Offshore wind power generation tower aging degree detection method and device

Abstract

The invention discloses an offshore wind power generation tower aging degree detection method and device, and the method comprises the steps: carrying out the inspection of a wind power generation tower in an offshore wind power plant according to an inspection path of the offshore wind power plant, so as to obtain a hyperspectral image of the surface of the wind power generation tower through airborne hyperspectral imaging equipment; performing image preprocessing on the hyperspectral image, and performing feature extraction on the hyperspectral image after image preprocessing to obtain hyperspectral image features of the hyperspectral image; determining the aging degree of the coating on the surface of the wind power generation tower according to the hyperspectral image features of the hyperspectral image; in this way, the aging degree of the coating on the surface of the offshore wind power generation tower can be accurately detected through the hyperspectral image of the surface of the wind power generation tower after image preprocessing on the basis of the characteristic that spectral characteristics of coatings with different aging degrees are greatly different, the whole process can be automatically achieved through the unmanned aerial vehicle, manual participation is not needed, and the detection efficiency is improved. And the detection efficiency is high.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of offshore equipment surface aging detection, and particularly relates to a method and device for detecting the aging degree of an offshore wind power tower. BACKGROUND

[0002] Since offshore wind farms are mainly distributed in the wind energy reserves of the coastal areas, the wind turbines therein are inevitably impacted and corroded by seawater, and surface coating (full name: paint coating) aging phenomena occur. If the coating aging is not timely noticed and treated, it will form a large fault after continuous accumulation, and eventually evolve into a serious accident.

[0003] Notably, the current detection of the surface coating aging degree of offshore wind turbines is generally performed by a patrol personnel operating a UAV to obtain images of the surface of the wind turbine through an onboard camera, and then manually determining the coating aging degree of the surface of the wind turbine according to the images. However, since coating aging often does not cause geometric deformation cracks and is not easy to be detected by the human eye, misjudgment often occurs. Patent CN113284102B discloses a wind turbine blade damage intelligent detection method and device based on a UAV. This patent uses a UAV to collect hyperspectral images of the wind turbine blade. In view of this, how to accurately detect the coating aging degree of the surface of offshore wind turbines has become a technical problem to be solved. SUMMARY

[0004] The purpose of the present application is to provide a method for detecting the aging degree of an offshore wind power tower, comprising the following steps:

[0005] S1, obtaining a patrol path of an offshore wind farm; along the patrol path of the offshore wind farm, a wind power tower in the offshore wind farm is patrolled, and a hyperspectral image of the surface of the wind power tower is obtained through an onboard hyperspectral imaging device;

[0006] S2, the hyperspectral image is preprocessed, and the hyperspectral image after image preprocessing is feature extracted to obtain a hyperspectral image feature;

[0007] S3, determining the coating aging degree of the surface of the wind turbine according to the hyperspectral image feature.

[0008] The present application provides a method for detecting the aging degree of an offshore wind power tower, which further comprises the following technical features:

[0009] Preferably, in S1, the onboard camera is used to shoot the offshore wind farm along the wind direction to obtain an offshore wind farm image;

[0010] The offshore wind farm image is subjected to image recognition to obtain the image size of the wind turbine in the offshore wind farm image.

[0011] Sort the wind turbines in the offshore wind farm images according to the order of image size from largest to smallest;

[0012] The positions of the wind turbines in the offshore wind farm image are connected in sequence according to the sorted order to generate an inspection path.

[0013] Preferably, the image preprocessing includes: brightness averaging processing based on the CLAHE algorithm, deblurring processing based on the GCANet algorithm, black and white correction of hyperspectral images, and multivariate scattering correction of hyperspectral images.

[0014] Preferably, the hyperspectral image features are input into a pre-trained coating aging degree detection model, and the coating aging degree detection model processes the hyperspectral image features to obtain the coating aging degree of the wind power tower.

[0015] Preferably, the coating aging degree detection model is trained by the following steps:

[0016] Obtaining a coating aging degree detection training set, wherein the coating aging degree detection training set includes multiple coating aging degree detection samples and their corresponding labels, the coating aging degree detection samples are hyperspectral images of the surface of the wind power tower, and the labels are the coating aging degrees of the surface of the wind power tower;

[0017] Perform image preprocessing on samples in the coating aging degree detection training set, and perform feature extraction on the preprocessed samples to obtain the hyperspectral image features of each sample;

[0018] The preset coating aging degree detection model is trained according to the hyperspectral image features of each sample and its corresponding label to obtain a trained coating aging degree detection model;

[0019] 100 wind turbine tower samples were tested and the difference in spectral reflectance of the aged coating in the 450-650nm band was about 20%-35%, achieving a significant differentiation effect in the difference in the detected spectrum. Aging conditions below 10% can be identified by the detection model.

[0020] Preferably, the coating aging degree detection model can be implemented by using machine learning or deep learning models, such as support vector machines (SVM) and convolutional neural networks (CNN), to identify the aging degree of the coating by extracting features from hyperspectral image data; the model includes the following steps:

[0021] Step 1: Data collection and annotation:

[0022] Sample data collection: Use drones equipped with hyperspectral imaging equipment to obtain images of coatings at different degrees of aging. The samples should cover coatings with mild, moderate, and severe aging to ensure comprehensive model training.

[0023] Data labeling: Label the coating according to its degree of aging, such as "unaged," "mildly aged," "moderately aged," and "severely aged." The labeling process can refer to manual inspection results or expert evaluation to generate a labeled data set.

[0024] The evaluation of coating degradation is usually carried out according to some international or national industry standards, which describe in detail the characteristics, classification and evaluation methods of coating degradation:

[0025] 1. GB / T 1766-2008 “Rating Method for Aging of Paints and Varnishes”

[0026] China's national standard for evaluating the degree of coating aging; it includes the following aspects:

[0027] Appearance changes: including color change, gloss change, cracking, blistering, rust, etc.;

[0028] Rating method: Determine the coating aging grade by comparing with standard pictures or on-site observation;

[0029] 2. ISO 4628-2016 Paints and varnishes – Methods for assessing degradation of coatings

[0030] This is a standard published by the International Organization for Standardization (ISO) and is widely used internationally to assess the degree of coating degradation. It includes the following parts:

[0031] ISO4628-1: Evaluation of cracking and blistering of coatings;

[0032] ISO4628-2: Evaluation of rusting of coatings;

[0033] ISO4628-3: Evaluate the chalking of coatings;

[0034] ISO4628-4: Evaluate coating shedding;

[0035] ISO 4628-5: Evaluation of stains and soiling of coatings;

[0036] 3. ASTM D 4214-17 Standard Practice for Evaluating the Weathering of Organic Coatings under Accelerated Exposure

[0037] Standards published by the American Society for Testing and Materials (ASTM) for evaluating the degradation of organic coatings under accelerated exposure conditions;

[0038] 4. ASTM D 5147-17 "Standard Practice for Rating Organic Coatings Exposed to Natural Weathering"

[0039] To evaluate the aging condition of organic coatings exposed in natural environment;

[0040] Evaluation basis and classification:

[0041] Aging area: Total area of coating aging region;

[0042] Aging number: Number of coating aging regions;

[0043] Aging size: Size of individual aging region;

[0044] Color difference: Color difference between aging region and non-aging region;

[0045] Gloss change: Change in gloss of coating due to aging;

[0046] Steps for implementing evaluation:

[0047] Image acquisition: Use hyperspectral imaging equipment or other image acquisition tools to obtain images of the coating;

[0048] Image analysis: Preprocessing and feature extraction of collected images;

[0049] Feature matching: Match extracted features with aging features described in the standard;

[0050] Grade assessment: According to the matching results and the description in the standard, evaluate the aging degree of the coating.

[0051] Reference to the above standard: According to the actual situation and test situation, this application proposes a classification of coating aging degree level:

[0052] New coating (not aged): Coating surface has no obvious wear and tear, smooth and uniform, color and initial state consistent;

[0053] Mild aging: Coating surface appears slight discoloration or gloss reduction, but no obvious peeling or cracking; Spectral reflectance has small changes compared to new coating, usually within 10%-20%;

[0054] Moderate aging: Coating surface has obvious discoloration and gloss loss, possibly with slight surface cracks or peeling; Spectral reflectance changes significantly, usually within 20%-40%;

[0055] Severe aging: Coating appears large area discoloration, cracking, peeling, coating thickness becomes thin or some areas have exposed substrate; Spectral reflectance changes greatly, more than 40%;

[0056] Evaluation basis and standard for coating aging degree:

[0057] (1) Spectral reflectance characteristics

[0058] Hyperspectral imaging is used to obtain the spectral characteristic curves of coatings with different degrees of aging. In the visible light band of 400-700nm, the degree of aging is evaluated using the following indicators:

[0059] Reflectivity difference: Coatings with different degrees of aging will have different reflectivity at specific wavelengths, such as 450nm in the blue light region and 620nm in the red light region. The more severe the aging, the higher the reflectivity.

[0060] Spectral slope: The aging condition is judged by the change in the slope of the reflectance curve. The spectral slope of the unaged coating changes gently, while the slope of the aged coating curve fluctuates more significantly.

[0061] Color change, based on the color feature changes of each pixel in the hyperspectral image, especially the changes in chromaticity and brightness:

[0062] Chromaticity change: This can be assessed using the CIELAB color model (L, a, b values). New coatings have minimal chromaticity differences, but with increasing aging, chromaticity changes increase.

[0063] Brightness decay: Coating aging is usually accompanied by brightness decay. Mild aging brightness decay is 5%-10%, moderate aging brightness decay is 10%-25%, and severe aging brightness decay is greater than 25%;

[0064] Physical defect detection:

[0065] For severely aged coatings, defects such as cracks, peeling, and flaking may occur. These defect characteristics can be detected through image processing and pattern recognition methods:

[0066] Crack detection: Count the number, length and density of surface cracks through edge detection algorithms;

[0067] Peeling and flaking area: By analyzing color differences and morphological characteristics, the area of ​​surface peeling or flaking is estimated. If the peeling rate reaches a certain percentage, it is considered severe aging;

[0068] Thickness variation:

[0069] The change in coating thickness is also an important basis for assessing aging. The coating thickness can be estimated by combining hyperspectral imaging with calculation methods:

[0070] Thickness attenuation rate: Taking the thickness of the new coating as the benchmark, the thickness change of mild aging is less than 10%, the thickness decreases by 10%-30% for moderate aging, and the thickness decreases by more than 30% for severe aging;

[0071] Step 2: Data preprocessing:

[0072] Image preprocessing: Use the CLAHE algorithm for brightness averaging, the GCANet algorithm for deblurring, black and white correction, and multivariate scattering correction to remove the effects of lighting and noise during the acquisition process;

[0073] Feature extraction: Extracting spectral features (such as wavelength-reflectance curves) or other morphological features (such as brightness changes, texture features, etc.) from hyperspectral images. These features can reflect the aging state of the coating;

[0074] Step 3: Model selection and training:

[0075] Model selection: Choose an appropriate model based on your needs. For example, support vector machines (SVMs) are suitable for hyperspectral data classification tasks. Convolutional neural networks (CNNs) can learn image features at multiple levels and are more effective for complex feature recognition.

[0076] Model training: Use labeled data for model training. The specific process is as follows:

[0077] Feature input: Input the hyperspectral image features into the model and train the model to learn the mapping relationship between aging degree and features;

[0078] Label association: For each input sample, the model adjusts its internal parameters by calculating the error so that it can correctly classify the aging degree of the sample;

[0079] Iterative optimization: Through multiple rounds of iteration and backpropagation, the model parameters are continuously optimized to improve the model's recognition accuracy for different degrees of aging;

[0080] Step 4: Model verification and evaluation:

[0081] Validation set testing: Use a portion of data that was not used in training as a validation set to test the accuracy of the model on different aging samples;

[0082] Performance evaluation: Evaluation indicators may include accuracy, recall, F1 value, false positive rate, etc. The model accuracy should be above 98%, and the false positive rate should be below 5%;

[0083] Step 5: Deployment and application:

[0084] Model deployment: Deploy the trained model to the edge computing device in the UAV system to achieve real-time aging detection;

[0085] Model update: To maintain model accuracy, new samples can be continuously collected and the dataset updated during the inspection process. The model can be retrained regularly to adapt to new environmental conditions or coating material properties.

[0086] Preferably, a coating aging assessment method is provided, wherein the coating is judged in the order of severe aging, moderate aging, and mild aging. If none of the conditions are met, the coating is judged to be unaged.

[0087] If reflectivity > 40%, brightness attenuation > 25%, peeling or crack area > 50%, thickness reduction > 30%, any of the above conditions is met, it is considered as severe aging;

[0088] 20% ≤ reflectivity ≤ 40%, 10% ≤ brightness attenuation ≤ 25%, 20% ≤ peeling or crack area ≤ 50%, 10% ≤ thickness reduction ≤ 30%. If any of the above conditions are met, it is considered moderate aging.

[0089] , no peeling, , if all the above are met, it is mild aging.

[0090] Preferably, the coating aging degree detection model includes multiple layers of convolution kernels for identifying subtle aging features;

[0091] The coating aging detection model includes an attention mechanism to enhance the discriminative ability of aging features;

[0092] The sample data requirements are as follows:

[0093] 300 wind turbine tower coating samples with different aging states were selected, including "unaged", "mildly aged", "moderately aged" and "severely aged";

[0094] Select nearshore and deep-sea sampling in different weather and seawater corrosion environments to improve the representativeness of sample data; regularly update sample data and collect new and aged samples in different environments;

[0095] The airborne hyperspectral imaging equipment obtains hyperspectral images of the surface of the wind turbine tower. Through multi-level spectral correction, the chromatic aberration interference in the image caused by distance changes is compensated to improve the accuracy of the spectral image.

[0096] Preferably, during the inspection process, the current remaining power and the recharging path from the current location to the charging point are obtained;

[0097] Calculate the flight distance based on the recharging path and obtain wind measurement data along the recharging path;

[0098] Calculate the power consumption from the current location to the charging point based on the flight distance, wind measurement data and the preset flight speed;

[0099] If the difference between the current remaining power and the power consumption is less than or equal to the preset power threshold, the inspection is suspended, the current position is recorded as the inspection interruption position, and the robot flies along the recharging path at the flying speed to the charging point for charging;

[0100] After charging is completed, fly to the inspection interruption location and continue the inspection.

[0101] Preferably, for any wind power tower in the offshore wind farm, if the aging degree of the coating on the surface of the wind power tower is greater than or equal to a preset coating aging degree threshold, the number of the corresponding wind power tower is recorded;

[0102] According to the recorded number of the wind power generation tower, a prompt message is sent to the user terminal of the maintenance personnel, prompting the maintenance personnel to perform coating maintenance on the wind power generation tower to which the number belongs.

[0103] Preferably, the spectral reflectance change value is: based on the spectral reflectance characteristics of coatings with different aging degrees, a reflectance difference of 20%-40% can be detected in the 400-700 nm band.

[0104] Preferably, preprocessing accuracy: CLAHE and GCANet algorithms are used for brightness averaging and deblurring. When the signal-to-noise ratio (SNR) reaches 40dB, the image clarity is improved by more than 25%.

[0105] Preferably, the aging detection accuracy is: through the feature extraction model, the error rate of aging degree detection is less than 5%, and can be within 3% in certain cases.

[0106] Preferably, a device for detecting damage to the surface coating of an offshore wind power tower comprises a drone, including:

[0107] An inspection module is used to inspect wind turbine towers in an offshore wind farm according to an inspection route of the offshore wind farm, so as to obtain hyperspectral images of the surfaces of the wind turbine towers through an airborne hyperspectral imaging device;

[0108] An extraction module is used to perform image preprocessing on the hyperspectral image and perform feature extraction on the hyperspectral image after image preprocessing to obtain hyperspectral image features of the hyperspectral image;

[0109] The determination module is used to determine the aging degree of the coating on the surface of the wind power tower according to the hyperspectral image characteristics of the hyperspectral image.

[0110] Preferably, an embodiment of the present disclosure provides an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method described above.

[0111] Preferably, an embodiment of the present disclosure provides a non-transitory computer-readable storage medium storing computer instructions, where the computer instructions are used to cause a computer to execute the method described above.

[0112] The beneficial effects of this application are:

[0113] 1. This application provides a method for detecting the aging degree of offshore wind power towers. Based on the fact that the spectral characteristics of coatings with different aging degrees vary greatly, the hyperspectral image of the wind power tower surface after image preprocessing is used to accurately detect the aging degree of the coating on the offshore wind power tower surface. The entire process can be automatically implemented by a drone, without manual participation, and the detection efficiency is high.

[0114] 2. Inspect wind turbine towers in offshore wind farms according to inspection routes of offshore wind farms to obtain hyperspectral images of the wind turbine tower surfaces using airborne hyperspectral imaging equipment; perform image preprocessing on the hyperspectral images and perform feature extraction on the preprocessed hyperspectral images to obtain hyperspectral image features of the hyperspectral images; and determine the degree of coating aging on the wind turbine tower surfaces based on the hyperspectral image features of the hyperspectral images;

[0115] 3. Comparative processing of coating images captured by a hyperspectral imager using the CLAHE and GCANet algorithms revealed that the detection accuracy for mildly, moderately, and severely aged samples was 95.5%, 98.2%, and 97.8%, respectively, with an average accuracy exceeding 97%.

[0116] 4. Compared with traditional manual inspection methods, the hyperspectral imaging system carried by drones reduces the inspection time of a single tower from 30 minutes for manual inspection to less than 8 minutes, improving overall inspection efficiency by approximately 75%;

[0117] After verification of 200 inspection experiments, the false detection rate of the solution of this application is only 2.3%, which is lower than the industry average false detection rate of 5%, greatly improving the reliability of detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0118] Figure 1 A flow chart showing a method for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown;

[0119] Figure 2 A structural diagram of a device for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown;

[0120] Figure 3 A structural diagram of an exemplary electronic device capable of implementing the embodiments of the present disclosure is shown;

[0121] Figure 4 A sample photo showing a method for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown;

[0122] Figure 5 A sample photo cropping diagram showing a method for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown;

[0123] Figure 6 A sample photo cropping diagram showing a method for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown;

[0124] Figure 7 A cropped photo of an aging sample of a method for detecting the aging degree of a surface coating of an offshore wind power tower provided by an embodiment of the present invention is shown;

[0125] Figure 8 A cropped photo of an aging sample of a method for detecting the aging degree of a surface coating of an offshore wind power tower provided by an embodiment of the present invention is shown;

[0126] Figure 9 A diagram showing the characteristics of aging samples of a method for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown;

[0127] Figure 10 A diagram showing the characteristics of aging samples of a method for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown;

[0128] Figure 11 A diagram showing the characteristics of aging samples of a method for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown;

[0129] Figure 12 A diagram showing the characteristics of aging samples of a method for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown;

[0130] Figure 13 A diagram showing the feature extraction of aging samples of a method for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown;

[0131] Figure 14 A diagram showing the feature extraction of aging samples of a method for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown;

[0132] Figure 15 A diagram showing the feature extraction of aging samples of a method for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown;

[0133] Figure 16A diagram showing the feature extraction of aging samples of a method for detecting the aging degree of surface coating of an offshore wind power tower provided by an embodiment disclosed in the present invention is shown. DETAILED DESCRIPTION

[0134] The following further describes the specific embodiments of the present application in conjunction with the accompanying drawings. These embodiments are only used to illustrate the present application and are not intended to limit the present invention.

[0135] In the description of the present invention, it should be noted that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer" and the like, indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings and are intended solely to facilitate and simplify the description of the present invention. They are not intended to indicate or imply that the devices or components referred to must have, be constructed, or operate in a specific orientation, and therefore should not be construed as limiting the present invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0136] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed, detachable, or integral connections; mechanical or electrical connections; direct or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0137] Furthermore, in the description of the present invention, unless otherwise specified, “plurality” means two or more.

[0138] like Figure 1-3 , a method and device for detecting the aging degree of the surface coating of an offshore wind power tower;

[0139] The drone inspects wind turbine towers in the offshore wind farm according to the inspection route of the offshore wind farm, so as to obtain hyperspectral images of the wind turbine tower surfaces using the onboard hyperspectral imaging equipment; performs image preprocessing on the hyperspectral images, and extracts features from the preprocessed hyperspectral images to obtain hyperspectral image features of the hyperspectral images; and determines the degree of coating aging on the wind turbine tower surfaces based on the hyperspectral image features of the hyperspectral images;

[0140] In this way, the degree of coating aging on the surface of offshore wind turbine towers can be accurately detected by using hyperspectral images of the wind turbine tower surface after image preprocessing, based on the fact that the spectral characteristics of coatings with different aging degrees vary greatly. The entire process can be automatically carried out by drones without the need for human intervention, and the detection efficiency is high.

[0141] Figure 1 FIG. 1 is a flow chart showing a method for detecting the aging degree of the surface coating of an offshore wind power tower provided by an embodiment of the present disclosure, as shown in FIG. Figure 1 As shown, the method 100 for detecting the aging degree of the surface coating of an offshore wind power tower can be applied to a drone and includes the following steps:

[0142] S110, inspecting the wind power towers in the offshore wind farm according to the inspection path of the offshore wind farm, so as to obtain a hyperspectral image of the surface of the wind power tower by an airborne hyperspectral imaging device, such as Figure 4 ;

[0143] The hyperspectral imaging device may be a hyperspectral imager, a hyperspectral camera, etc., which are not limited here;

[0144] It is worth noting that before S110, the inspection path can be effectively obtained through the following methods:

[0145] The offshore wind farm is photographed along the wind direction by an airborne camera to obtain an offshore wind farm image. The offshore wind farm image is then subjected to image recognition to obtain the image size of the wind turbine towers in the offshore wind farm image. The wind turbine towers in the offshore wind farm image are then sorted in descending order of image size. The positions of the wind turbine towers in the offshore wind farm image are connected in sequence according to the sorted order to generate an inspection path.

[0146] Obviously, considering that the wind turbine towers in the offshore wind farm will adjust their orientation according to the wind direction, the wind turbine towers in the offshore wind farm image obtained by shooting the offshore wind farm along the wind direction are mostly from the front, so the image size of the wind turbine tower in the offshore wind farm image can be accurately identified (the side reflects the distance between the wind turbine tower and the drone), and then the positions of the wind turbine towers in the offshore wind farm are connected in order from large to small (that is, from far to near), so as to generate the inspection path simply and quickly.

[0147] S120, performing image preprocessing on the hyperspectral image, and extracting features from the preprocessed hyperspectral image to obtain hyperspectral image features of the hyperspectral image; Figure 5 、 6 , 7, 8, from Figure 4 In the cutting, further obtain Figure 9 、10 , 11, 12, get the feature range map, and then further extract to get Figure 13 、 14 , 15, 16, i.e., feature maps;

[0148] Among them, image preprocessing can include: brightness averaging processing based on CLAHE algorithm, deblurring processing based on GCANet algorithm, restoration processing based on mathematical morphology algorithm, hyperspectral image black and white correction, hyperspectral image multivariate scattering correction;

[0149] (1) To address the problem of uneven image brightness, the CLAHE algorithm is used to segment the hyperspectral image into numerous small regions. The local histogram of each sub-region is enhanced while limiting the height of the local histogram. Interpolation is then used to optimize the transition between sub-regions. This approach not only balances the brightness but also limits the contrast, preserves the image details, and limits noise amplification, effectively improving the problem of uneven image brightness.

[0150] (2) To address the problem of blurriness in some hyperspectral images, the GCANet algorithm was used to directly learn the residual between the blurred image and the unblurred image to implement an end-to-end deblurring model, which was then used to deblur the hyperspectral images.

[0151] (3) To address the problem of partial detail loss in hyperspectral images, mathematical morphology algorithms are used to perform restoration and enhancement.

[0152] (4) The black and white correction of hyperspectral images can be expressed as follows:

[0153]

[0154] Among them, Q0 represents the original hyperspectral image, I white represents the white frame calibration image, I dark represents the black frame calibration image, and Q represents the corrected image.

[0155] (5) Multivariate scatter correction of hyperspectral images is used to eliminate spectral differences caused by different scattering levels. The specific steps include:

[0156] Assuming that the average value of all hyperspectral image spectral data is an ideal spectrum, it can be expressed as follows:

[0157]

[0158] in, Represents the average spectral vector of the spectral data of all hyperspectral images, that is, the ideal spectrum, S c Represents the spectral vector of the cth hyperspectral image, which is 1*q-dimensional data, where q represents the number of spectral wavelengths and p represents the number of spectral lines of the selected hyperspectral image.

[0159] Perform a linear regression between each spectral data of the hyperspectral image and the ideal spectrum to obtain the baseline shift b of each spectrum. c and offset a c , which is expressed as follows:

[0160]

[0161] Get the baseline translation b c and offset a c After that, the hyperspectral image spectral curve is corrected and expressed as follows:

[0162]

[0163] Among them, S c，MSC represents the spectral vector of the cth hyperspectral image after multivariate scattering correction.

[0164] Exemplarily, the feature extraction here can be implemented using a preset hyperspectral image feature extraction algorithm such as a spectral reflectance method, a principal component analysis method, a sparse expression method, etc., which is not limited here.

[0165] S130, determining an aging degree of a coating on a surface of a wind turbine tower according to hyperspectral image features of the hyperspectral image;

[0166] In this embodiment, this technical solution has the following technical features and requirements:

[0167] 1. Sample data

[0168] Data volume: 300 wind turbine tower coating samples with different aging states were selected, including "unaged," "mildly aged," "moderately aged," and "severely aged."

[0169] Collection environment: Select different weather and seawater corrosion environments (nearshore and deep sea) to collect data to ensure the representativeness of the sample data;

[0170] Data category: Based on the collection of hyperspectral images, the spectral range is the visible light region of 400-700nm;

[0171] 2. Technical evaluation indicators

[0172] Reflectivity difference: The reflectivity difference range of samples with different aging degrees (for example: mild aging is 10%-20%, moderate aging is 20%-40%, and severe aging is more than 40%);

[0173] Color brightness change: The brightness change of the aged coating is evaluated by the L value (brightness) of the CIELAB model; the brightness decreases by 5%-10% for mild aging; 10%-25% for moderate aging; and greater than 25% for severe aging;

[0174] Crack detection accuracy: The crack detection accuracy under mild aging is above 92%, moderate aging is above 96%, and severe aging is above 98%;

[0175] False detection rate: The false detection rate of aging detection achieved by the machine learning model is less than 2.5%;

[0176] 3. Technical issues to be solved by this application and implementation path:

[0177] Main technical issues:

[0178] Accuracy of aged coating detection: Coatings with different degrees of aging have small differences in spectral characteristics and are easily affected by the environment;

[0179] Reliability of real-time detection: When drones conduct inspections in a marine environment, power consumption and adverse weather conditions can affect inspection continuity.

[0180] Effectiveness of data processing: Coating aging characteristics are complex, and image preprocessing and feature extraction must be fast and effective to support real-time applications;

[0181] Implementation path: Hyperspectral imaging acquisition and preprocessing: Use the CLAHE algorithm and GCANet algorithm to perform brightness and deblurring processing to reduce external light and blur interference and enhance the clarity of spectral images;

[0182] Aging detection based on classification models: Using CNN or SVM models, extracting features after image preprocessing, inputting spectral curves into the model, and improving the model's aging recognition rate through repeated training;

[0183] Power management and route optimization: Real-time power monitoring and recharge route calculation modules are added to the inspection route to ensure the continuity of the inspection task.

[0184] Specifically, in one embodiment of the present application, the hyperspectral image features can be input into a pre-trained coating aging degree detection model, and the coating aging degree detection model processes the hyperspectral image features to accurately obtain the coating aging degree of the wind turbine tower.

[0185] Specifically, in one embodiment of the present application, the coating aging degree detection model can be trained by the following steps:

[0186] Obtain a coating aging degree detection training set; wherein the coating aging degree detection training set includes multiple coating aging degree detection samples and their corresponding labels, the coating aging degree detection samples are hyperspectral images of the surface of the wind power tower, and the labels are the coating aging degree of the surface of the wind power tower;

[0187] Perform image preprocessing on samples in the coating aging degree detection training set, and perform feature extraction on the preprocessed samples to obtain the hyperspectral image features of each sample. Then, based on the hyperspectral image features of each sample and its corresponding labels, a preset coating aging degree detection model, such as SVM and convolutional neural network, is trained to obtain a trained coating aging degree detection model.

[0188] Based on a comparative experiment of 100 sample data, the spectral curve characteristics of unaged and aged coatings showed clear differentiation. After analysis by the coating aging degree detection model, the recognition rate reached 96%, and the false alarm rate was controlled within 2%, verifying the effectiveness and stability of the detection model.

[0189] It is understandable that the image preprocessing and feature extraction here are consistent with the image preprocessing and feature extraction details in S120, and for the sake of brevity, they are not described here in detail.

[0190] Specifically, in one embodiment of the present application, based on the fact that the spectral characteristics of coatings with different aging degrees vary greatly, the aging degree of the coating on the surface of an offshore wind power tower can be accurately detected by using a hyperspectral image of the wind power tower surface after image preprocessing. The entire process can be automatically implemented by a drone without the need for human intervention, and the detection efficiency is high.

[0191] Specifically, in one embodiment of the present application, considering that the offshore wind farm area is large, the drone may need to be charged midway during the inspection process. To further improve automation, the offshore wind power tower surface coating aging degree detection method 100 may further include:

[0192] During the inspection process, the current remaining power and the recharging path from the current location to the charging point are obtained. The flight distance is calculated based on the recharging path, and wind measurement data along the recharging path, such as wind speed, wind direction, and vertical airflow, are obtained.

[0193] Calculate the power consumption from the current location to the charging point based on the flight distance, wind measurement data, and the preset flight speed, and calculate the difference between the current remaining power and the power consumption;

[0194] If the difference between the current remaining power and the power consumption is less than or equal to the preset power threshold, the inspection is suspended, the current position is recorded as the inspection interruption position, and the robot flies along the recharging path at the flying speed to the charging point for charging;

[0195] After the charging is completed, fly to the inspection interruption position, continue to inspect.

[0196] Specifically, in one embodiment of the present application, in order to improve efficiency, the offshore wind power tower surface coating aging degree detection method 100 can further include:

[0197] When the inspection interruption position is recorded, the standby unmanned aerial vehicle is sent the inspection path and the inspection interruption position, so that it replaces itself to start inspection from the inspection interruption position, and after the charging is completed, fly to the position of the standby unmanned aerial vehicle, and continue to inspect.

[0198] Specifically, in one embodiment of the present application, in order to timely repair the wind power tower with high coating aging degree, the offshore wind power tower surface coating aging degree detection method 100 can further include:

[0199] For any wind power tower in the offshore wind farm, if the coating aging degree of the wind power tower surface is greater than or equal to the preset coating aging degree threshold, the number of the corresponding wind power tower is recorded;

[0200] According to the recorded number of the wind power tower, a prompt information is sent to the user terminal of the maintenance personnel, prompting the maintenance personnel to timely maintain the coating of the wind power tower to which the number belongs.

[0201] Specifically, in one embodiment of the present application, Figure 2 The structure diagram of the offshore wind power tower surface coating aging degree detection device provided by the embodiment of the present application is shown, as Figure 2 As shown in the figure, the offshore wind power tower surface coating aging degree detection device 200 can be applied to an unmanned aerial vehicle, and includes:

[0202] The inspection module 210 is configured to inspect the wind power towers in the offshore wind farm according to the inspection path of the offshore wind farm, so as to obtain the hyperspectral image of the surface of the wind power tower by the airborne hyperspectral imaging device;

[0203] The extraction module 220 is configured to perform image preprocessing on the hyperspectral image, and perform feature extraction on the hyperspectral image after image preprocessing, to obtain the hyperspectral image feature of the hyperspectral image.

[0204] The determination module 230 is configured to determine the coating aging degree of the surface of the wind power tower according to the hyperspectral image feature of the hyperspectral image.

[0205] Specifically, in one embodiment of the present application, the offshore wind power tower surface coating aging degree detection device 200 further includes:

[0206] The shooting module is configured to shoot the offshore wind farm along a wind direction by using an airborne camera before the wind power towers in the offshore wind farm are inspected according to the inspection path of the offshore wind farm, so as to obtain an offshore wind farm image.

[0207] The identification module is configured to perform image recognition on the offshore wind farm image, so as to obtain an image size of the wind power tower in the offshore wind farm image.

[0208] The sorting module is configured to sort the wind power towers in the offshore wind farm image according to an order from large to small of the image sizes.

[0209] The generation module is configured to sequentially connect positions of the wind power towers in the offshore wind farm image in the offshore wind farm according to the sorted order, so as to generate the inspection path.

[0210] Specifically, in one embodiment of the present application, the determination module 230 is specifically configured to:

[0211] input the hyperspectral image feature into a pre-trained coating aging degree detection model, process the hyperspectral image feature by using the coating aging degree detection model, and obtain the coating aging degree of the wind power tower.

[0212] In some embodiments, the offshore wind power tower surface coating aging degree detection device 200 further comprises:

[0213] The acquisition module is configured to acquire a current remaining power and a return charging path from a current position to a charging point during the inspection process.

[0214] The calculation module is configured to calculate a flight distance according to the return charging path, and acquire wind measurement data on the return charging path.

[0215] The calculation module is further configured to calculate a power consumption from the current position to the charging point according to the flight distance, the wind measurement data, and a preset flight speed.

[0216] The return charging module is configured to, if a difference between the current remaining power and the power consumption is less than or equal to a preset power threshold, pause the inspection, record the current position as an inspection interruption position, fly to the charging point at the flight speed along the return charging path to charge.

[0217] The continuation module is configured to, after the charging is completed, fly to the inspection interruption position to continue the inspection.

[0218] Specifically, in one embodiment of the present application, the offshore wind power tower surface coating aging degree detection device 200 further comprises:

[0219] The recording module is configured to record the number of the wind power generation tower if the aging degree of the coating on the surface of the wind power generation tower is greater than or equal to the preset coating aging degree threshold.

[0220] The prompting module is configured to send prompt information to a user terminal of the maintenance personnel, prompting the maintenance personnel to perform coating maintenance on the wind power generation tower to which the number belongs, according to the recorded number of the wind power generation tower.

[0221] Specifically, in one embodiment of the present application, Figure 2 Each module / unit in the offshore wind power generation tower surface coating aging degree detection device 200 shown has the function of implementing Figure 1 Each step in the offshore wind power generation tower surface coating aging degree detection method 100 shown has the function of implementing

[0222] Specifically, in one embodiment of the present application, Figure 3 A structural diagram of an exemplary electronic device capable of implementing an embodiment of the present disclosure is shown; the electronic device 300 is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers; the electronic device 300 can also represent various forms of mobile devices, such as personal digital processors, cellular telephones, smart phones, wearable devices, and other similar computing devices; the components shown herein, their connections, and relationships, and their functions, are merely examples and are not intended to limit the implementations described and / or claimed in this document;

[0223] As shown in the figure, Figure 3 The electronic device 300 can include a computing unit 301, which can perform various appropriate actions and processes according to computer programs stored in a read-only memory (ROM) 302 or computer programs loaded into a random access memory (RAM) 303 from a storage unit 308; various programs and data required for the operation of the electronic device 300 can also be stored in the RAM 303; the computing unit 301, the ROM 302, and the RAM 303 are connected to each other through a bus 304; an input / output (I / O) interface 305 is also connected to the bus 304;

[0224] The plurality of components in the electronic device 300 are connected to the I / O interface 305, including: an input unit 306, such as a keyboard, a mouse, etc.; an output unit 307, such as various types of displays, speakers, etc.; a storage unit 308, such as a magnetic disk, an optical disk, etc.; and a communication unit 309, such as a network card, a modem, a wireless communication transceiver, etc.; the communication unit 309 allows the electronic device 300 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunication networks;

[0225] The computing unit 301 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities; some examples of the computing unit 301 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc.; the computing unit 301 performs various methods and processes described above, such as the method 100; for example, in some embodiments, the method 100 can be implemented as a computer program product, including a computer program tangibly embodied in a computer-readable medium, such as the storage unit 308; in some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 300 via the ROM 302 and / or the communication unit 309; when the computer program is loaded into the RAM 303 and executed by the computing unit 301, one or more steps of the method 100 described above can be performed.

[0226] In particular, in one embodiment of the present application, the computing unit 301 can be configured to perform the method 100 by any other appropriate means, such as by means of firmware.

[0227] The various embodiments described above herein can be implemented in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a programmable logic device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0228] Program code for carrying out methods of the present disclosure can be written in any combination of one or more programming languages. The program code can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the program code, when executed by the processor or controller, produces a means for implementing the functions / operations specified in the flowchart diagrams and / or block diagrams. The program code can be executed entirely on a machine, partially on a machine, partially on a machine as a stand-alone software package, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0229] In the context of the present disclosure, a computer-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the computer-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0230] In particular, in one embodiment of the present application, the present disclosure further provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to make a computer execute the method 100, and achieve the corresponding technical effects achieved by the embodiments of the present disclosure in executing the method. For brevity, the description will not be repeated here.

[0231] In particular, in one embodiment of the present application, the present disclosure further provides a computer program product, which includes a computer program, and the computer program, when executed by a processor, implements the method 100.

[0232] To provide for interaction with a user, the above described embodiments can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0233] The above described embodiments can be implemented in a computing system that includes a back-end component, e.g., as a data server, or that includes a middleware component, e.g., an application server, or that includes a front-end component, e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here, or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication, e.g., a communication network. Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0234] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0235] Overall, a kind of offshore wind power tower surface coating aging degree detection method and device are applied to wind power generation technical field;The method is applied to unmanned aerial vehicle, comprising: according to the inspection path of offshore wind farm, the wind power tower in offshore wind farm is inspected, so as to obtain the hyperspectral image of wind power tower surface by airborne hyperspectral imaging equipment;Image pre-processing is carried out on the hyperspectral image, and the hyperspectral image features of the hyperspectral image are obtained by feature extraction on the hyperspectral image after image pre-processing;The coating aging degree of wind power tower surface is determined according to the hyperspectral image features of the hyperspectral image;In this way, the coating aging degree of offshore wind power tower surface can be accurately detected based on the characteristics of the large spectral characteristics difference of different aging degree coatings through the hyperspectral image of wind power tower surface after image pre-processing, and the whole process can be automatically realized by unmanned aerial vehicle without manual intervention, and the detection efficiency is higher.

[0236] The above only describes the preferred embodiments of the present application, and it should be pointed out that for ordinary skilled in the art, without departing from the technical principles of the present application, a number of improvements and substitutions can be made, and these improvements and substitutions should also be considered as the protection scope of the present application.

Claims

1. A method for detecting the aging degree of an offshore wind power tower, characterized in that: The following steps are involved: S1, obtaining an inspection route of an offshore wind farm; inspecting wind turbine towers in the offshore wind farm along the inspection route, and obtaining hyperspectral images of the wind turbine tower surfaces using an airborne hyperspectral imaging device; S2, performing image preprocessing on the hyperspectral image, and extracting features from the preprocessed hyperspectral image to obtain hyperspectral image features; S3, determining the coating aging degree on the surface of the wind turbine based on the hyperspectral image features.

2. The method for detecting the aging degree of an offshore wind power tower according to claim 1, wherein: In S1, obtaining an inspection route for an offshore wind farm includes the following steps: Acquiring images of offshore wind farms: Using an onboard camera to photograph offshore wind farms along the wind direction, collecting image data including wind turbine towers; Image recognition: Perform image recognition processing on offshore wind farm images to extract and identify the image size information of wind turbine towers in the images; Size sorting: Based on the extracted wind turbine tower image size information, the wind turbine towers are sorted from large to small according to image size; Generate an inspection route: Connect the location points of each wind turbine tower in the offshore wind farm in sequence according to the sorted order to form an inspection route that adapts to the actual layout of the wind farm.

3. The method for detecting the aging degree of an offshore wind power tower according to claim 1, wherein: Image preprocessing includes: brightness averaging based on the CLAHE algorithm, deblurring based on the GCANet algorithm, black and white correction of hyperspectral images, and multivariate scattering correction of hyperspectral images; The hyperspectral image features are input into a pre-trained coating aging degree detection model, which processes the hyperspectral image features to obtain the coating aging degree of the wind turbine tower.

4. The method for detecting the aging degree of an offshore wind power tower according to claim 1, wherein: The coating aging degree detection model is trained through the following steps: Obtaining a coating aging degree detection training set, wherein the coating aging degree detection training set includes multiple coating aging degree detection samples and their corresponding labels, the coating aging degree detection samples are hyperspectral images of the surface of the wind power tower, and the labels are the coating aging degrees of the surface of the wind power tower; Perform image preprocessing on samples in the coating aging degree detection training set, and perform feature extraction on the preprocessed samples to obtain the hyperspectral image features of each sample; The preset coating aging degree detection model is trained according to the hyperspectral image features of each sample and its corresponding label to obtain a trained coating aging degree detection model.

5. The method for detecting the aging degree of an offshore wind power tower according to claim 4, wherein: The coating aging degree detection model uses machine learning or deep learning models to identify the aging degree of the coating by extracting features from hyperspectral image data; it includes the following steps: Step 1: Data collection and annotation: Sample data collection: Use drones equipped with hyperspectral imaging equipment to obtain images of coatings at different degrees of aging. The samples should cover coatings with mild, moderate, and severe aging to ensure comprehensive model training. Data labeling: Label according to the degree of coating aging, and classify and label according to "no aging", "mild aging", "moderate aging", and "severe aging"; Step 2: Data preprocessing: Image preprocessing: Use the CLAHE algorithm for brightness averaging, the GCANet algorithm for deblurring, black and white correction, and multivariate scattering correction to remove the effects of lighting and noise during the acquisition process; Feature extraction: Extract spectral features or morphological features from hyperspectral images to reflect the aging state of the coating; Step 3: Model selection and training: Model selection: Choose the appropriate model based on your needs; support vector machines are used for hyperspectral data classification tasks; convolutional neural networks learn image features at multiple levels; Model training, use labeled data for model training: Feature input: Input the hyperspectral image features into the model and train the model to learn the mapping relationship between aging degree and features; Label association: For each input sample, the model adjusts its internal parameters by calculating the error so that it can correctly classify the aging degree of the sample; Iterative optimization: Through multiple rounds of iteration and backpropagation, the model parameters are continuously optimized to improve the model's recognition accuracy for different degrees of aging; Step 4: Model verification and evaluation: Validation set testing: Use a portion of data that was not used in training as a validation set to test the accuracy of the model on different aging samples; Performance evaluation: Evaluation indicators include accuracy, recall, F1 value, and false positive rate; the model accuracy reaches more than 98% and the false positive rate is less than 5%.

6. A method for detecting the aging degree of an offshore wind power tower according to claim 5, characterized in that: Including the evaluation method of coating aging, and judging in the order of severe aging, moderate aging, and mild aging. If none of them meet the requirements, it is judged as unaged coating; If reflectivity > 40%, brightness attenuation > 25%, peeling or crack area > 50%, thickness reduction > 30%, any of the above conditions is met, it is considered as severe aging; 20% ≤ reflectivity ≤ 40%, 10% ≤ brightness attenuation ≤ 25%, 20% ≤ peeling or crack area ≤ 50%, 10% ≤ thickness reduction ≤ 30%. If any of the above conditions are met, it is considered moderate aging. 10% ≤ reflectivity 5% ≤ brightness attenuation No peeling, thickness reduction If all the conditions are met, it is mild aging.

7. The method for detecting the aging degree of an offshore wind power tower according to claim 5, wherein: The coating aging degree detection model includes multiple layers of convolution kernels to identify subtle aging features; The coating aging detection model includes an attention mechanism to enhance the discriminative ability of aging features; The sample data requirements are as follows: 300 wind turbine tower coating samples with different aging states were selected, including "unaged", "mildly aged", "moderately aged" and "severely aged"; Select nearshore and deep-sea samples in different weather and seawater corrosion environments to improve the representativeness of sample data; regularly update sample data and collect new and aged samples in different environments; Airborne hyperspectral imaging equipment obtains hyperspectral images of the surface of wind turbine towers and uses multi-level spectral correction to compensate for chromatic aberration interference in the image caused by distance changes, thereby improving the accuracy of the spectral image.

8. The method for detecting the aging degree of an offshore wind power tower according to claim 1, wherein: During the inspection process, obtain the current remaining power and the recharging path from the current location to the charging point; Calculate the flight distance based on the recharging path and obtain wind measurement data along the recharging path; Calculate the power consumption from the current location to the charging point based on the flight distance, wind measurement data and the preset flight speed; If the difference between the current remaining power and the power consumption is less than or equal to the preset power threshold, the inspection is suspended, the current position is recorded as the inspection interruption position, and the robot flies along the recharging path at the flying speed to the charging point for charging; After charging is completed, fly to the inspection interruption location and continue the inspection.

9. The method for detecting the aging degree of an offshore wind power tower according to claim 1, wherein: If the coating aging degree on the surface of the wind turbine tower is greater than or equal to the preset coating aging degree threshold, the number of the corresponding wind turbine tower is recorded; According to the recorded number of the wind power generation tower, a prompt message is sent to the user terminal of the maintenance personnel, prompting the maintenance personnel to perform coating maintenance on the wind power generation tower to which the number belongs.

10. A device for detecting damage to the surface coating of an offshore wind power tower, using the method for detecting aging of an offshore wind power tower according to any one of claims 1 to 9, comprising a drone, characterized in that: An acquisition module is used to obtain the inspection path of the offshore wind farm; An inspection module is used to inspect wind turbine towers in an offshore wind farm according to an inspection route of the offshore wind farm, so as to obtain hyperspectral images of the surfaces of the wind turbine towers through an airborne hyperspectral imaging device; An extraction module is used to perform image preprocessing on the hyperspectral image and perform feature extraction on the hyperspectral image after image preprocessing to obtain hyperspectral image features of the hyperspectral image; The determination module is used to determine the aging degree of the coating on the surface of the wind power tower according to the hyperspectral image characteristics of the hyperspectral image.

Images