Fan blade icing state intelligent classification method and system based on image analysis

Abstract

The invention provides a fan blade icing state intelligent classification method and system based on image analysis, and relates to the technical field of image analysis, and the method comprises the steps: obtaining initial image information when a fan blade is iced; target image information is obtained through defogging processing, denoising processing, edge sharpening processing and graying processing; classifying according to the texture features to obtain image classification information; performing image edge detection on the second image information and calculating the icing thickness so as to perform grade division on the icing to obtain image grading information; and making a decision to determine a corresponding processing strategy. From two different angles, icing states of different types and different severity degrees are comprehensively considered, a comprehensive decision is made according to the image classification information and the image grading information, processing can be performed by using unlabeled data, the detection precision can be ensured, the icing state of the fan blade can be accurately judged, and the detection accuracy is improved. And a corresponding strategy is adopted to deal with the icing phenomenon of the fan blade.

Description

Technical Field

[0001] The present application relates to the technical field of image analysis, and in particular to an intelligent classification method and system for the icing status of wind turbine blades based on image analysis. Background Art

[0002] As a clean and efficient renewable energy source, wind energy plays an important role in energy conservation and emission reduction. The scale and number of wind farms are increasing year by year. With the increase in the number of wind farms put into operation, their safe operation has also received widespread attention. Wind farms generally have the problem of wind turbine blades covered with ice in winter. The types of ice cover include rime and fog, and rime has the greatest impact on wind farms. Blade ice cover will not only lead to the deterioration of the aerodynamic performance of the wind turbine and the reduction of wind energy capture ability, but also may threaten the lives of workers and local residents.

[0003] In the related art, wind turbine blade icing detection is mainly carried out by analyzing SCADA data sets and image data, and by studying characteristic values ​​to detect the real-time icing status or predict the icing status in the future, so as to take timely protective measures. However, the related art has the following defects in the process of wind turbine blade icing detection:

[0004] (1) Existing studies have not considered that different types and severity of wind turbine blade icing require different deicing strategies. Using a single deicing strategy to deal with different types and severity of icing may result in a waste of resources and reduced efficiency, and may cause unnecessary damage to wind turbine blades. If the deicing strategy cannot adapt to the specific icing situation, it will not only fail to effectively solve the problem, but may also increase maintenance costs and safety risks.

[0005] (2) Existing studies often use machine learning models such as convolutional neural networks and support vector machines in the process of training models. These models usually require a large amount of labeled data to ensure the effectiveness of training and the accuracy of the model. However, obtaining a large amount of labeled data is not only time-consuming and labor-intensive, but also costly. Summary of the invention

[0006] In response to the shortcomings of the prior art, the present application provides a method and system for intelligent classification of the icing status of wind turbine blades based on image analysis, which solves the problems of low efficiency in the de-icing process and the high cost of current icing detection of wind turbine blades that relies on labeled data.

[0007] To achieve the above objectives, this application is implemented through the following technical solutions:

[0008] In a first aspect, an embodiment of the present application provides an intelligent classification method for the icing status of wind blades based on image analysis, the intelligent classification method for the icing status of wind blades comprising: obtaining initial image information when the wind blades are covered with ice, the initial image information comprising a windward surface image and a cross-sectional image, wherein the cross-sectional image is obtained by photographing the windward surface image at a preset target distance from the wind blades by a photographing device; performing defogging, denoising, edge sharpening, and grayscale processing on the initial image information to obtain target image information; wherein the target image information comprises first image information and second image information corresponding to the windward surface image and the cross-sectional image, respectively; classifying the first image information according to texture features to obtain image classification information; performing image edge detection on the second image information, calculating the ice thickness according to the edge detection result to determine the ice weight, and grading the ice according to the ratio of the ice weight to the weight of the wind blades to obtain image grading information; making a decision based on the image classification information and the image grading information to determine the corresponding processing strategy to deal with the icing phenomenon of the wind blades.

[0009] In a second aspect, an embodiment of the present application provides an intelligent classification system for the icing status of wind turbine blades based on image analysis, and the intelligent classification system for the icing status of wind turbine blades includes: a data acquisition module, an image preprocessing module, a type recognition module, a grade classification module and a decision output module.

[0010] Specifically, the data acquisition module is used to obtain initial image information when the wind turbine blades are covered with ice, and the initial image information includes a windward surface image and a cross-sectional image, wherein the cross-sectional image is obtained by photographing the windward surface image at a preset target distance from the wind turbine blades by a photographing device; the image preprocessing module is used to perform defogging, denoising, edge sharpening, and grayscale processing on the initial image information to obtain target image information; wherein the target image information includes first image information and second image information corresponding to the windward surface image and the cross-sectional image, respectively; the type recognition module is used to classify the first image information according to texture features to obtain image classification information; the grading module is used to perform image edge detection on the second image information, calculate the ice thickness according to the edge detection result to determine the ice weight, and grade the ice according to the ratio of the ice weight to the wind turbine blade weight to obtain image grading information; the decision output module is used to make a decision based on the image classification information and the image grading information to determine the corresponding processing strategy to deal with the ice covering phenomenon on the wind turbine blades.

[0011] In a third aspect, an embodiment of the present application provides an electronic device, comprising: a processor, a memory, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the method for intelligently classifying the icing status of wind turbine blades based on image analysis in the aforementioned first aspect is implemented.

[0012] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium storing a program or instruction, which, when executed by a processor, implements the method for intelligent classification of wind turbine blade icing status based on image analysis in the aforementioned first aspect.

[0013] This application provides a method and system for intelligently classifying the ice-covered state of wind turbine blades based on image analysis. Compared with the prior art, it has the following beneficial effects:

[0014] The present application starts from two different perspectives, comprehensively considering different types and severity of icing conditions to decide on the corresponding de-icing strategy; after the initial image information is defogged, denoised, edge sharpened, and grayed to obtain the target image information, the first image information in the target image information is classified according to texture features to obtain image classification information; image edge detection is performed on the second image information in the target image information, and image classification information is obtained through analysis; the present application makes a comprehensive decision based on the image classification information and the image classification information, can use unlabeled data for processing and ensure detection accuracy, can accurately judge the icing state of the wind turbine blades, and adopt corresponding strategies to deal with the icing phenomenon of the wind turbine blades. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 It is a flow chart of an intelligent classification method for wind turbine blade icing status based on image analysis provided in an embodiment of the present application;

[0017] Figure 2 yes Figure 1 An exemplary flow chart of S130;

[0018] Figure 3 It is a structural schematic diagram of an intelligent classification system for wind turbine blade icing status based on image analysis provided in an embodiment of the present application;

[0019] Figure 4 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0022] The embodiments of the present application provide a method and system for intelligently classifying the icing status of wind turbine blades based on image analysis, thereby solving the problems of low efficiency in the deicing process and high cost of current icing detection of wind turbine blades which relies on labeled data.

[0023] The technical solution in the embodiment of the present application is to solve the above technical problems, and the overall idea is as follows:

[0024] As a clean and efficient renewable energy, wind energy plays an important role in energy conservation and emission reduction. The scale and number of wind farms are increasing year by year. With the increase in the number of wind farms put into operation, their safe operation has also received widespread attention. Wind farms generally have the problem of ice covering the wind turbine blades in winter.

[0025] Ice covering types include rime and rime. Rime refers to the formation of white ice crystals when supercooled fog droplets condense on objects in a low-temperature and humid environment. Such ice crystals have a fluffy structure and are easy to melt. Rime refers to the formation of a hard ice layer when supercooled ice and snow adhere to objects or the ground below 0°C during the descent process. It has a dense structure and is not easy to melt. Rime has a greater impact on wind farms. Blade icing will not only deteriorate the aerodynamic performance of the wind turbine and reduce its ability to capture wind energy, but may also threaten the lives of workers and local residents.

[0026] In the related art, wind turbine blade icing detection is mainly carried out by analyzing SCADA data sets and image data, and by studying characteristic values ​​to detect the real-time icing status or predict the icing status in the future, so as to take timely protective measures. However, the related art has the following defects in the process of wind turbine blade icing detection:

[0027] (1) Existing studies have not considered that different types and severity of wind turbine blade icing require different deicing strategies. Using a single deicing strategy to deal with different types and severity of icing may result in a waste of resources and reduced efficiency, and may cause unnecessary damage to wind turbine blades. If the deicing strategy cannot adapt to the specific icing situation, it will not only fail to effectively solve the problem, but may also increase maintenance costs and safety risks.

[0028] (2) Existing studies often use machine learning models such as convolutional neural networks and support vector machines in the process of training models. These models usually require a large amount of labeled data to ensure the effectiveness of training and the accuracy of the model. However, obtaining a large amount of labeled data is not only time-consuming and labor-intensive, but also costly.

[0029] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0030] The following first introduces an intelligent classification method for wind turbine blade icing status based on image analysis provided in an embodiment of the present application.

[0031] The present application embodiment provides a flow chart of a method for intelligently classifying the icing status of wind turbine blades based on image analysis, as shown in FIG. Figure 1 As shown, the method for intelligently classifying wind turbine blade icing status based on image analysis may include the following steps S110-S150.

[0032] S110, obtaining initial image information of wind turbine blades covered with ice, the initial image information comprising a windward surface image and a cross-sectional image, wherein the cross-sectional image is obtained by photographing a camera at a preset target distance from the wind turbine blades.

[0033] S120, performing defogging, denoising, edge sharpening, and grayscale processing on the initial image information to obtain target image information; wherein the target image information includes first image information and second image information corresponding to the windward surface image and the cross-sectional image, respectively.

[0034] S130: Classify the first image information according to texture features to obtain image classification information.

[0035] S140, performing image edge detection on the second image information, calculating the ice thickness according to the edge detection result to determine the ice weight, and classifying the ice according to the ratio of the ice weight to the wind turbine blade weight to obtain image classification information.

[0036] S150: Make a decision based on the image classification information and the image grading information to determine a corresponding processing strategy to handle the icing phenomenon on the wind turbine blades.

[0037] The above is a specific implementation method of an intelligent classification method for wind turbine blade icing status based on image analysis provided in an embodiment of the present application. It can be understood that the present application starts from two different angles and comprehensively considers icing status of different types and severity to decide on the corresponding de-icing strategy; after the initial image information is dehazed, denoised, edge sharpened, and grayed to obtain the target image information, the first image information in the target image information is classified according to texture features to obtain image classification information; image edge detection is performed on the second image information in the target image information, and image grading information is obtained through analysis.

[0038] The present application collects ice-covered images of wind turbine blades at close range to perform relatively accurate type identification and grade classification. The present application makes comprehensive decisions based on image classification information and image grading information, and can accurately determine the ice coverage status of wind turbine blades. Different de-icing strategies are proposed according to different ice coverage situations. This can not only improve the targeted de-icing operations, but also effectively reduce maintenance costs and reduce unnecessary damage to wind turbine blades.

[0039] It should be noted that the present application utilizes image classification technology and image edge detection technology to identify the types and classify the levels of ice covering wind turbine blades, and provides corresponding countermeasures based on the identification and classification results to improve the efficiency of the de-icing process and reduce the maintenance costs during the de-icing process.

[0040] In some embodiments, Figure 2 As shown, the first image information is classified according to the texture feature to obtain image classification information, that is, the above S130 may specifically include the following steps:

[0041] S210, extracting picture texture features in the first image information using a gray level co-occurrence matrix GLCM;

[0042] S220, analyzing the extracted image texture features to determine the relationship between the image texture features and different types of ice coverage;

[0043] S230, using the image texture features and combining with the semi-supervised learning model for training to classify the ice coverage on the wind turbine blades and obtain image classification information;

[0044] S240, verifying the validity of the model through pre-acquired actual ice-covered images, and optimizing the model according to the verification results to improve the accuracy of classification.

[0045] In the embodiments of the present application, it can be understood that the present application adopts a semi-supervised learning method in the image classification process, and improves the accuracy and robustness of the classification model by combining a large amount of unlabeled data, thereby improving the automation level and processing speed of wind turbine blade ice type detection. The semi-supervised learning model performs in-depth model training and optimization on the features extracted from the wind turbine blade ice image to complete the identification of the wind turbine blade ice type.

[0046] It should be noted that this application adopts a semi-supervised learning method, which makes full use of the information of unlabeled data, avoids data waste, and improves the performance of the model. At the same time, it can also utilize a large amount of unlabeled data, reducing the dependence on labeled data and the need for full data labeling, thereby reducing the cost of data labeling, speeding up model training, and shortening detection time.

[0047] In some embodiments, the gray level co-occurrence matrix GLCM is used to extract the image texture features in the first image information, that is, the aforementioned S210 may specifically include the following steps:

[0048] S310, using a gray level co-occurrence matrix GLCM calculation method to calculate a gray level co-occurrence matrix for the first image information at multiple different directions and different distances;

[0049] S320, extracting image texture features from the gray level co-occurrence matrix GLCM, where the image texture features include energy, contrast, correlation, and entropy to describe the uniformity, clarity, randomness, and complexity of the image texture.

[0050] In the embodiments of the present application, it can be understood that the present application uses a gray level co-occurrence matrix to describe the texture characteristics of an image. The gray level co-occurrence matrix (GLCM) effectively describes the texture characteristics of an image by analyzing the grayscale relationship between pixels in an image and pixels in its surrounding neighborhood. The gray level co-occurrence matrix GLCM can extract multiple key texture parameters including contrast, uniformity, energy, and entropy, and then establish a model based on the key texture parameters to identify the type of ice on wind turbine blades. These parameters play a vital role in identifying and classifying image texture patterns.

[0051] In some embodiments, the aforementioned image edge detection is performed on the second image information, the ice thickness is calculated according to the edge detection result to determine the ice weight, and the ice is graded according to the ratio of the ice weight to the wind turbine blade weight to obtain image grading information, that is, the aforementioned S140 may specifically include the following steps:

[0052] S410, performing image edge detection on the second image information using an edge detection operator to obtain target edge information;

[0053] S420, estimating the ice thickness of the wind turbine blades according to the target edge information;

[0054] S430, determining the ice weight of the wind turbine blades according to the ice thickness;

[0055] S440, determining the ice coverage percentage of the fan blades based on the ice coverage weight and the total weight of the fan blades;

[0056] S450: Determine image classification information according to the ice coverage percentage of the wind turbine blades; the image classification information includes light ice coverage, moderate ice coverage, and heavy ice coverage.

[0057] In the embodiments of the present application, it can be understood that the image edge detection technology can effectively identify and extract the contours of objects in the image for subsequent image analysis and feature extraction. The edge detection operator is an edge detection algorithm that can not only capture subtle edge changes in the image, but also effectively suppress noise to ensure the accuracy and robustness of edge detection.

[0058] This application uses an edge detection operator to capture the edges in the image, uses the relationship between the captured image and the actual object to estimate the thickness of the ice covering the wind blades, and then estimates the weight of the ice covering the wind blades, and classifies the wind blades according to the ratio of the weight of the ice covering the wind blades to the weight of the wind blades.

[0059] In one example, the ice thickness of the wind turbine blade is L h , the ice weight of the wind turbine blade is recorded as W, where W = ρ × A × L h , A is the surface area of ​​the fan blade, ρ is the estimated ice density. The ice coverage percentage of the fan blade is the ratio of the ice weight W to the total weight of the fan blade.

[0060] In another example, light icing corresponds to an ice coverage percentage of 0%-5% on the fan blades, moderate icing corresponds to an ice coverage percentage of 5%-10% on the fan blades, and heavy icing corresponds to an ice coverage percentage of more than 10% on the fan blades.

[0061] In some embodiments, the aforementioned edge detection operator is used to perform image edge detection on the second image information to obtain target edge information, that is, the aforementioned S410 may specifically include the following steps:

[0062] S510, using the Sobel operator to calculate the gradient size and direction of each pixel in the second image information, and obtaining a gradient image to characterize the strength and direction of the edge;

[0063] S520, performing non-maximum suppression on the gradient image to refine the edge to ensure that the edge is a single pixel width and the gradient direction is consistent with the edge direction;

[0064] S530, applying a dual threshold method to set a high threshold and a low threshold, determining that a pixel point greater than the high threshold is a strong edge, determining that a pixel point less than the low threshold is excluded, and determining that a pixel point between the high threshold and the low threshold is a weak edge;

[0065] S540, using hysteresis technology to track edges, marking weak edges connected to strong edges as edges, and deleting other weak edges to obtain target edge information to retain complete edge information.

[0066] In some embodiments, the aforementioned estimation of the ice thickness of the wind turbine blades by using the target edge information, that is, the aforementioned S420 may specifically include the following steps:

[0067] S610, photographing a cube calibration block with a side length of L1 at a distance from the target using a photographing device with the same parameters;

[0068] S620, using an edge detection operator to perform edge detection on the captured image of the cube calibration block to obtain a side length L2 of the cube calibration block in the image;

[0069] S630, measuring the length between the ice-covered edge of the wind turbine blade and the trailing edge of the blade in the image according to the target edge information, and recording it as L3;

[0070] S640, measuring the actual length L4 between the ice-covered edge of the wind turbine blade and the trailing edge of the blade, the actual length L4 satisfies the expression:

[0071] S650: Calculate the ice thickness of the fan blade based on the thickness D of the fan blade. The ice thickness L h Satisfies the expression: L h =L4-D.

[0072] In some embodiments, based on the image classification information and the image grading information, different types and levels of ice cover correspond to different countermeasures, as shown in Table 1:

[0073] Table 1 Corresponding countermeasures for different levels and types of icing

[0074]

[0075]

[0076] In some embodiments, the present application provides an intelligent classification system 700 for wind turbine blade icing status based on image analysis, such as Figure 3As shown, the wind turbine blade icing state intelligent classification system 700 may include the following modules:

[0077] The data acquisition module 710 is used to acquire initial image information when the wind turbine blade is covered with ice, wherein the initial image information includes a windward surface image and a cross-sectional image, wherein the cross-sectional image is acquired by photographing a camera at a preset target distance from the wind turbine blade;

[0078] The image preprocessing module 720 is used to perform defogging, denoising, edge sharpening, and grayscale processing on the initial image information to obtain target image information; wherein the target image information includes first image information and second image information corresponding to the windward surface image and the cross-sectional image respectively;

[0079] A type identification module 730 is used to classify the first image information according to the texture feature to obtain image classification information;

[0080] The grading module 740 is used to perform image edge detection on the second image information, calculate the ice thickness according to the edge detection result to determine the ice weight, and grade the ice according to the ratio of the ice weight to the wind turbine blade weight to obtain image grading information;

[0081] The decision output module 750 is used to make a decision based on the image classification information and the image grading information to determine a corresponding processing strategy to handle the icing phenomenon of the wind turbine blades.

[0082] According to an embodiment of the present application, any multiple modules among the data acquisition module 710, the image preprocessing module 720, the type identification module 730, the level classification module 740 and the decision output module 750 can be combined into one module for implementation, or any one of the modules can be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module.

[0083] In some embodiments, the type identification module 730 may specifically include the following units:

[0084] The texture feature extraction unit 731 is used to extract the image texture feature in the first image information using the gray level co-occurrence matrix GLCM;

[0085] A texture feature analysis unit 732, configured to analyze the extracted image texture features to determine the relationship between the image texture features and different types of ice coverage;

[0086] The classification unit 733 is used to classify the ice coverage on the wind turbine blades by using the image texture features and combining with the semi-supervised learning model for training, so as to obtain image classification information;

[0087] The optimization unit 734 is used to verify the validity of the model through pre-acquired actual ice-covered images, and optimize the model according to the verification results to improve the accuracy of classification.

[0088] In some embodiments, the texture feature extraction unit 731 may be specifically used for:

[0089] Using a gray level co-occurrence matrix GLCM calculation method, the gray level co-occurrence matrix is ​​calculated for the first image information at multiple different directions and different distances;

[0090] Image texture features are extracted from the gray-level co-occurrence matrix GLCM. Image texture features include energy, contrast, correlation and entropy to describe the uniformity, clarity, randomness and complexity of image texture.

[0091] In some embodiments, the level classification module 740 may specifically include the following units:

[0092] The edge detection unit 741 is used to perform image edge detection on the second image information using an edge detection operator to obtain target edge information;

[0093] An ice thickness estimation unit 742, used to estimate the ice thickness of the wind turbine blades by using target edge information;

[0094] An ice-covered weight estimation unit 743, used to determine the ice-covered weight of the wind turbine blades according to the ice-covered thickness;

[0095] An ice coverage percentage determination unit 744, configured to determine an ice coverage percentage of the fan blades based on the ice coverage weight and the total weight of the fan blades;

[0096] The image grading unit 745 is used to determine image grading information according to the ice coverage percentage of the wind turbine blades; the image grading information includes light ice coverage, moderate ice coverage and heavy ice coverage.

[0097] In some embodiments, the edge detection unit 741 may be specifically used for:

[0098] Using the Sobel operator to calculate the gradient magnitude and direction of each pixel in the second image information, and obtaining a gradient image to characterize the strength and direction of the edge;

[0099] Perform non-maximum suppression on the gradient image to refine the edge to ensure that the edge is single-pixel wide and the gradient direction is consistent with the edge direction;

[0100] A double threshold method is applied to set a high threshold and a low threshold, and pixels greater than the high threshold are determined as strong edges, pixels less than the low threshold are excluded, and pixels between the high threshold and the low threshold are determined as weak edges;

[0101] The hysteresis technique is used to track the edges, and the weak edges connected to the strong edges are also marked as edges, while other weak edges are deleted to obtain the target edge information to retain the complete edge information.

[0102] In some embodiments, the ice thickness estimation unit 742 may be specifically used to:

[0103] Use a camera with the same parameters to shoot a cube calibration block with a side length of L1 at a distance from the target;

[0104] Use the edge detection operator to perform edge detection on the captured image of the cube calibration block to obtain the side length L2 of the cube calibration block in the image;

[0105] According to the target edge information, measure the length between the ice-covered edge of the wind turbine blade and the trailing edge of the blade in the image, which is recorded as L3;

[0106] Calculate the actual length L4 between the ice-covered edge of the wind turbine blade and the trailing edge of the blade. The actual length L4 satisfies the expression:

[0107] Based on the thickness D of the wind turbine blade, calculate the ice thickness of the wind turbine blade, ice thickness L h Satisfies the expression: L h =L4-D.

[0108] Figure 3 Each module in the system shown has the function of implementing each step in the aforementioned method for intelligent classification of wind turbine blade icing status based on image analysis, and can achieve its corresponding technical effect. For the sake of concise description, it will not be repeated here.

[0109] In some embodiments, the present application provides an electronic device, the structure diagram of the electronic device is as follows Figure 4 shown.

[0110] The electronic device may include a processor 810 and a memory 820 storing computer program instructions.

[0111] Specifically, the processor 810 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0112] The memory 820 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 820 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 820 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 820 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 820 is a non-volatile solid-state memory.

[0113] The memory 820 may include a read-only memory (ROM), a random access memory (RAM), a disk storage medium device, an optical storage medium device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Therefore, generally, the memory 820 includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer executable instructions, and when the software is executed (e.g., by one or more processors), it can perform the operations described in any one of the above-mentioned embodiments of the method for intelligent classification of wind turbine blade icing status based on image analysis.

[0114] The processor 810 reads and executes the computer program instructions stored in the memory 820 to implement any one of the methods for intelligently classifying wind turbine blade icing states based on image analysis in the above-mentioned embodiments.

[0115] In one example, the electronic device may further include a communication interface 830 and a bus 800. Figure 4 As shown, the processor 810, the memory 820, and the communication interface 830 are connected via a bus 800 and communicate with each other.

[0116] The communication interface 830 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0117] Bus 800 includes hardware, software or both, and the parts of online data flow billing equipment are coupled to each other. For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front-end bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 800 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the application considers any suitable bus or interconnection.

[0118] In addition, in combination with the above-mentioned method for intelligently classifying the icing status of wind turbine blades based on image analysis, the present application embodiment can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when the computer program instructions are executed by the processor, any one of the above-mentioned methods for intelligently classifying the icing status of wind turbine blades based on image analysis is implemented.

[0119] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0120] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0121] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0122] Aspects of the present disclosure are described above with reference to the flowchart and / or block diagram of the method, device (system) and computer program product according to the embodiment of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the function / action specified in one or more boxes of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0123] In summary, compared with the prior art, this application has the following beneficial effects:

[0124] 1. This application starts from two different perspectives and comprehensively considers different types and severity of icing conditions to decide on the corresponding de-icing strategy; this application makes a comprehensive decision based on image classification information and image grading information, can use unlabeled data for processing and ensure detection accuracy, can accurately judge the icing status of wind turbine blades, and adopt corresponding strategies to deal with the icing phenomenon of wind turbine blades.

[0125] 2. This application uses image classification technology and image edge detection technology to identify the types and grade the ice covering on wind turbine blades, and provides corresponding countermeasures based on the identification and classification results to improve the efficiency of the de-icing process and reduce the maintenance cost during the de-icing process.

[0126] 3. This application adopts a semi-supervised learning method, which makes full use of the information of unlabeled data, avoids data waste, and improves the performance of the model. At the same time, it can also utilize a large amount of unlabeled data, reducing the dependence on labeled data and the need for full data labeling, thereby reducing the cost of data labeling, speeding up model training, and shortening detection time.

[0127] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent classification method for wind turbine blade icing status based on image analysis, characterized in that: include: Acquire initial image information when the wind turbine blade is covered with ice, wherein the initial image information includes a windward surface image and a cross-sectional image, wherein the cross-sectional image is obtained by photographing a camera at a preset target distance from the wind turbine blade; Performing defogging, denoising, edge sharpening, and grayscale processing on the initial image information to obtain target image information; wherein the target image information includes first image information and second image information corresponding to the windward surface image and the cross-sectional image, respectively; Classifying the first image information according to texture features to obtain image classification information; Performing image edge detection on the second image information, calculating the ice thickness according to the edge detection result to determine the ice weight, and classifying the ice according to the ratio of the ice weight to the wind turbine blade weight to obtain image classification information; According to the image classification information and the image grading information, a decision is made to determine a corresponding processing strategy to handle the icing phenomenon of the wind turbine blades.

2. The method for intelligently classifying the icing status of wind turbine blades based on image analysis according to claim 1, characterized in that: The classifying the first image information according to the texture feature to obtain image classification information includes: Using a gray level co-occurrence matrix GLCM to extract picture texture features in the first image information; Analyzing the extracted image texture features to determine the relationship between the image texture features and different types of ice coverage; The image texture features are used in combination with a semi-supervised learning model for training to classify ice coverage on wind turbine blades and obtain image classification information; The effectiveness of the model is verified by pre-acquired actual ice-covered images, and the model is optimized based on the verification results to improve the classification accuracy.

3. The method for intelligent classification of wind turbine blade icing status based on image analysis according to claim 2, characterized in that: The extracting the picture texture features in the first image information by using the gray level co-occurrence matrix GLCM includes: Using a gray level co-occurrence matrix GLCM calculation method, the gray level co-occurrence matrix of the first image information is calculated at multiple different directions and different distances; Image texture features are extracted from the gray level co-occurrence matrix GLCM, wherein the image texture features include energy, contrast, correlation and entropy value to describe the uniformity, clarity, randomness and complexity of the image texture.

4. The method for intelligently classifying wind turbine blade icing status based on image analysis according to claim 1, characterized in that: The performing of image edge detection on the second image information, calculating the ice thickness according to the edge detection result to determine the ice weight, and classifying the ice according to the ratio of the ice weight to the wind turbine blade weight to obtain image classification information includes: Using an edge detection operator to perform image edge detection on the second image information to obtain target edge information; estimating the ice thickness of the wind turbine blades by using the target edge information; Determining the ice weight of the wind turbine blade according to the ice thickness; Determining an ice coverage percentage of the fan blade based on the ice coverage weight and the total weight of the fan blade; Image grading information is determined according to the ice coverage percentage of the wind turbine blades; the image grading information includes light ice coverage, moderate ice coverage and heavy ice coverage.

5. The method for intelligently classifying the icing status of wind turbine blades based on image analysis according to claim 4, characterized in that: The light icing corresponds to an ice coverage percentage of 0%-5% on the fan blades, the moderate icing corresponds to an ice coverage percentage of 5%-10% on the fan blades, and the heavy icing corresponds to an ice coverage percentage of more than 10% on the fan blades.

6. The method for intelligently classifying wind turbine blade icing status based on image analysis according to claim 4, characterized in that: The using an edge detection operator to perform image edge detection on the second image information to obtain target edge information includes: Using the Sobel operator to calculate the gradient magnitude and direction of each pixel in the second image information, and obtaining a gradient image to characterize the strength and direction of the edge; Performing non-maximum suppression on the gradient image to refine the edge to ensure that the edge is a single pixel width and the gradient direction is consistent with the edge direction; A double threshold method is applied to set a high threshold and a low threshold, and pixels greater than the high threshold are determined as strong edges, pixels less than the low threshold are determined to be excluded, and pixels between the high threshold and the low threshold are determined to be weak edges; The hysteresis technique is used to track the edges, and the weak edges connected to the strong edges are also marked as edges, while other weak edges are deleted to obtain the target edge information to retain the complete edge information.

7. The method for intelligently classifying wind turbine blade icing status based on image analysis according to claim 4, characterized in that: The estimating the ice thickness of the wind turbine blade by using the target edge information includes: Using a camera with the same parameters to shoot a cube calibration block with a side length of L1 at a distance from the target; Use the edge detection operator to perform edge detection on the captured image of the cube calibration block to obtain the side length L2 of the cube calibration block in the image; According to the target edge information, the length between the ice-covered edge of the wind turbine blade and the trailing edge of the blade in the image is measured and recorded as L3; The actual length L4 between the ice-covered edge of the wind turbine blade and the trailing edge of the blade is calculated. The actual length L4 satisfies the expression: Based on the thickness D of the wind turbine blade, the ice thickness of the wind turbine blade is calculated. The ice thickness L h Satisfies the expression: L h =L4-D.

8. An intelligent classification system for wind turbine blade icing status based on image analysis, characterized in that: include: A data acquisition module, used to acquire initial image information when the wind turbine blade is covered with ice, wherein the initial image information includes a windward surface image and a cross-sectional image, wherein the cross-sectional image is obtained by photographing a camera at a preset target distance from the wind turbine blade; An image preprocessing module, configured to perform defogging, denoising, edge sharpening, and grayscale processing on the initial image information to obtain target image information; wherein the target image information includes first image information and second image information corresponding to the windward surface image and the cross-sectional image, respectively; A type recognition module, used to classify the first image information according to texture features to obtain image classification information; a grading module, configured to perform image edge detection on the second image information, calculate the ice thickness according to the edge detection result to determine the ice weight, and grade the ice according to the ratio of the ice weight to the wind turbine blade weight to obtain image grading information; The decision output module is used to make a decision based on the image classification information and the image grading information to determine a corresponding processing strategy to handle the icing phenomenon of the wind turbine blades.

9. An electronic device, characterized in that: include: A processor, a memory, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the method for intelligently classifying the icing status of wind turbine blades based on image analysis as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program or instruction, and when the program or instruction is executed by the processor, the method for intelligent classification of wind turbine blade icing status based on image analysis as described in any one of claims 1 to 7 is implemented.

Images