type capacitive icing sensor based on the historical rotational speed of the wind turbine and the distance.

[0122] In one embodiment, the rotational speed of each patch-type capacitive icing sensor is determined by the following formula:

[0123] Formula (2)

[0124] in, Let be the rotational speed of the i-th patch-type capacitive icing sensor. This refers to the distance between the i-th surface-mount capacitive icing sensor and the center point of the wind turbine hub. This refers to the historical rotational speed of the wind turbine where the i-th patch-type capacitive icing sensor is located.

[0125] In one embodiment, determining the icing monitoring value at any blade surface location based on the current icing correction value and the current icing static value includes: determining the total value between the current icing correction value and the current icing static value as the icing monitoring value.

[0126] This application also provides a computer program product that, when executed on a data processing device, is adapted to perform a program that initializes a method for detecting icing in a wind turbine.

[0127] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0128] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0129] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0130] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0131] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0132] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0133] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0134] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0135] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for detecting icing on wind turbines, characterized in that, The icing detection method includes: Acquire the position of any blade surface of the wind turbine, the current image data of the top of the wind turbine nacelle, and the current meteorological data; Determine the current motion data of the wind turbine, which includes the rotational speed at any blade surface position, the current yaw angle, and the current pitch angle of any blade. Preprocessing is performed on the surface position of any blade, the current meteorological data, the current image data, and the current motion data; The preprocessed position of any blade surface, current motion data, current image data, and current meteorological data are input into the icing detection model to obtain the current icing correction value at any blade surface position. Obtain the current static icing value of the top of the nacelle of the wind turbine; The icing monitoring value at any blade surface location is determined based on the current icing correction value and the current static icing value. The aforementioned icing detection method for wind turbines includes at least one patch-type capacitive icing sensor installed on the wind turbine. The icing detection model is trained through the following steps: The training data includes the installation location of each patch-type capacitive icing sensor on the wind turbine, historical image data of the top surface of the wind turbine nacelle, historical meteorological data of the wind turbine, and historical motion data of the wind turbine. The training data is input into a preset neural network so that the preset neural network outputs historical icing correction values; The historical static value of icing on the top of the nacelle of the wind turbine is obtained, as well as the blade surface icing monitoring value collected by each patch capacitive icing sensor. The loss value of the preset neural network is determined based on the historical icing correction value, the historical icing static value, and all leaf surface icing monitoring values. If the loss value meets the preset conditions, the preset neural network training is determined to be complete, and the ice detection model is obtained. The loss value of the preset neural network is determined by the following formula (1): Formula (1) in, This refers to the preset loss value of the neural network. This refers to the number of surface-mount capacitive icing sensors. This refers to the historical icing correction value. This refers to historical static values ​​of icing. This refers to the leaf surface icing monitoring value collected by the i-th patch-type capacitive icing sensor; In addition, the wind turbine is equipped with meteorological sensors and image acquisition equipment, and the training data to be determined includes: Obtain the installation location of each surface-mount capacitive icing sensor on the wind turbine; Receive historical image data of the cabin roof surface acquired by the image acquisition device; Receive historical meteorological data collected by the meteorological sensor; The historical rotational speed, yaw angle, and pitch angle of each blade containing a patch capacitive icing sensor of the wind turbine are obtained. The rotational speed of the blade where each patch capacitive icing sensor is located is determined based on the historical rotational speed and the installation position of each patch capacitive icing sensor. The historical motion data of the wind turbine are determined based on the rotational speed, the yaw angle, and the pitch angle. The installation location, historical image data, historical meteorological data, and historical motion data are preprocessed to obtain the training data; The step of determining the icing monitoring value at any blade surface location based on the current icing correction value and the current static icing value includes: The total value between the current icing correction value and the current icing static value is determined as the icing monitoring value.

2. The method for detecting icing in wind turbines according to claim 1, characterized in that, The determination of the rotational speed of the blade where each patch-type capacitive icing sensor is located, based on the historical rotational speed and the installation position of each patch-type capacitive icing sensor, includes: Determine the distance between the installation location of each patch capacitive icing sensor and the center point of the wind turbine hub; The rotational speed of each patch-type capacitive icing sensor is determined based on the historical rotational speed of the wind turbine and the distance.

3. The icing detection method for wind turbines according to claim 2, characterized in that, The rotational speed of each patch-type capacitive icing sensor is determined by the following formula: Formula (2) in, Let be the rotational speed of the i-th patch-type capacitive icing sensor. This refers to the distance between the i-th patch-type capacitive icing sensor and the center point of the wind turbine hub. This refers to the historical rotational speed of the wind turbine where the i-th patch-type capacitive icing sensor is located.

4. An icing detection device for wind turbines, characterized in that, The icing detection device includes: The memory is configured to store instructions; The processor is configured to retrieve the instructions from the memory and, when executing the instructions, to implement the icing detection method for a wind turbine according to any one of claims 1 to 3.

5. An icing detection system for wind turbines, characterized in that, include: Ice detection device for wind turbines according to claim 4; The microwave icing monitoring sensor is installed on the top of the wind turbine nacelle to collect static icing values ​​on the top of the wind turbine nacelle. A weather sensor is installed on the top of the wind turbine nacelle to collect current weather data on the top surface of the wind turbine nacelle. At least one patch-type capacitive icing sensor is installed on the outer surface of the wind turbine blade to collect the blade surface icing monitoring value. Image acquisition equipment, installed on the top of the wind turbine nacelle, is used to acquire current image data of the top surface of the wind turbine nacelle; The data acquisition unit is used to receive and forward the static icing value, the current meteorological data, the leaf surface icing monitoring value, and the current image data.

6. A machine-readable storage medium storing instructions thereon, characterized in that, When executed by a processor, this instruction causes the processor to be configured to perform the icing detection method for a wind turbine according to any one of claims 1 to 3.

Images