Fan tower drum damage evaluation method and system comprehensively considering multiple indexes

Through the combination of convolutional neural network and finite element analysis, the mechanical performance parameters of the fan tower are obtained and the tower damage index is calculated, which solves the problem of inaccurate evaluation results in the existing technology, and achieves the accuracy of the damage to the fan tower. Assessment reduces detection costs and risks.

CN119066918APending Publication Date: 2024-12-03SHANDONG ELECTRIC POWER ENG CONSULTING INST CORP
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411130672.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

The existing fan tower damage diagnosis method only considers image features and fails to effectively combine the internal mechanical properties of the tower, resulting in inaccurate evaluation results and lacks an evaluation scheme that comprehensively considers mechanical properties and crack characteristics.

Method used

Convolutional neural network combined with finite element analysis is used to obtain the crack size and morphological parameters in the tower image, and the mechanical performance parameters are calculated through the finite element model to determine the tower structural integrity index, damage comprehensive coefficient and safety coefficient, and conduct comprehensive evaluation.

Benefits of technology

The accuracy of the damage evaluation of fan tower is improved, the relationship between the internal mechanical properties of the tower and the crack characteristics is taken into account, the detection cost and risk are reduced, and the safe and stable operation of the wind farm is ensured.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119066918A_ABST
    Figure CN119066918A_ABST
Patent Text Reader

Abstract

The invention belongs to the technical field of new energy equipment detection and maintenance, and provides a fan tower drum damage evaluation method and system comprehensively considering multiple indexes, and the method comprises the steps: obtaining the mechanical performance parameters of a fan tower drum through a finite element model on the basis of obtaining the size and morphological parameters of a crack through image segmentation; according to the obtained size and morphological parameters and mechanical property parameters of the crack, determining three indexes of a tower drum structure integrity index, a tower drum damage comprehensive coefficient and a tower drum safety coefficient; carrying out weighted summation on three indexes of the tower drum structure integrity index, the tower drum damage comprehensive coefficient and the tower drum safety coefficient to obtain a comprehensive evaluation index; and finally, performing fan tower drum damage evaluation by using the comprehensive evaluation index. When the damage of the tower drum of the fan is evaluated, the relationship between the characteristics of the tower drum, cracks and the like in the image and the internal mechanical property of the tower drum is considered, and a plurality of indexes are comprehensively considered, so that the accuracy of evaluating the damage condition of the tower drum is ensured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of new energy equipment detection and maintenance, and in particular relates to a wind turbine tower damage evaluation method and system that comprehensively considers multiple indicators. Background Art

[0002] In the context of the current global energy structure transformation, wind power generation, as a clean and renewable energy form, has been widely used and developed rapidly. As the key structure supporting the wind turbine generator set, the stability and safety of the wind turbine tower are directly related to the normal operation of the entire wind turbine generator set.

[0003] The inventors have found that in the current intelligent diagnosis method for wind turbine tower damage, after obtaining the tower image, the image is usually processed and recognized with the help of a trained neural network model to obtain an evaluation result of the tower damage; however, in the above method, only the situation of the image itself is considered, such as characteristic factors such as the area and depth of the crack, and the relationship between the features in the image and the mechanical properties inside the tower itself is not considered. There is a large difference between the considered factors and the actual internal stress conditions of the tower, and the evaluation indicators are relatively single, resulting in inaccurate evaluation results of the tower damage; and because it is difficult and costly to implement on-site detection of mechanical performance parameters by sensors, there is currently no wind turbine tower damage evaluation scheme that obtains mechanical performance parameters and comprehensively considers the relationship between mechanical performance and crack characteristics. Summary of the invention

[0004] In order to solve the above problems, the present invention proposes a wind turbine tower damage evaluation method and system that comprehensively considers multiple indicators. The present invention proposes a wind turbine tower damage evaluation method and system that comprehensively considers multiple indicators, considers the relationship between features such as the middle tower and cracks and the internal mechanical properties of the tower itself, and integrates multiple indicators such as the tower structure integrity index, the tower damage comprehensive coefficient and the tower safety factor, thereby ensuring the accuracy of the evaluation of the tower damage.

[0005] In order to achieve the above object, the present invention is implemented through the following technical solutions:

[0006] In a first aspect, the present invention provides a wind turbine tower damage evaluation method that comprehensively considers multiple indicators, including:

[0007] Acquire wind turbine tower images;

[0008] According to the acquired wind turbine tower image and the preset convolutional neural network model, the size and morphological parameters of the cracks in the wind turbine tower are obtained;

[0009] Finite element analysis is performed based on the obtained crack size and morphological parameters and the pre-established finite element model to obtain the mechanical performance parameters of the wind turbine tower;

[0010] According to the obtained crack size and morphological parameters, as well as mechanical performance parameters, three indicators, namely, tower structure integrity index, tower damage comprehensive coefficient and tower safety factor, are determined; the three indicators, namely, tower structure integrity index, tower damage comprehensive coefficient and tower safety factor, are weighted and summed to obtain a comprehensive evaluation index;

[0011] The comprehensive evaluation index is used to evaluate wind turbine tower damage.

[0012] Furthermore, the contextual information of the wind tower image is extracted by contracting the path, and then the details and spatial resolution of the wind tower image are restored by expanding the path. The fully connected layer in the traditional convolutional neural network is replaced by the fully convolutional layer to achieve end-to-end segmentation of wind tower images of any size. The atrous convolution is combined to expand the receptive field and obtain semantic information, and the encoder-decoder structure is used to refine the segmentation boundary. After the wind tower image is segmented, the size of the crack is obtained by proportional calculation with the help of the known tower size.

[0013] Furthermore, when acquiring the image of the wind turbine tower, the tower is divided into multiple layers according to a preset distance, and a front image of the tower is collected around each layer, and an image of the tower at a preset angle is collected on each layer.

[0014] Furthermore, an attention module is added to the convolutional neural network to focus on important areas related to towers and cracks in the image and suppress interference from irrelevant background. Convolution kernels of different sizes are used to extract multi-scale feature information and fuse them to capture the characteristics of towers and cracks at different scales. Depthwise separable convolution is used instead of traditional convolution to reduce the number of parameters and computation.

[0015] Furthermore, during the finite element analysis, a denser grid is used in the crack tip area, and singular elements are used to simulate stress singularity phenomena; boundary constraints and loads are applied according to the actual installation conditions and operating conditions, including wind loads, deadweight, and temperature changes.

[0016] Furthermore, the tower structural integrity index = (1-(total crack area / tower surface area)) × (1-maximum equivalent plastic strain) × (1-maximum stress intensity factor); tower damage comprehensive coefficient = (1-average crack width / tower wall thickness) × (1-average cracking energy release rate / reference energy release rate) × (1-average damage variable); tower safety factor = (original strength / residual strength) × (1-total crack length / tower circumference) × (1-maximum stress concentration factor).

[0017] In a second aspect, the present invention further provides a wind turbine tower damage evaluation system that comprehensively considers multiple indicators, including:

[0018] The image acquisition module is configured to: acquire an image of a wind turbine tower;

[0019] The image segmentation module is configured to: obtain the size and morphological parameters of the cracks in the wind turbine tower according to the acquired wind turbine tower image and a preset convolutional neural network model;

[0020] The finite element analysis module is configured to: perform finite element analysis according to the obtained size and shape parameters of the crack and the pre-established finite element model to obtain the mechanical performance parameters of the wind turbine tower;

[0021] The evaluation index determination module is configured to: determine the three indicators of tower structure integrity index, tower damage comprehensive coefficient and tower safety factor according to the obtained crack size and morphological parameters and mechanical performance parameters; perform weighted summation on the three indicators of tower structure integrity index, tower damage comprehensive coefficient and tower safety factor to obtain a comprehensive evaluation index;

[0022] The evaluation module is configured to: use the comprehensive evaluation index to evaluate the damage of the wind turbine tower.

[0023] In a third aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the wind turbine tower damage evaluation method that comprehensively considers multiple indicators as described in the first aspect.

[0024] In a fourth aspect, the present invention further provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein when the processor executes the program, the steps of the wind turbine tower damage assessment method that comprehensively considers multiple indicators as described in the first aspect are implemented.

[0025] In a fifth aspect, the present invention further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps of the wind turbine tower damage evaluation method that comprehensively considers multiple indicators as described in the first aspect are implemented.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] The present invention obtains the size and morphological parameters of cracks through image segmentation, and obtains the mechanical performance parameters of the wind turbine tower by using a finite element model; determines three indicators, namely, a tower structure integrity index, a tower damage comprehensive coefficient and a tower safety factor, according to the obtained crack size and morphological parameters and mechanical performance parameters; performs weighted summation on the three indicators, namely, the tower structure integrity index, the tower damage comprehensive coefficient and the tower safety factor, to obtain a comprehensive evaluation index; finally, uses the comprehensive evaluation index to evaluate the wind turbine tower damage; when evaluating the wind turbine tower damage, uses a finite element model to obtain the mechanical performance parameters, and considers the relationship between features such as the middle tower and cracks and the internal mechanical properties of the tower itself without increasing the difficulty and cost of on-site implementation, and combines multiple indicators, namely, the tower structure integrity index, the tower damage comprehensive coefficient and the tower safety factor, to ensure the accuracy of the evaluation of the tower damage. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The drawings in the specification that constitute a part of this embodiment are used to provide a further understanding of this embodiment. The schematic embodiments of this embodiment and their descriptions are used to explain this embodiment and do not constitute improper limitations on this embodiment.

[0029] Figure 1 This is a flow chart of Embodiment 1 of the present invention;

[0030] Figure 2 This is a schematic diagram of crack identification in Example 1 of the present invention;

[0031] Figure 3 This is a schematic diagram of the finite element analysis of tower cracks in Example 1 of the present invention. DETAILED DESCRIPTION

[0032] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0033] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0034] Embodiment 1:

[0035] In the prior art, the diagnostic methods for wind turbine tower damage mainly include manual inspection and some traditional detection methods. Manual inspection usually requires inspectors to climb up the tower and conduct inspections through naked eye observation and simple tools. This method is not only inefficient and labor-intensive, but also has great safety risks. At the same time, the results of manual inspections largely depend on the experience and skills of the inspectors, which are highly subjective and difficult to ensure the accuracy and consistency of the inspections. In addition, some traditional inspection technologies, such as ultrasonic inspection and magnetic particle inspection, can detect tower damage to a certain extent, but these methods often require multiple inspection points to be set on the tower surface, which is complicated to operate and takes a long time to inspect a large area of ​​the tower surface. With the development of science and technology, some image-based detection methods have gradually been applied to wind turbine tower damage diagnosis. However, these methods still have many problems. For example, in terms of image acquisition, commonly used fixed cameras or ground shooting equipment often cannot fully and clearly obtain the overall image of the tower, especially for the upper part of the tower with a high height and difficult-to-access parts, which are prone to shooting blind spots. In the image analysis and processing stage, the existing image segmentation algorithm is not accurate enough, and it is difficult to accurately separate the tower and cracks from the complex background. This leads to large errors in the subsequent measurement of crack size and morphology, affecting the accuracy of damage diagnosis. When establishing a finite element model for analysis, the model used in the prior art is not accurate enough to fully consider the actual working environment and stress conditions of the tower, resulting in deviations between the analysis results and the actual situation. In addition, with regard to the definition of damage indicators and the establishment of corresponding relationships, the existing methods lack scientificity and systematicity, and cannot effectively and accurately associate the morphological parameters of the cracks with the degree of damage to the tower, thereby failing to provide an accurate and reliable basis for the maintenance and repair of the tower.

[0036] In summary, the existing wind turbine tower damage diagnosis technology has many shortcomings and cannot meet the needs of rapid and accurate diagnosis of tower damage in practical applications. In addition, in the current intelligent diagnosis method for wind turbine tower damage, after obtaining the tower image, the image is usually processed and recognized with the help of a trained neural network model to obtain the evaluation result of the tower damage; however, in the above method, only the image itself is considered, such as the area and depth of the crack and other characteristic factors, but the relationship between the features in the image and the mechanical properties of the tower itself is not considered. The factors considered are quite different from the internal stress conditions of the actual tower, and the evaluation index is relatively single, resulting in inaccurate evaluation results of the tower damage.

[0037] In order to solve at least one of the above problems, this embodiment provides a wind turbine tower damage evaluation method that comprehensively considers multiple indicators, the purpose of which is to obtain a full range of images of the wind turbine tower by using a camera mounted on an unmanned aerial vehicle, overcome the limitations and blind spots of existing image acquisition methods, and achieve full coverage of the tower. An advanced image segmentation algorithm based on a convolutional neural network is used to improve the accuracy and precision of tower and crack segmentation, providing a reliable basis for subsequent measurement and analysis. With the help of accurate tower size information, the actual size of the crack is accurately obtained to reduce measurement errors. A more accurate and practical finite element model is established using ABAQUS software to conduct a comprehensive and in-depth parametric analysis of the cracking of the tower. A scientific and reasonable damage index is customized to establish an accurate correspondence between the crack morphology parameters and the tower damage index, so that the degree of damage of the tower can be accurately and quickly evaluated, providing strong technical support and decision-making basis for the maintenance and repair of the wind turbine tower. This embodiment aims to improve the efficiency and accuracy of wind turbine tower damage diagnosis, reduce detection costs and risks, ensure the safe and stable operation of wind farms, extend the service life of wind turbine towers, and promote the sustainable development of the wind power industry.

[0038] And, in this embodiment, based on the size and morphological parameters of the cracks obtained by image segmentation, the finite element model is used to obtain the mechanical performance parameters of the wind turbine tower; according to the size and morphological parameters of the cracks obtained, as well as the mechanical performance parameters, three indicators of the tower structure integrity index, the tower damage comprehensive coefficient and the tower safety factor are determined; the three indicators of the tower structure integrity index, the tower damage comprehensive coefficient and the tower safety factor are weighted and summed to obtain a comprehensive evaluation index; finally, the wind turbine tower damage is evaluated using the comprehensive evaluation index; when evaluating the wind turbine tower damage, the relationship between features such as the middle tower and cracks and the internal mechanical properties of the tower itself is considered, and multiple indicators such as the tower structure integrity index, the tower damage comprehensive coefficient and the tower safety factor are combined to ensure the accuracy of the evaluation of the tower damage. The method in this embodiment includes:

[0039] S1. Image acquisition:

[0040] Optionally, in the image acquisition phase, a drone equipped with a high-definition camera is used. The drone takes all-round and multi-angle photos of the wind turbine tower through a preset flight path and shooting angle. The camera resolution and shooting parameters are debugged to ensure that clear and detailed images are obtained. In this step, the drone's flight control system and the camera parameter settings work together to ensure the integrity and quality of the image, providing a reliable data basis for subsequent analysis.

[0041] In order to improve the comprehensiveness and clarity of wind turbine tower images, a shooting method of multi-layer surround shooting, oblique shooting and / or dynamic adjustment of shooting parameters is adopted; when obtaining tower images, the tower is divided into multiple layers according to a preset distance, and the front image of the tower is collected on each layer, and the image of the tower at a certain angle is collected on each layer. Specifically, in the multi-layer surround shooting method, the drone is allowed to perform multiple layers of surround flights at different heights around the tower to ensure that every angle and height can be photographed. For example, starting from the bottom of the tower, a layer of surround shooting is performed at a certain height, and gradually rises to the top, such as a layer of surround shooting every 5 meters; in addition to horizontal surround, a certain angle of oblique shooting is introduced to capture more details on the side of the tower and reduce shadows and blurred areas caused by light and viewing angle; the shooting parameters are dynamically adjusted, and the camera parameters such as focal length, aperture and shutter speed are adjusted in real time according to the distance and angle between the drone and the tower to ensure image clarity and uniform exposure.

[0042] In order to further improve the comprehensiveness and clarity of wind turbine tower images, high-performance shooting equipment should be selected, including high-resolution cameras, wide-angle lenses and / or cameras with good low-light performance. Specifically, cameras with high pixel counts, such as 4K or even 8K resolution cameras, should be selected to obtain clearer and more detailed images and capture tiny cracks and defects; wide-angle lenses should be used to expand the shooting field of view and reduce the missed areas in a single shot, especially for the upper part of the tower and difficult-to-access areas, so that more complete images can be obtained; cameras that can still maintain good imaging effects under low-light conditions should be selected to meet the shooting needs under different weather and time conditions.

[0043] In some embodiments, advanced image enhancement technology is introduced, including multi-frame synthesis technology and / or image sharpening and deblurring algorithms. Specifically, during the shooting process, multiple images are taken continuously, and then these images are synthesized through an image synthesis algorithm to remove noise and enhance the clarity and contrast of the image; the acquired original image is post-processed, and a sharpening algorithm is applied to highlight the edges and details in the image, and a deblurring algorithm is used to eliminate image blur caused by the shaking of the drone or the influence of wind.

[0044] In some embodiments, other auxiliary equipment and technologies are combined, including laser rangefinders, positioning and attitude control systems and / or thermal imaging technologies. Specifically, the laser rangefinder is installed on the drone and works in conjunction with the camera to measure the distance between the drone and the tower in real time, providing accurate data for adjusting the shooting parameters; the positioning and attitude control system is used to ensure the stable positioning and accurate attitude control of the drone during flight, reduce the deviation and jitter of the shooting angle, and thus improve the quality of the image; combined with a thermal imaging camera, the temperature distribution on the surface of the tower is detected to help find potential damage areas, especially those problems that are not obvious in visible light images.

[0045] In some embodiments, optimizing the shooting environment and time includes selecting appropriate weather conditions and / or controlling the shooting time. Specifically, shooting is performed in clear, windless or breezy weather to avoid the effects of cloud cover, rain or strong wind on image quality; and shooting is performed in a time of the day when the light is uniform and soft, such as in the morning or afternoon, to avoid excessive contrast between highlights and shadows caused by direct sunlight.

[0046] By comprehensively applying the above solutions, the comprehensiveness and clarity of wind turbine tower images can be effectively improved, providing a more reliable basis for subsequent damage diagnosis.

[0047] S2. Image processing:

[0048] After acquiring the image data, the acquired image is processed using an image segmentation algorithm based on a convolutional neural network. Optionally, the constructed convolutional neural network model is trained with a large amount of image data labeled with towers and cracks, and can accurately identify and segment the towers and crack areas. In this process, through multi-layer convolution and pooling operations, the features of the image are extracted to achieve effective separation of the complex background.

[0049] Optionally, in the convolutional neural network, an attention module is added to the convolutional neural network so that the network can pay more attention to important areas related to towers and cracks in the image and suppress interference from irrelevant background; convolution kernels of different sizes are used to extract multi-scale feature information and fuse them to better capture the characteristics of towers and cracks at different scales; depthwise separable convolution is used instead of traditional convolution to reduce the number of parameters and calculations while maintaining or improving segmentation performance; residual connections are added to solve the gradient vanishing problem caused by increased network depth, promote the effective transmission of information, and improve the training effect and generalization ability of the model.

[0050] The image segmentation algorithm can adopt: U-Net structure, which extracts the contextual information of the image by contracting the path, and then restores the details and spatial resolution of the image by expanding the path to achieve accurate pixel-level segmentation; using the fully convolutional layer to replace the fully connected layer in the traditional convolutional neural network to achieve end-to-end segmentation of input images of any size; DeepLab series, such as DeepLabv3+, combines void convolution to expand the receptive field and obtain richer semantic information, while using the encoder-decoder structure to refine the segmentation boundaries.

[0051] The detailed process of image segmentation is:

[0052] S2.1. Data preparation:

[0053] Collect a large number of images containing wind turbine towers and cracks, annotate them, and mark the tower and crack areas in the images as training data. Perform preprocessing on the data, such as cropping, rotating, flipping, etc., to increase data diversity and reduce the risk of overfitting.

[0054] S2.2, Model training:

[0055] Input the prepared training data into the improved convolutional neural network for training. Use stochastic gradient descent (SGD) or other optimization algorithms to update the parameters of the network to minimize the loss function, such as cross entropy loss or Dice loss.

[0056] During the training process, validation is performed regularly, and hyperparameters such as learning rate and convolution kernel size are adjusted based on performance indicators on the validation set, such as accuracy, recall, and Dice coefficient.

[0057] S2.3. Model prediction:

[0058] The wind turbine tower image to be segmented is input into the trained image segmentation model. The image segmentation model outputs the probability of each pixel belonging to the tower or crack through forward propagation calculation. According to the set threshold, the probability value is converted into a binary image to obtain the final segmentation result, that is, the tower and crack areas are distinguished.

[0059] S2.4, post-processing:

[0060] Morphological operations such as dilation, erosion, opening and closing are performed on the segmentation results to remove noise and fill holes, making the segmentation results smoother and more continuous. The segmentation results are further modified and optimized based on prior knowledge and geometric constraints, such as adjusting the segmentation boundaries according to the shape and structural characteristics of the tower.

[0061] Through the above improvements, algorithm selection and detailed segmentation process, the accuracy and reliability of wind turbine tower image segmentation can be improved, providing a more precise basis for subsequent damage diagnosis.

[0062] S2.5. After the image segmentation is completed, the size of the crack is accurately obtained through proportional calculation with the help of the known tower size. This step requires the establishment of a mathematical relationship between the tower size and the corresponding size in the image, and the accuracy of the crack size is ensured through precise measurement and calculation.

[0063] S3. Finite element simulation analysis:

[0064] According to the actual situation of the crack, its possible shape and characteristics are defined in advance, such as rectangle, ellipse, etc. The finite element model is established using ABAQUS software, and the parameters such as the size and shape of the crack are input into the model. In the process of model establishment, the material properties, geometric shape, boundary conditions and actual stress conditions of the tower are fully considered to achieve accurate parametric analysis of cracking. The steps for establishing the finite element model are as follows:

[0065] S3.1. Geometric modeling:

[0066] Use 3D scanning technology to obtain the actual geometric shape of the tower, or use CAD software for high-precision modeling to ensure the consistency between the model and the actual structure.

[0067] S3.2. Reasonable material definition:

[0068] The real mechanical performance parameters of the tower material, including elastic modulus, yield strength, Poisson’s ratio, etc., are obtained through experimental testing and accurately input into ABAQUS, taking into account the nonlinear behavior of the material, such as plastic deformation and damage evolution.

[0069] S3.3. Fine grid division:

[0070] In the crack tip area, a dense grid is used and singular elements are used to simulate stress singularity and improve calculation accuracy.

[0071] S3.4, Boundary conditions and load simulation:

[0072] Accurately apply boundary constraints and loads, including wind loads, deadweight, and temperature changes, based on actual installation conditions and operating conditions.

[0073] Through precise finite element modeling, the expansion trend of tower cracks and the remaining life of the structure can be predicted more accurately, providing a more reliable basis for maintenance and replacement decisions. Multi-condition simulation can also be performed to analyze the response of the tower under different load combinations and optimize the structural design. In addition, it helps to deeply understand the formation mechanism and development law of tower cracks, and provide guidance for improving manufacturing processes and material selection.

[0074] S4. Evaluation:

[0075] Finally, a damage index that takes multiple factors into consideration is defined. This index covers key parameters such as crack size, shape, location, and mechanical properties obtained by finite element analysis. Through a large amount of experimental data and analysis, an accurate correspondence between crack morphology parameters and tower damage index is established.

[0076] S4.1. Through a large number of numerical simulation analyses of the obtained wind turbine tower crack images, the following corresponding relationship between the crack morphology parameters and the tower damage index is given:

[0077] Definition: Tower Structural Integrity Index (TSII) = (1-(total crack area / tower surface area)) × (1-maximum equivalent plastic strain) × (1-maximum stress intensity factor). Among the crack morphology parameters, the total crack area is the total area of ​​the cracks on the tower surface obtained through Abaqus simulation; the maximum equivalent plastic strain value calculated in the simulation; the maximum stress intensity factor value calculated in the simulation.

[0078] In the corresponding relationship, the larger the total crack area, the smaller the TSII value, indicating that the tower damage is more serious; the larger the maximum equivalent plastic strain, the smaller the TSII value, and the more serious the damage; the larger the maximum stress intensity factor, the smaller the TSII value, and the more serious the damage; when the TSII value is close to 1, it indicates that the tower damage is slight; when the TSII value is between 0.5-1, it is moderate damage; when the TSII value is less than 0.5, it is severe damage.

[0079] Definition: Tower Damage Comprehensive Coefficient (TDC) = (1-average crack width / tower wall thickness) × (1-average cracking energy release rate / reference energy release rate) × (1-average damage variable). Among the crack morphology parameters, the average crack width calculated in the simulation; the average cracking energy release rate obtained in the simulation; the average damage variable calculated in the simulation.

[0080] In the corresponding relationship, the larger the average crack width, the smaller the TDC value, and the more serious the damage; the larger the average cracking energy release rate, the smaller the TDC value, and the more serious the damage; the larger the average damage variable, the smaller the TDC value, and the more serious the damage; a TDC value greater than 0.8 indicates mild damage, 0.5-0.8 indicates moderate damage, and less than 0.5 indicates severe damage.

[0081] Definition: Tower Safety Factor (TSF) = (original strength / residual strength) × (1-total crack length / tower circumference) × (1-maximum stress concentration factor). Among the crack morphology parameters, the sum of all crack lengths obtained in the simulation; the maximum stress concentration factor calculated in the simulation.

[0082] In the corresponding relationship, the larger the total crack length, the smaller the TSF value, and the more serious the damage; the larger the maximum stress concentration factor, the smaller the TSF value, and the more serious the damage; the smaller the residual strength, the smaller the TSF value, and the more serious the damage; a TSF value greater than 2 indicates safety, between 1 and 2 indicates need for attention, and less than 1 indicates a dangerous state.

[0083] S4.2, indicator fusion:

[0084] The tower structural integrity index (TSII), tower damage coefficient (TDC) and tower safety factor (TSF) can be integrated into a comprehensive evaluation index by weighted summation. Assume that the weights of the three indicators are w1, w2 and w3, and w1+w2+w3=1.

[0085] Comprehensive evaluation index (CI) = w1×TSII+w2×TDC+w3×TSF

[0086] S4.3. Determine weights using the analytic hierarchy process (AHP):

[0087] S4.3.1. Establish a hierarchical structure: place the goal (comprehensive assessment of the degree of tower damage) at the top level and the three indicators at the bottom level.

[0088] S4.3.2. Construct a judgment matrix: Through expert scoring or based on experience, compare the relative importance of the three indicators in assessing the degree of tower damage to obtain a judgment matrix.

[0089] S4.3.3. Calculate weights: solve the eigenvector of the judgment matrix and obtain the weight values ​​w1, w2 and w3 of the three indicators.

[0090] Taking the analytic hierarchy process (AHP) as an example, the weight values ​​of the indicators are determined, and the weight values ​​of the tower structural integrity index (TSII), the tower damage coefficient (TDC) and the tower safety factor (TSF) are determined.

[0091] A hierarchical structure is established, including a target layer and a criterion layer. The target layer is to comprehensively evaluate the damage degree of the wind turbine tower, and the criterion layer is TSII, TDC and TSF.

[0092] Construct a judgment matrix. Optionally, invite several experts with rich experience in wind turbine tower damage assessment to compare the three indicators in pairs and score them according to the 1-9 scale (1 means that the two indicators are equally important, and 9 means that one indicator is extremely important than the other indicator). The judgment matrix A is shown in Table 1:

[0093] Table 1 Judgment Matrix

[0094] TSII TDC TSF TSII 1 3 5 TDC 1 / 3 1 3 TSF 1 / 5 1 / 3 1

[0095] Calculate the product of each row of the judgment matrix A:

[0096] M1=1×3×5=15

[0097]

[0098] Calculate M i The nth root of (n is the order of the judgment matrix, here n = 3):

[0099]

[0100] Pair Vector Normalize it and get the weight vector W:

[0101]

[0102] Therefore, the weight values ​​obtained are: TSII is about 0.637, TDC is about 0.258, and TSF is about 0.105.

[0103] Conduct consistency test:

[0104] Calculate the maximum eigenvalue λ of the judgment matrix max :

[0105] First, calculate the product AW of the judgment matrix A and the weight vector W:

[0106]

[0107] Then calculate λ max :

[0108]

[0109] Calculate the consistency index CI:

[0110]

[0111] Find the average random consistency index RI (for a 3-order matrix, RI is about 0.58)

[0112] Calculate the consistency ratio CR:

[0113]

[0114] Since CR < 0.1, the judgment matrix passes the consistency test and the obtained weights are valid.

[0115] The final weight values ​​are: TSII is about 0.637, TDC is about 0.258, and TSF is about 0.105. This means that when comprehensively evaluating the damage degree of wind turbine towers, TSII is relatively more important, followed by TDC, and TSF is relatively lower. It should be noted that the specific weight values ​​may vary depending on expert opinions and actual application scenarios, and need to be adjusted and verified according to specific circumstances.

[0116] S4.4. Prediction of wind turbine tower damage status:

[0117] Optionally, the obtained comprehensive evaluation index (CI) is compared with the preset index range to predict the damage state of the wind turbine tower; specifically, when the comprehensive evaluation index is less than the minimum value of the preset index range, the wind turbine tower damage state is safe, and no maintenance or further observation is required; when the comprehensive evaluation index is within the preset index range, the wind turbine tower damage state is low risk, and the inspection frequency of the wind turbine tower needs to be increased; when the comprehensive evaluation index is greater than the maximum value of the preset index range, the wind turbine tower damage state is predicted to be high risk, and further detailed observation and inspection of the wind turbine tower, or shutdown for maintenance, is required. The preset index range can be determined by historical evaluation index data and actual risk accident situations, and adjusted according to data updates.

[0118] In some other embodiments, in terms of image acquisition, multiple drones can be used for collaborative shooting, or higher performance photographic equipment can be used to obtain higher quality images. For image segmentation algorithms, in addition to convolutional neural networks, other advanced machine learning algorithms can also be tried, such as support vector machines, decision trees, etc., or combined with multiple algorithms for optimization. When establishing a finite element model, other similar finite element analysis software can also be selected and adjusted according to actual needs and software characteristics. When defining damage indicators, more influencing factors can be introduced, such as environmental factors, the service life of the tower, etc., to further improve the accuracy of damage diagnosis.

[0119] Embodiment 2:

[0120] This embodiment provides a wind turbine tower damage assessment system that comprehensively considers multiple indicators, including:

[0121] The image acquisition module is configured to: acquire an image of a wind turbine tower;

[0122] The image segmentation module is configured to: obtain the size and morphological parameters of the cracks in the wind turbine tower according to the acquired wind turbine tower image and a preset convolutional neural network model;

[0123] The finite element analysis module is configured to: perform finite element analysis according to the obtained size and shape parameters of the crack and the pre-established finite element model to obtain the mechanical performance parameters of the wind turbine tower;

[0124] The evaluation index determination module is configured to: determine the three indicators of tower structure integrity index, tower damage comprehensive coefficient and tower safety factor according to the obtained crack size and morphological parameters and mechanical performance parameters; perform weighted summation on the three indicators of tower structure integrity index, tower damage comprehensive coefficient and tower safety factor to obtain a comprehensive evaluation index;

[0125] The evaluation module is configured to: use the comprehensive evaluation index to evaluate the damage of the wind turbine tower.

[0126] The working method of the system is the same as the wind turbine tower damage evaluation method that comprehensively considers multiple indicators in Example 1, and will not be repeated here.

[0127] Embodiment 3:

[0128] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps of the wind turbine tower damage evaluation method that comprehensively considers multiple indicators as described in Example 1 are implemented.

[0129] Embodiment 4:

[0130] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the program, the steps of the wind turbine tower damage assessment method that comprehensively considers multiple indicators as described in Example 1 are implemented.

[0131] Embodiment 5:

[0132] This embodiment provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps of the wind turbine tower damage assessment method that comprehensively considers multiple indicators as described in Example 1 are implemented.

[0133] The above description is only a preferred embodiment of the present embodiment and is not intended to limit the present embodiment. For those skilled in the art, the present embodiment may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present embodiment shall be included in the protection scope of the present embodiment.

Claims

1. A wind turbine tower damage evaluation method that comprehensively considers multiple indicators is characterized by: include: Acquire wind turbine tower images; According to the acquired wind turbine tower image and the preset convolutional neural network model, the size and morphological parameters of the cracks in the wind turbine tower are obtained; Finite element analysis is performed based on the obtained crack size and morphological parameters and the pre-established finite element model to obtain the mechanical performance parameters of the wind turbine tower; According to the obtained crack size and morphological parameters, as well as mechanical performance parameters, three indicators, namely, tower structure integrity index, tower damage comprehensive coefficient and tower safety factor, are determined; the three indicators, namely, tower structure integrity index, tower damage comprehensive coefficient and tower safety factor, are weighted and summed to obtain a comprehensive evaluation index; The comprehensive evaluation index is used to evaluate wind turbine tower damage.

2. The wind turbine tower damage assessment method taking into account multiple indicators as claimed in claim 1 is characterized in that: The context information of the wind tower image is extracted by contracting the path, and the details and spatial resolution of the wind tower image are restored by expanding the path. The fully connected layer in the traditional convolutional neural network is replaced by the fully convolutional layer to achieve end-to-end segmentation of wind tower images of any size. The atrous convolution is used to expand the receptive field and obtain semantic information, and the encoder-decoder structure is used to refine the segmentation boundary. After the wind tower image is segmented, the size of the crack is obtained by proportional calculation with the help of the known tower size.

3. The wind turbine tower damage assessment method taking into account multiple indicators as claimed in claim 1 is characterized in that: When acquiring the image of the wind turbine tower, the tower is divided into multiple layers according to the preset distance, and the front image of the tower is collected around each layer, and the image of the tower at a preset angle is collected on each layer.

4. The wind turbine tower damage evaluation method taking into account multiple indicators as claimed in claim 1 is characterized in that: An attention module is added to the convolutional neural network to focus on important areas related to towers and cracks in the image and suppress interference from irrelevant background. Convolution kernels of different sizes are used to extract multi-scale feature information and fuse them to capture the characteristics of towers and cracks at different scales. Depthwise separable convolution is used instead of traditional convolution to reduce the number of parameters and computation.

5. The wind turbine tower damage assessment method taking into account multiple indicators as claimed in claim 1, characterized in that: During the finite element analysis, a denser grid is used in the crack tip area, and singular elements are used to simulate stress singularity phenomena. Boundary constraints and loads are applied according to the actual installation conditions and operating conditions, including wind loads, deadweight, and temperature changes.

6. The wind turbine tower damage assessment method taking into account multiple indicators as claimed in claim 1, characterized in that: Tower structural integrity index = (1-(total crack area / tower surface area)) × (1-maximum equivalent plastic strain) × (1-maximum stress intensity factor); tower damage comprehensive coefficient = (1-average crack width / tower wall thickness) × (1-average cracking energy release rate / reference energy release rate) × (1-average damage variable); tower safety factor = (original strength / residual strength) × (1-total crack length / tower circumference) × (1-maximum stress concentration factor).

7. A wind turbine tower damage evaluation system that comprehensively considers multiple indicators is characterized by: include: The image acquisition module is configured to: acquire an image of a wind turbine tower; The image segmentation module is configured to: obtain the size and morphological parameters of the cracks in the wind turbine tower according to the acquired wind turbine tower image and a preset convolutional neural network model; The finite element analysis module is configured to: perform finite element analysis according to the obtained size and shape parameters of the crack and the pre-established finite element model to obtain the mechanical performance parameters of the wind turbine tower; The evaluation index determination module is configured to: determine the three indicators of tower structure integrity index, tower damage comprehensive coefficient and tower safety factor according to the obtained crack size and morphological parameters and mechanical performance parameters; perform weighted summation on the three indicators of tower structure integrity index, tower damage comprehensive coefficient and tower safety factor to obtain a comprehensive evaluation index; The evaluation module is configured to: use the comprehensive evaluation index to evaluate the damage of the wind turbine tower.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the wind turbine tower damage evaluation method taking into account multiple indicators are implemented as described in any one of claims 1 to 6.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that: When the processor executes the program, the steps of the wind turbine tower damage evaluation method that comprehensively considers multiple indicators as described in any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, the steps of the wind turbine tower damage evaluation method taking into account multiple indicators are implemented as described in any one of claims 1 to 6.

Citation Information

Cited By

  • Damage assessment method and system for offshore wind power switching section

    CN122471264A

  • A method and system for damage assessment of an offshore wind power adapter segment

    CN122471264B