Method, device and storage medium for identifying icing thickness of fan blade

Abstract

The present invention provides a method, device and storage medium for identifying the icing thickness of a fan blade. The method includes: collecting a first image containing an iced fan blade and performing noise reduction processing on the first image to obtain a second image; extracting edge features from the second image based on the Otsu threshold segmentation algorithm to obtain a first edge contour of the iced fan blade; extracting edge features from the second image based on a coupling algorithm of the Otsu threshold segmentation algorithm and a particle swarm optimization algorithm with improved weights to obtain a second edge contour of the iced fan blade; performing integration processing on the first edge contour and the second edge contour to obtain an edge contour of the iced fan blade; and determining the icing thickness of the fan blade according to the edge contour of the iced fan blade and the edge contour of the non-iced fan blade. The present invention can obtain the accurate thickness of the icing on the surface of the fan blade in the air-cooled island.

Description

Technical Field

[0001] The present invention relates to the technical field of power transmission and distribution, and particularly to a method, device and storage medium for identifying the icing thickness of a fan blade. Background Art

[0002] The fan blade of an air-cooled island is a device for cooling high-temperature steam and has been widely used in the air-cooling system of a power plant.

[0003] Under extreme winter environmental conditions, the fan blade of an air-cooled island will be iced. When the icing thickness is too thick, it will seriously affect its normal operation and may cause safety accidents. At present, mainly by increasing the thermal resistance on the fan blade of the air-cooled island to delay the icing rate, and then regularly checking the surface of the fan blade manually according to experience to judge whether it is necessary to stop the machine for deicing.

[0004] However, only deicing by manual experience regularly cannot obtain the accurate thickness of the ice on the surface of the fan blade, and there may be situations of premature deicing or deicing too late; while using the traditional numerical simulation method cannot simulate the icing situation of the fan blade and cannot obtain the icing thickness on the surface of the fan blade. And how to obtain the accurate thickness of the ice on the surface of the fan blade of the air-cooled island has become a technical problem that needs to be solved urgently at present. Summary of the Invention

[0005] Embodiments of the present invention provide a method, device and storage medium for identifying the icing thickness of a fan blade to solve the problem that the accurate thickness of the ice on the surface of the fan blade of the air-cooled island cannot be obtained at present.

[0006] In a first aspect, embodiments of the present invention provide a method for identifying the icing thickness of a fan blade, including:

[0007] Collect a first image including an iced fan blade, and perform noise reduction processing on the first image to obtain a second image; wherein, the iced fan blade is a fan blade with ice on its surface;

[0008] Based on the maximum between-class variance threshold segmentation algorithm, extract the edge features of the second image to obtain the first edge contour of the iced fan blade; based on the coupling algorithm of the maximum between-class variance threshold segmentation algorithm and the particle swarm algorithm with improved weights, extract the edge features of the second image to obtain the second edge contour of the iced fan blade; perform integration processing on the first edge contour and the second edge contour to obtain the edge contour of the iced fan blade;

[0009] Determine the icing thickness of the fan blade according to the edge contour of the iced fan blade and the edge contour of the non-iced fan blade.

[0010] In a possible implementation manner, before performing edge feature extraction on the second image by using a coupling algorithm of the maximum inter-class variance threshold segmentation algorithm and the particle swarm optimization algorithm with improved weights to obtain the second edge contour of the ice-covered fan blade, the method further includes:

[0011] Input the second image into a pre-trained blade ice-covering prediction model to obtain the ice-covering probability of each pixel point of the second image;

[0012] Performing edge feature extraction on the second image by using a coupling algorithm of the maximum inter-class variance threshold segmentation algorithm and the particle swarm optimization algorithm with improved weights to obtain the second edge contour of the ice-covered fan blade, including:

[0013] Performing edge feature extraction on the second image by using the coupling algorithm of the ice-covering probability of each pixel point, the maximum inter-class variance threshold segmentation algorithm, and the particle swarm optimization algorithm with improved weights to obtain the second edge contour of the ice-covered fan blade.

[0014] In a possible implementation manner, performing edge feature extraction on the second image by using a coupling algorithm of the ice-covering probability of each pixel point, the maximum inter-class variance threshold segmentation algorithm, and the particle swarm optimization algorithm with improved weights to obtain the second edge contour of the ice-covered fan blade, including:

[0015] When the ice-covering probability of the target pixel point is greater than the preset threshold, set the decimal digit number of the iterative inertia weight value of the target pixel point to the first preset digit number, and perform edge feature extraction on the edge of the second image by using the maximum inter-class variance threshold segmentation algorithm and the particle swarm optimization algorithm with the improved weights of the first preset digit number to obtain the second edge contour of the ice-covered fan blade; wherein, the target pixel point is any pixel point on the edge of the second image;

[0016] When the ice-covering probability of the target pixel point is less than the preset threshold, set the decimal digit number of the iterative inertia weight value of the target pixel point to the second preset digit number, and perform edge feature extraction on the edge of the second image by using the maximum inter-class variance threshold segmentation algorithm and the particle swarm optimization algorithm with the improved weights of the second preset digit number to obtain the second edge contour of the ice-covered fan blade; wherein, the first preset digit number is greater than the second preset digit number.

[0017] In a possible implementation manner, performing integration processing on the first edge contour and the second edge contour to obtain the edge contour of the ice-covered fan blade, including:

[0018] Determine the contour of the intersection area of the area within the first edge contour and the area within the second edge contour as the edge contour of the ice-covered fan blade.

[0019] In a possible implementation, after performing coordinate correction processing on the edge contour of the ice-covered fan blade and the edge contour of the non-ice-covered fan blade respectively, the ice thickness of the fan blade is obtained, including:

[0020] Convert the edge contour of the ice-covered fan blade and the edge contour of the non-ice-covered fan blade from the pixel coordinate system to the world coordinate system;

[0021] In the world coordinate system, for each pair of pixel points corresponding to the same blade position on the edge contour of the ice-covered fan blade and the edge contour of the non-ice-covered fan blade, multiply the number of pixels between the two pixel points by the actual length of a single pixel to obtain the ice thickness of the blade position corresponding to the two pixel points.

[0022] In a possible implementation, the first image is denoised to obtain a second image, including:

[0023] Based on the non-local mean filtering method, the first image is denoised to obtain a second image; wherein, the first image is an image in the thickness direction of the fan blade.

[0024] In a possible implementation, the blade ice prediction model is a support vector machine model, and the support vector machine model is trained and tested through multiple historical ice-covered images that have been denoised; wherein, the historical ice-covered images include ice-covered images of the fan that stopped for maintenance due to excessive ice thickness.

[0025] In a possible implementation, after determining the ice thickness of the fan blade according to the edge contour of the ice-covered fan blade and the edge contour of the non-ice-covered fan blade, it further includes:

[0026] When the ice thickness of the fan blade is greater than the first preset ice thickness, turn on the heating device on the fan blade, and turn off the heating device until the ice thickness of the fan blade is equal to the second preset ice thickness.

[0027] In a second aspect, an embodiment of the present invention provides an identification device for the ice thickness of a fan blade, including:

[0028] An image processing module, configured to collect a first image including an ice-covered fan blade, and perform denoising processing on the first image to obtain a second image; wherein, the ice-covered fan blade is a fan blade with ice on its surface;

[0029] A first edge extraction module, configured to perform edge feature extraction on the second image based on the maximum between-class variance threshold segmentation algorithm to obtain the first edge contour of the ice-covered fan blade;

[0030] The second edge extraction module is used to extract the edge features of the second image based on the coupled algorithm of the maximum inter-class variance threshold segmentation algorithm and the particle swarm algorithm with improved weights, so as to obtain the second edge contour of the ice-covered fan blade;

[0031] The edge contour extraction module is used to integrally process the first edge contour and the second edge contour to obtain the edge contour of the ice-covered fan blade;

[0032] The ice thickness determination module is used to determine the ice thickness of the fan blade according to the edge contour of the ice-covered fan blade and the edge contour of the non-ice-covered fan blade.

[0033] In a possible implementation manner, the second edge extraction module is further used for:

[0034] Input the second image into a pre-trained blade ice coverage prediction model to obtain the ice coverage probability of each pixel point of the second image;

[0035] Based on the ice coverage probability of each pixel point, the maximum inter-class variance threshold segmentation algorithm, and the coupled algorithm of the particle swarm algorithm with improved weights, extract the edge features of the second image to obtain the second edge contour of the ice-covered fan blade.

[0036] In a possible implementation manner, the second edge extraction module is further used for:

[0037] When the ice coverage probability of the target pixel point is greater than the preset threshold, set the decimal digit number of the iterative inertia weight value of the target pixel point to the first preset digit number, and based on the maximum inter-class variance threshold segmentation algorithm and the particle swarm algorithm with improved weights of the first preset digit number, extract the edge features of the edge of the second image to obtain the second edge contour of the ice-covered fan blade; wherein, the target pixel point is any pixel point on the edge of the second image;

[0038] When the ice coverage probability of the target pixel point is less than the preset threshold, set the decimal digit number of the iterative inertia weight value of the target pixel point to the second preset digit number, and based on the maximum inter-class variance threshold segmentation algorithm and the particle swarm algorithm with improved weights of the second preset digit number, extract the edge features of the edge of the second image to obtain the second edge contour of the ice-covered fan blade; wherein, the first preset digit number is greater than the second preset digit number.

[0039] In a possible implementation manner, the edge contour extraction module is further used for:

[0040] Determine the contour of the intersection area of the area within the first edge contour and the area within the second edge contour as the edge contour of the ice-covered fan blade.

[0041] In a possible implementation manner, the ice thickness determination module is further used for:

[0042] Convert the edge contours of the ice-covered fan blade and the edge contours of the non-ice-covered fan blade from the pixel coordinate system to the world coordinate system;

[0043] In the world coordinate system, for every two pixel points corresponding to the same blade position on the edge contour of the ice-covered fan blade and the edge contour of the non-ice-covered fan blade, multiply the number of pixels between the two pixel points by the actual length of a single pixel to obtain the ice thickness at the blade position corresponding to the two pixel points.

[0044] In a possible implementation, the image processing module is further configured to:

[0045] Based on the non-local means filtering method, perform noise reduction processing on the first image to obtain a second image; wherein, the first image is an image in the thickness direction of the fan blade.

[0046] In a possible implementation, the blade ice coverage prediction model is a support vector machine model, and the support vector machine model is trained and tested through multiple denoised historical ice coverage images; wherein, the historical ice coverage images include ice coverage images of the fan that stopped for maintenance due to excessive ice thickness.

[0047] In a possible implementation, the determining thickness module is further configured to:

[0048] When the ice thickness of the fan blade is greater than the first preset ice thickness, turn on the heating device on the fan blade, and turn off the heating device until the ice thickness of the fan blade is equal to the second preset ice thickness.

[0049] In a third aspect, an embodiment of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the method described in the first aspect or any possible implementation of the first aspect above are implemented.

[0050] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the method described in the first aspect or any possible implementation of the first aspect above are implemented.

[0051] An embodiment of the present invention provides a method, device, and storage medium for identifying the icing thickness of a fan blade. First, a first image including an iced fan blade is collected, and the first image is denoised to obtain a second image. Then, based on the Otsu threshold segmentation algorithm, edge features of the second image are extracted to obtain a first edge contour of the iced fan blade. After that, based on a coupling algorithm of the Otsu threshold segmentation algorithm and an improved-weight particle swarm optimization algorithm, edge features of the second image are extracted to obtain a second edge contour of the iced fan blade. Next, the first edge contour and the second edge contour are integrated to obtain an edge contour of the iced fan blade. Finally, according to the edge contour of the iced fan blade and the edge contour of the non-iced fan blade, the icing thickness of the fan blade is determined.

[0052] Due to the icing on the fan blades of the air-cooled island under extreme winter environmental conditions, but currently, it is impossible to accurately identify the icing thickness on the surface of the fan blades. The present invention performs a series of processes on the collected images of the iced fan blades, so as to accurately extract the edge contour of the iced fan blade, and then finally determine the icing thickness of the fan blade. Thus, the accurate thickness of the icing on the surface of the fan blade can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0054] Figure 1 is a flowchart for implementing the method for identifying the icing thickness of a fan blade provided by an embodiment of the present invention;

[0055] Figure 2 is a schematic diagram of the edge contour in the thickness direction of the fan blade provided by an embodiment of the present invention;

[0056] Figure 3 is a schematic diagram of coordinate system conversion provided by an embodiment of the present invention;

[0057] Figure 4 is a schematic structural diagram of the device for identifying the icing thickness of a fan blade provided by an embodiment of the present invention;

[0058] Figure 5 is a schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0059] In the following description, specific details such as specific system structures and technologies are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present invention. However, those skilled in the art should clearly understand that the present invention can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present invention.

[0060] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will be described through specific embodiments in conjunction with the accompanying drawings.

[0061] As described in the background art, currently, it is impossible to accurately obtain the ice thickness on the surface of the fan blade. Only the ice thickness is determined through manual experience for deicing. Therefore, there is an urgent need for a method for identifying the ice thickness of the fan blade, which can accurately obtain the ice thickness on the surface of the fan blade, so as to prepare for subsequent timely deicing.

[0062] To solve the problems of the prior art, embodiments of the present invention provide a method, device, and storage medium for identifying the ice thickness of a fan blade. First, the method for identifying the ice thickness of a fan blade provided by the embodiments of the present invention will be introduced below.

[0063] The execution subject of the method for identifying the ice thickness of a fan blade can be a device for identifying the ice thickness of a fan blade. The device for identifying the ice thickness of a fan blade can be an electronic device with a processor and a memory, such as a millimeter-wave radar. The embodiments of the present invention do not make specific limitations.

[0064] See Figure 1 , which shows the implementation flowchart of the method for identifying the ice thickness of a fan blade provided by the embodiments of the present invention, and is described in detail as follows:

[0065] Step S110: Collect a first image containing an ice-covered fan blade, and perform noise reduction processing on the first image to obtain a second image.

[0066] Under extreme winter environmental conditions, since the surface of the fan blade is extremely prone to icing due to low temperature and high humidity, when the ice thickness is too thick, it will cause a failure and the fan cannot operate normally.

[0067] By using an identification device for the ice thickness of a fan blade to collect a first image including an ice-covered fan blade, where the ice-covered fan blade is a fan blade with ice on its surface. Since the collected first image is affected by various external factors, it is necessary to perform noise reduction processing on the collected first image to obtain a processed second image. It should be noted that the first image in the present invention is an image taken in the thickness direction of the fan blade. When ice covers the fan blade, the thickness of the fan blade will increase. By using the identification device for the ice thickness of the fan blade to take an image in the thickness direction of the fan blade, after processing, the ice thickness of the fan blade can be obtained. Of course, the user can also take an image in the width direction of the fan blade according to different application scenarios, which will not be elaborated in the present invention.

[0068] In order to maximally preserve the detail features of the first image while denoising, optionally, the non-local means filtering method can be used to perform noise reduction processing on the first image to obtain a second image. The specific noise reduction process is as follows:

[0069] Assume that the first image is v = {v(i, j)|(i, j) ∈ I}, and the value of any pixel point i in the first image after denoising is:

[0070]

[0071] where w(i, j) is the weight value in the first image that depends on pixel point j relative to pixel point i:

[0072]

[0073]

[0074] In the formula, 0 ≤ w(i, j) ≤ 1 is satisfied, Z(i) is the normalization constant, and h is the filtering parameter that controls the attenuation of the weight function. When h is small, the denoising effect on the first image is not obvious. When h is large, phenomena such as blurring and loss of edge detail information are likely to occur in the first image. For example, through actual data collection and comparison, it is found that when processing the first image of the fan blade in the air-cooled island, the best denoising effect can be achieved when h = 22.

[0075] Step S120: Based on the maximum inter-class variance threshold segmentation algorithm, perform edge feature extraction on the second image to obtain the first edge contour of the ice-covered fan blade.

[0076] In some embodiments, after performing noise reduction processing on the first image to obtain a second image, the maximum inter-class variance threshold segmentation algorithm can be used to perform edge feature extraction on the second image.

[0077] Specifically, the method for feature extraction using the Otsu thresholding algorithm is as follows:

[0078] Let f(x, y) be the gray value at the position (x, y) of the image I M×N The number of all pixels with gray level i is f i , then the probability of the occurrence of the i-th gray level is:

[0079] where i = 0, 1,..., L - 1.

[0080] The pixels in the first image are divided into two categories by the threshold t according to the gray level, namely the background C0 and the target C1. The gray levels corresponding to the background C0 are 0 to t - 1, and the gray levels corresponding to the target C1 are t to L - 1.

[0081] The probability of the occurrence of the background C0 is:

[0082]

[0083] The probability of the occurrence of the target C1 is:

[0084]

[0085] The average gray value of the background C0 is:

[0086]

[0087] The average gray value of the target C1 is:

[0088]

[0089] The total average gray value of the first image is:

[0090]

[0091] The between-class variance of the background and the target in the first image is:

[0092] δ 2 (k) = ω0(μ - μ0) 2 + ω1(μ - μ1) 2 ;

[0093] where the value of k ranges from 0 to L - 1, and the between-class variance δ 2 (k) is calculated for different values of k. When δ 2 (k) is the largest, the value of k is the optimal threshold sought.

[0094] Thus, using the obtained optimal threshold, the second image is segmented into a background and an object, and a corresponding binary image is generated. Thus, the blade image after separating the background can be obtained, and the first edge contour can be obtained, and the coordinate data of the blade boundary is also included in the image.

[0095] Step S130: Based on a coupling algorithm of the Otsu threshold segmentation algorithm and the particle swarm optimization algorithm with improved weights, edge features of the second image are extracted to obtain a second edge contour of the ice-covered fan blade.

[0096] When only the edge features of the first image are extracted by the Otsu threshold segmentation algorithm, due to the influence of the edge feature resolution, the edge features cannot be accurately extracted. Therefore, in the present invention, while the edge features are extracted by the Otsu threshold segmentation algorithm, in order to more accurately obtain the edge contour of the fan blade, the particle swarm optimization algorithm with improved weights is added. Through the coupling algorithm of these two algorithms, namely the Otsu threshold segmentation algorithm and the particle swarm optimization algorithm with improved weights, the accuracy of edge feature extraction for the first image can be improved.

[0097] Since when extracting the edge features of the first image and using the above coupling algorithm, the coupling algorithm will increase the data processing computing power, making the ice-covered thickness recognition device of the fan blade unable to calculate in time, a blade ice coverage prediction model is introduced. According to the blade ice coverage prediction model, the ice coverage probability at all positions in the first image is predicted. The recognition accuracy can be increased in the area with a high ice thickness probability and decreased in the area with a low ice thickness probability, thus maximizing the computing power.

[0098] After the blade ice coverage prediction model is trained multiple times, the detection results obtained by the blade ice coverage prediction model will be more accurate. Therefore, the accuracy of the early training is particularly important. The monitoring variables introduced in the learning are: wind direction angle, blade angle, ambient temperature, blade acceleration, blade size, etc., and different monitoring variables can also be introduced according to different scenarios. By obtaining multiple existing first images containing ice-covered fan blades, including images of ice-covered fan blades of existing fans that have stopped running due to excessive ice coverage. After obtaining the images, it is also necessary to delete the invalid images and further simplify the remaining images using the undersampling algorithm to determine the finally usable training set images.

[0099] Specifically, the blade ice coverage prediction model in the present invention is a support vector machine model. By inputting the training set images into the support vector machine model, the ice coverage probability of all pixel points can be obtained. And the blade ice coverage prediction model is tested using the test set to determine the accuracy of the blade ice coverage prediction model test.

[0100] By inputting the second image into a pre-trained ice accretion prediction model for wind turbine blades, the ice accretion probability of each pixel in the second image can be obtained. Based on the ice accretion probability of each pixel, the coupling algorithm of the Otsu threshold segmentation algorithm and the particle swarm optimization algorithm with improved weights, the edge features of the second image are extracted to obtain the second edge contour of the ice-covered wind turbine blade.

[0101] Specifically, the particle swarm optimization algorithm with improved weights can maintain good global search ability and avoid falling into local optimal solutions. Therefore, when the Otsu threshold segmentation algorithm is used to solve the optimal threshold, it can not only maintain good global search ability but also maintain local search ability, thereby improving the accuracy of edge feature extraction and obtaining a more accurate edge contour.

[0102] For example, the particle swarm optimization algorithm with improved weights can be expressed as:

[0103]

[0104] where w start is the initial value of the inertia weight, generally taken as 0.9; w end is the inertia weight value at the end of iteration, generally taken as 0.4; t is the current iteration number, t max is the maximum iteration number, and w is the iteration inertia weight value. By changing the decimal places of the iteration inertia weight value w, the recognition accuracy can be adjusted.

[0105] In the second image, the ice accretion probability of the pixels on the wind turbine blade is very high, and the ice accretion probability of the pixels far from the wind turbine blade is very low. The key of the present invention is to extract the edge contour of the ice-covered wind turbine blade. Therefore, it is necessary to perform high-precision calculation on the edge of the second image with a high ice accretion probability to accurately obtain the edge contour.

[0106] In order to more accurately identify the edge contour, the recognition accuracy of any pixel on the edge of the second image can be adjusted, thereby improving the recognition accuracy and speed of the overall edge contour. For the sake of description, the target pixel is set as any pixel on the edge of the second image here.

[0107] When the ice accretion probability of the target pixel is greater than the preset threshold, that is, the probability of ice accretion of this pixel is very high, the decimal places of the iteration inertia weight value w of this pixel can be increased. According to the user's needs, the decimal places are adjusted to the first preset number of digits, so that the extraction accuracy of this pixel can be higher. When using the coupling algorithm of the particle swarm optimization algorithm with improved weights and the Otsu threshold segmentation algorithm for this pixel, the decimal places of the iteration inertia weight value w in the particle swarm optimization algorithm with improved weights adopt the first preset number of digits, so that it can be more accurately determined whether this pixel is the target object.

[0108] When the icing probability of the target pixel is less than the preset threshold, that is, the probability of icing of this pixel is very low, the decimal places of the iterative inertia weight value w of this pixel can be adjusted to a smaller value. According to the user's requirements, the decimal places are adjusted to the second preset number of digits, so that the extraction accuracy of this pixel can be reduced and excessive computing power does not need to be consumed. When using the coupled algorithm of the improved weight particle swarm algorithm and the maximum inter-class variance threshold segmentation algorithm for this pixel, the decimal places of the iterative inertia weight value w in the improved weight particle swarm algorithm adopt the second preset number of digits, thus saving operations.

[0109] Through the above processing, the second edge contour of the icing fan blade can be obtained.

[0110] Step S140: Integrate the first edge contour and the second edge contour to obtain the edge contour of the icing fan blade.

[0111] Since the first edge contour is the edge contour extracted by the maximum inter-class variance threshold segmentation algorithm, and the second edge contour is the edge contour extracted by the coupled algorithm of the maximum inter-class variance threshold segmentation algorithm and the improved weight particle swarm algorithm, the image resolutions of the two contours are different, so the obtained edge contours are not very consistent.

[0112] In some embodiments, the contour of the intersection area of the area within the first edge contour and the area within the second edge contour can be determined as the edge contour of the icing fan blade, so as to obtain a more accurate edge contour of the icing fan blade. It can make the subsequent calculation of the icing thickness more accurate.

[0113] Step S150: Determine the icing thickness of the fan blade according to the edge contour of the icing fan blade and the edge contour of the non-icing fan blade.

[0114] Perform coordinate transformation on the obtained edge contour of the icing fan blade and the original edge contour of the non-icing fan blade to facilitate the calculation of the icing thickness. As Figure 2 shown, the extracted edge contour of the icing fan blade and the edge contour of the non-icing fan blade are both contours in the thickness direction of the fan blade. Among them, Figure 2 the length of the edge contour of the non-icing fan blade is s and the thickness is d.

[0115] Of course, the user can extract the contour in the width direction of the fan blade according to the requirements, and the present invention will not elaborate.

[0116] First, convert the edge contour of the icing fan blade and the edge contour of the non-icing fan blade from the pixel coordinate system to the world coordinate system.

[0117] Then, in the world coordinate system, for every two pixel points corresponding to the same blade position on the edge contour of the ice-covered fan blade and the edge contour of the non-ice-covered fan blade, multiply the number of pixels between the two pixel points by the actual length of a single pixel to obtain the ice thickness at the blade position corresponding to the two pixel points.

[0118] Specifically, referring to Figure 3 the coordinate system conversion relationship diagram in, the conversion relationship from pixel coordinates to world coordinates can be obtained as:

[0119]

[0120] where M1 is the internal parameter matrix of the recognition entity for the ice thickness of the fan blade, and M2 is the external parameter matrix of the recognition entity for the ice thickness of the fan blade.

[0121] In the present invention, the actual length occupied by a single pixel of the recognition entity for the ice thickness of the fan blade is 0.571 mm. Therefore, the ice thickness is obtained by calculating the product of the number of pixels between two pixel points at the same position before and after icing and 0.571.

[0122] However, it should be noted that the actual length of a single pixel of the recognition entity for the ice thickness of different fan blades is different and can be calculated according to the actual usage scenario. It is not limited in the present invention.

[0123] In the embodiment of the present invention, first, a first image including an ice-covered fan blade is collected, and the first image is denoised to obtain a second image. Then, based on the maximum inter-class variance threshold segmentation algorithm, edge features of the second image are extracted to obtain the first edge contour of the ice-covered fan blade. After that, based on the coupled algorithm of the maximum inter-class variance threshold segmentation algorithm and the particle swarm algorithm with improved weights, edge features of the second image are extracted to obtain the second edge contour of the ice-covered fan blade. Then, the first edge contour and the second edge contour are integrated to obtain the edge contour of the ice-covered fan blade. Finally, according to the edge contour of the ice-covered fan blade and the edge contour of the non-ice-covered fan blade, the ice thickness of the fan blade is determined.

[0124] Due to the extreme environmental conditions in winter, the fan blades of the air-cooled island will be covered with ice. However, currently, it is impossible to accurately identify the ice thickness on the surface of the fan blades. The present invention performs a series of processing on the collected images of the ice-covered fan blades, so that the edge contour of the ice-covered fan blades can be accurately extracted, and then the ice thickness of the fan blades can be finally determined. Thus, the accurate ice thickness on the surface of the fan blades can be obtained.

[0125] It should be understood that the sequence numbers of the steps in the above embodiments do not indicate the order of execution, and the execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0126] Based on the method for identifying the icing thickness of a wind turbine blade provided in the above embodiment, correspondingly, the present invention also provides a specific implementation manner of an identification device for the icing thickness of a wind turbine blade applied to the method for identifying the icing thickness of a wind turbine blade. Please refer to the following embodiments.

[0127] As Figure 4 shown, an identification device 400 for the icing thickness of a wind turbine blade is provided. The device includes:

[0128] An image processing module 410, configured to collect a first image including an ice-covered wind turbine blade, and perform noise reduction processing on the first image to obtain a second image; wherein, the ice-covered wind turbine blade is a wind turbine blade with ice on its surface;

[0129] A first edge extraction module 420, configured to perform edge feature extraction on the second image based on the maximum inter-class variance threshold segmentation algorithm to obtain a first edge contour of the ice-covered wind turbine blade;

[0130] A second edge extraction module 430, configured to perform edge feature extraction on the second image based on a coupling algorithm of the maximum inter-class variance threshold segmentation algorithm and an improved weight particle swarm algorithm to obtain a second edge contour of the ice-covered wind turbine blade;

[0131] An edge contour extraction module 440, configured to perform integration processing on the first edge contour and the second edge contour to obtain an edge contour of the ice-covered wind turbine blade;

[0132] A thickness determination module 450, configured to determine the icing thickness of the wind turbine blade according to the edge contour of the ice-covered wind turbine blade and the edge contour of the non-iced wind turbine blade.

[0133] In a possible implementation manner, the second edge extraction module 430 is further configured to:

[0134] Input the second image into a pre-trained blade icing prediction model to obtain the icing probability of each pixel point of the second image;

[0135] Based on the icing probability of each pixel point, the maximum inter-class variance threshold segmentation algorithm, and a coupling algorithm of an improved weight particle swarm algorithm, perform edge feature extraction on the second image to obtain a second edge contour of the ice-covered wind turbine blade.

[0136] In a possible implementation manner, the second edge extraction module 430 is further configured to:

[0137] When the icing probability of the target pixel is greater than the preset threshold, set the decimal places of the iterative inertia weight value of the target pixel to the first preset number of places, and based on the maximum inter-class variance threshold segmentation algorithm and the particle swarm optimization algorithm with the improved weight of the first preset number of places, extract the edge features of the second image to obtain the second edge contour of the icing fan blade; where the target pixel is any pixel on the edge of the second image.

[0138] When the icing probability of the target pixel is less than the preset threshold, set the decimal places of the iterative inertia weight value of the target pixel to the second preset number of places, and based on the maximum inter-class variance threshold segmentation algorithm and the particle swarm optimization algorithm with the improved weight of the second preset number of places, extract the edge features of the second image to obtain the second edge contour of the icing fan blade; where the first preset number of places is greater than the second preset number of places.

[0139] In a possible implementation, the edge contour extraction module 440 is further configured to:

[0140] Determine the contour of the intersection area of the area within the first edge contour and the area within the second edge contour as the edge contour of the icing fan blade.

[0141] In a possible implementation, the thickness determination module 450 is further configured to:

[0142] Convert the edge contour of the icing fan blade and the edge contour of the non-icing fan blade from the pixel coordinate system to the world coordinate system.

[0143] In the world coordinate system, for every two pixel points corresponding to the same blade position on the edge contour of the icing fan blade and the edge contour of the non-icing fan blade, multiply the number of pixels between the two pixel points by the actual length of a single pixel to obtain the icing thickness at the blade position corresponding to the two pixel points.

[0144] In a possible implementation, the image processing module 410 is further configured to:

[0145] Perform noise reduction processing on the first image based on the non-local means filtering method to obtain the second image.

[0146] In a possible implementation, the blade icing prediction model is a support vector machine model, and the support vector machine model is trained and tested through multiple historical icing images that have undergone noise reduction processing; after the image to be processed passes through the blade icing prediction model, the icing probabilities of all pixel points on the image to be processed can be obtained; where the historical icing images include but are not limited to icing images of fans that have stopped for maintenance due to excessive icing thickness.

[0147] In a possible implementation, the thickness determination module 450 is further configured to:

[0148] When the ice thickness on the fan blade is greater than the first preset ice thickness, turn on the heating device on the fan blade, and turn off the heating device until the ice thickness on the fan blade is equal to the second preset ice thickness.

[0149] Figure 5 is a schematic diagram of an electronic device provided by an embodiment of the present invention. As Figure 5 shown, the electronic device 5 of this embodiment includes: a processor 50, a memory 51, and a computer program 52 stored in the memory 51 and executable on the processor 50. When the processor 50 executes the computer program 52, the steps in the above-described embodiments of the method for identifying the ice thickness of each fan blade are implemented, for example Figure 1 the steps 110 to 150 shown. Alternatively, when the processor 50 executes the computer program 52, the functions of each module in the above-described device embodiments are implemented, for example Figure 4 the functions of the modules 410 to 450 shown.

[0150] Exemplarily, the computer program 52 may be divided into one or more modules, and the one or more modules are stored in the memory 51 and executed by the processor 50 to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program 52 in the electronic device 5. For example, the computer program 52 may be divided into Figure 4 the modules 410 to 450 shown.

[0151] The electronic device 5 may include, but is not limited to, a processor 50 and a memory 51. Those skilled in the art can understand that Figure 5 merely examples of the electronic device 5 do not constitute a limitation on the electronic device 5, and may include more or fewer components than shown in the figure, or combine certain components, or different components. For example, the electronic device may further include input / output devices, network access devices, a bus, etc.

[0152] The so-called processor 50 may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor, or the processor may also be any conventional processor, etc.

[0153] The memory 51 may be an internal storage unit of the electronic device 5, such as the hard disk or memory of the electronic device 5. The memory 51 may also be an external storage device of the electronic device 5, such as a plug-in hard disk equipped on the electronic device 5, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. Further, the memory 51 may also include both the internal storage unit and the external storage device of the electronic device 5. The memory 51 is used to store the computer program and other programs and data required by the electronic device. The memory 51 may also be used to temporarily store data that has been output or will be output.

[0154] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above division of each functional unit and module is used as an example. In practical applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiments can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of this application. The specific working processes of the units and modules in the above system can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated here.

[0155] In the above embodiments, the descriptions of the various embodiments have their own emphases. For parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0156] Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. A professional technician can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0157] In the embodiments provided by the present invention, it should be understood that the disclosed device / electronic device and method can be implemented in other ways. For example, the device / electronic device embodiments described above are merely illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections between each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical or other forms.

[0158] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or they can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0159] In addition, the functional units in each embodiment of the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0160] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes in the above-described embodiment methods of the present invention, it can also be completed by a computer program instructing relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-described embodiments of the method for identifying the icing thickness of each wind turbine blade can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice within the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0161] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for identifying the icing thickness of a fan blade, characterized in that, Including: Collect a first image including an ice-covered fan blade, and perform noise reduction processing on the first image to obtain a second image; wherein, the ice-covered fan blade is a fan blade with ice on its surface. Based on the Otsu threshold segmentation algorithm, extract edge features from the second image to obtain a first edge contour of the ice-covered fan blade. Input the second image into a pre-trained blade ice prediction model to obtain the ice coverage probability of each pixel point in the second image. When the ice coverage probability of a target pixel point is greater than a preset threshold, set the decimal places of the iterative inertia weight value of the target pixel point to a first preset number of digits, and based on the Otsu threshold segmentation algorithm and the particle swarm optimization algorithm with the improved weight of the first preset number of digits, extract edge features from the edge of the second image to obtain a second edge contour of the ice-covered fan blade; wherein, the target pixel point is any pixel point on the edge of the second image. When the ice coverage probability of the target pixel point is less than the preset threshold, set the decimal places of the iterative inertia weight value of the target pixel point to a second preset number of digits, and based on the Otsu threshold segmentation algorithm and the particle swarm optimization algorithm with the improved weight of the second preset number of digits, extract edge features from the edge of the second image to obtain a second edge contour of the ice-covered fan blade; wherein, the first preset number of digits is greater than the second preset number of digits. Determine the contour of the intersection area between the area within the first edge contour and the area within the second edge contour as the edge contour of the ice-covered fan blade. Convert the edge contour of the ice-covered fan blade and the edge contour of the non-ice-covered fan blade from the pixel coordinate system to the world coordinate system; in the world coordinate system, for every two pixel points corresponding to the same blade position on the edge contour of the ice-covered fan blade and the edge contour of the non-ice-covered fan blade, multiply the number of pixels between the two pixel points by the actual length of a single pixel to obtain the ice thickness at the blade position corresponding to the two pixel points.

2. The method for identifying the icing thickness of a fan blade according to claim 1, characterized in that The performing noise reduction processing on the first image to obtain a second image includes: Based on the non-local means filtering method, perform noise reduction processing on the first image to obtain a second image; wherein, the first image is an image in the thickness direction of the fan blade.

3. The method for identifying the icing thickness of a fan blade according to claim 1, characterized in that The blade ice prediction model is a support vector machine model, and the support vector machine model is trained and tested by multiple historical ice-covered images that have undergone noise reduction processing; wherein, the historical ice-covered images include ice-covered images of the fan when it stops for maintenance due to excessive ice thickness.

4. The method for identifying the icing thickness of a fan blade according to claim 1, characterized in that, After determining the ice thickness of the fan blade according to the edge contour of the ice-covered fan blade and the edge contour of the non-ice-covered fan blade, further including: When the ice thickness of the fan blade is greater than a first preset ice thickness, turn on the heating device on the fan blade, and turn off the heating device until the ice thickness of the fan blade is equal to a second preset ice thickness.

5. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 4.

6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

Images