Fan tower welding seam and bolt detection method and system
By combining a wall-climbing adsorption robot with data processing equipment, defects in the welds and bolts of wind turbine towers can be automatically identified, solving the problems of high difficulty, high cost, and low accuracy in existing technologies, and achieving efficient and accurate defect detection.
Patent Information
- Application Number
- CN202411444254.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-16
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-10-16
AI Technical Summary
In existing technologies, the inspection of welds and bolts in wind turbine towers is difficult, costly, cumbersome, and inaccurate, relying on the subjective judgment of inspection personnel.
An ultrasonic image is obtained by scanning the surface of the wind turbine tower using a wall-climbing adsorption robot. The image is then processed by anisotropic diffusion noise reduction using data processing equipment, and a defect detection model is used to identify defects, including automatic detection of defects such as weld incomplete fusion, weld incomplete penetration, weld cracks, weld fatigue, bolt breakage, and bolt loosening.
It achieves high accuracy and efficiency in detecting defects in welds and bolts of wind turbine towers, avoiding the difficulties and high costs of manual operations, and improving the flexibility and accuracy of detection.
Smart Images

Figure CN119021840A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of defect detection, and in particular to a method and system for detecting welds and bolts of a wind turbine tower. Background Art
[0002] The wind turbine tower is a structural support system that installs the wind turbine generator set high above the ground. It can bear the weight of the wind turbine generator set and lift the wind turbine generator set to a higher height. While ensuring the stability and safety of the entire wind turbine system, the wind turbine generator set can more effectively capture the wind flow in the sky, thereby improving the power generation efficiency. Welds and bolts are important parts of the wind turbine tower support structure. If the weld quality is poor or the bolts are not properly tightened, the wind turbine tower may become loose or deformed during operation, affecting the stability and working efficiency of the entire wind turbine system. Therefore, it is necessary to regularly inspect the welds and bolts of the wind turbine tower to detect problems in time and ensure the long-term stable operation of the wind turbine tower.
[0003] In the related art, inspectors with professional qualifications work in high altitude and narrow spaces to inspect the welds and bolts of wind turbine towers. Due to the limited space for personnel to move around in the air, the difficulty in setting up the inspection equipment, and the potential safety hazards of working at high altitudes, the inspection is difficult, costly, and cumbersome to operate. In addition, the judgment of weld and bolt defects depends on the subjective ideas of the inspectors, and it is difficult to ensure accuracy. Therefore, there is an urgent need for a method that can automatically detect welds and bolts of wind turbine towers and automatically perform defect identification and analysis. Summary of the invention
[0004] In view of this, an object of the present invention is to provide a method and system for detecting welds and bolts of a wind turbine tower, so as to solve the problems of difficulty in detecting welds and bolts of a wind turbine tower, complicated operation and low accuracy.
[0005] In a first aspect, the present invention provides a method for detecting welds and bolts of a wind turbine tower, the method being executed by a data processing device of a wind turbine tower weld and bolt detection system; the wind turbine tower weld and bolt detection system further comprises a wall-climbing adsorption robot; the wall-climbing adsorption robot is communicatively connected with the data processing device; the method comprises:
[0006] Acquire an ultrasonic image of the surface of the wind turbine tower; the ultrasonic image is obtained by the wall-climbing adsorption robot scanning the surface of the wind turbine tower;
[0007] Performing anisotropic diffusion denoising processing on the ultrasonic image to obtain a processed ultrasonic smoothed image;
[0008] The ultrasonic smoothed image is processed by using a defect detection model to obtain a defect detection result; the defect detection model is a neural network model that has been pre-trained with sample images of the wind turbine tower surface.
[0009] In an optional implementation, performing anisotropic diffusion denoising on the ultrasound image includes:
[0010] Get the anisotropic diffusion denoising model;
[0011] The ultrasonic image is input into an anisotropic diffusion denoising model, and the ultrasonic image is iterated through the diffusion coefficient, gradient operator and divergence operator of the anisotropic diffusion denoising model to obtain a processed ultrasonic smoothed image.
[0012] In an optional implementation, the using the defect detection model to process the ultrasonic smoothed image to obtain the defect detection result includes:
[0013] Performing sliding cropping on the ultrasonic smoothed image based on a target step size to generate sub-images of the ultrasonic smoothed image;
[0014] Using the defect detection model to process each sub-image of the ultrasonic smoothed image, respectively obtain the probability value of each defect classification corresponding to each sub-image; wherein each sub-image corresponds to the probability value of each defect classification; the defect classification includes weld lack of fusion, weld lack of penetration, weld crack, weld fatigue, bolt breakage, bolt loosening, and bolt preload reduction;
[0015] The voting method is used to take the mean of the probability values of each defect classification, and the defect classification with the largest mean is determined as the defect detection result.
[0016] In a second aspect, the present invention provides a wind turbine tower weld and bolt detection system, comprising:
[0017] A wall-climbing adsorption robot comprises a walking adsorption mechanism, a driving mechanism, an ultrasonic detection transducer and an ultrasonic signal transmission mechanism; the driving mechanism is used to control the walking adsorption mechanism so that the wall-climbing adsorption robot is adsorbed on the surface of a wind turbine tower and moves along a target route; the ultrasonic detection transducer is used to collect ultrasonic signals from the surface of the wind turbine tower and transmit them to the driving mechanism through the ultrasonic signal transmission mechanism; the driving mechanism also comprises a signal transmitting and receiving device for transmitting ultrasonic signals to a ground control platform;
[0018] The ground control platform comprises a data receiving module and a data processing device; the data receiving module is used to receive the ultrasonic signal transmitted by the signal transmitting and receiving device; the data processing device comprises:
[0019] A data acquisition module, used to acquire an ultrasonic image of the surface of the wind turbine tower; the ultrasonic image is obtained by processing the ultrasonic signal;
[0020] A denoising module, used for performing anisotropic diffusion denoising on the ultrasonic image to obtain a processed ultrasonic smoothed image;
[0021] The defect detection module is used to process the ultrasonic smoothed image using a defect detection model to obtain a defect detection result; the defect detection model is a neural network model that has been pre-trained with sample images of the wind turbine tower surface.
[0022] In an optional embodiment, the walking adsorption mechanism also includes two symmetrical track modules for moving on the surface of the wind turbine tower; the walking adsorption mechanism also includes a permanent magnet adsorption assembly; the permanent magnet adsorption assembly is placed at the center of gravity of the wall-climbing adsorption robot to provide the same driving torque for the two track modules.
[0023] In an optional embodiment, the driving mechanism also includes a controller, which is provided with a surface motion control algorithm for calculating the output torque according to the inclination angle of the wind turbine tower surface, the friction and contact friction coefficient between the track module and the wind turbine tower surface, the gravity of the wall-climbing adsorption robot, the turning radius, the motor and gear transmission ratio, and the motor and gear transmission efficiency, and controlling the walking adsorption mechanism through the output torque.
[0024] In an optional embodiment, the ultrasonic detection transducer is arranged at the front end of the wall-climbing adsorption robot and close to the surface of the wind turbine tower to transmit ultrasonic waves to the surface of the wind turbine tower and receive ultrasonic signals reflected by the surface of the wind turbine tower.
[0025] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor are communicatively connected to each other, computer instructions are stored in the memory, and the processor executes the wind turbine tower weld and bolt detection method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.
[0026] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the wind turbine tower weld and bolt detection method of the first aspect or any corresponding embodiment thereof.
[0027] In a fifth aspect, the present invention provides a computer program product, comprising computer instructions for causing a computer to execute the wind turbine tower weld and bolt detection method of the first aspect or any corresponding embodiment thereof.
[0028] The technical solution provided by the present invention may include the following beneficial effects:
[0029] The wind turbine tower weld and bolt detection method provided by the present invention is executed by a data processing device of a wind turbine tower weld and bolt detection system. The wind turbine tower weld and bolt detection system also includes a wall-climbing adsorption robot. The wall-climbing adsorption robot scans the surface of the wind turbine tower to obtain an ultrasonic image of the surface of the wind turbine tower. The ultrasonic image is then anisotropically diffused and denoised by the data processing device to obtain a processed ultrasonic smooth image. The ultrasonic smooth image is processed by a defect detection model to obtain a defect detection result, thereby avoiding the problems of great detection difficulty, high cost, cumbersome operation and low accuracy caused by manual work. The scheme is flexible, and the accuracy and efficiency are high when performing wind turbine tower weld and bolt defect detection. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0031] Figure 1 1 is a flow chart of a method for detecting welds and bolts of a wind turbine tower according to an embodiment of the present invention;
[0032] Figure 2 is a flow chart of another method for detecting welds and bolts of a wind turbine tower according to an embodiment of the present invention;
[0033] Figure 3 is a structural block diagram of a wind turbine tower weld and bolt detection system according to an embodiment of the present invention;
[0034] Figure 4 is a schematic structural diagram of a wall-climbing adsorption robot involved in an embodiment of the present invention;
[0035] Figure 5 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION
[0036] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.
[0037] In the related art, inspectors with professional qualifications work at high altitudes and in narrow spaces to inspect the welds and bolts of wind turbine towers. Due to the limited space for personnel to move around at high altitudes, the difficulty in setting up detection instruments, and the potential safety hazards of working at high altitudes, detection is difficult, costly, and cumbersome to operate. In addition, the judgment of weld and bolt defects depends on the subjective ideas of inspectors, and it is difficult to ensure accuracy. Therefore, an embodiment of the present invention provides a method for inspecting welds and bolts of wind turbine towers, which uses a wall-climbing adsorption robot to scan the surface of the wind turbine tower to obtain an ultrasonic image, and uses a data processing device to process the ultrasonic image to obtain a defect detection result, so that the accuracy and efficiency of the weld and bolt defect detection of wind turbine towers are high.
[0038] According to an embodiment of the present invention, an embodiment of a method for detecting welds and bolts of a wind turbine tower is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0039] In this embodiment, a method for detecting welds and bolts of a wind turbine tower is provided, which is executed by a data processing device of a wind turbine tower weld and bolt detection system, wherein the wind turbine tower weld and bolt detection system further comprises a wall-climbing adsorption robot, which is communicatively connected to the data processing device. The data processing device may be an industrial computer, a laptop computer, a desktop computer, etc. Figure 1 is a flow chart of a method for detecting welds and bolts of a wind turbine tower according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:
[0040] Step S101, acquiring an ultrasonic image of the surface of a wind turbine tower.
[0041] The ultrasonic image is obtained by scanning the surface of the wind turbine tower by the wall-climbing adsorption robot. The wall-climbing adsorption robot can move on the surface of the wind turbine tower, emit ultrasonic waves to the surface of the wind turbine tower, and receive ultrasonic signals reflected by the surface of the wind turbine tower. Then, the data processing device can receive the ultrasonic signal to obtain an ultrasonic image of the surface of the wind turbine tower.
[0042] Step S102: performing anisotropic diffusion denoising processing on the ultrasonic image to obtain a processed ultrasonic smoothed image.
[0043] Exemplarily, the diffusion coefficient and diffusion steps of anisotropic diffusion are first selected. The diffusion coefficient controls the propagation speed of information in different directions during the anisotropic diffusion process, affecting the denoising effect and the degree of image edge retention. The diffusion steps are the number of times the diffusion process is applied to each pixel or each region. Next, an anisotropic diffusion algorithm is selected, such as a nonlinear diffusion algorithm or a structural tensor anisotropic diffusion algorithm. Finally, the ultrasound image is denoised by using the selected anisotropic diffusion algorithm, which is iteratively applied to each pixel or each region of the image until a preset convergence condition is reached.
[0044] Step S103: Process the ultrasonic smoothed image using the defect detection model to obtain a defect detection result.
[0045] The defect detection model is a neural network model that has been pre-trained with sample images of the wind tower surface. The sample images of the wind tower surface include various defect images of welds and bolts on the surface of the wind tower, such as cracked welds, non-compliance with specifications, loose bolts, etc. The defect detection model trained with sample images of the wind tower surface can accurately identify various defects of wind tower welds and bolts to obtain defect detection results.
[0046] The wind turbine tower weld and bolt detection method provided in the present embodiment is executed by a data processing device of a wind turbine tower weld and bolt detection system. The wind turbine tower weld and bolt detection system also includes a wall-climbing adsorption robot. The wall-climbing adsorption robot scans the surface of the wind turbine tower to obtain an ultrasonic image of the surface of the wind turbine tower. The ultrasonic image is then subjected to anisotropic diffusion denoising processing by the data processing device to obtain a processed ultrasonic smooth image. The ultrasonic smooth image is processed using a defect detection model to obtain a defect detection result. The problems of difficulty, high cost, cumbersome operation and low accuracy of detection caused by manual work are avoided. The scheme is flexible, and the accuracy and efficiency of wind turbine tower weld and bolt defect detection are good.
[0047] In this embodiment, a method for detecting welds and bolts of a wind turbine tower is provided, which is executed by a data processing device of a wind turbine tower weld and bolt detection system, wherein the wind turbine tower weld and bolt detection system further comprises a wall-climbing adsorption robot, which is communicatively connected to the data processing device. The data processing device may be an industrial computer, a laptop computer, a desktop computer, etc. Figure 2 is a flow chart of a method for detecting welds and bolts of a wind turbine tower according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:
[0048] Step S201, obtaining a defect detection model.
[0049] The defect detection model is a neural network model that has been pre-trained using sample images of the wind turbine tower surface.
[0050] Optionally, sample images of the wind turbine tower surface corresponding to various defects of the welds and bolts on the wind turbine tower surface are collected to form a sample data set, such as unfused welds, incomplete weld penetration, weld cracks, weld fatigue, as well as bolt breakage, bolt loosening, and decreased bolt preload. Next, obtain a defect detection model to be trained, which can be a deep learning classification model, a convolutional neural network model, etc. In order to reduce the amount of training, a pre-trained model with image detection function can also be obtained as the defect detection model to be trained. Finally, the sample data set composed of the collected sample images of the wind turbine tower surface is used to train the defect detection model to be trained. The sample images of the wind turbine tower surface are first preprocessed and then normalized as input data. The normalization formula is as follows:
[0051]
[0052] Among them, x1 represents the minimum value of the input data, and x2 represents the maximum value of the input data.
[0053] The preprocessed sample data set is then input into the defect detection model to be trained, and the cross entropy cost function is used to update the model parameters using the stochastic gradient descent method. The back propagation algorithm is used for supervised fine-tuning to optimize the model, and finally the defect detection model is obtained. The defect detection model obtained after training can detect various defects corresponding to the welds and bolts on the surface of the wind turbine tower.
[0054] For example, deep learning has the characteristics of feature extraction and data dimensionality reduction, and can autonomously learn to obtain feature expressions from sample data. For deep learning classification models, convolutional neural networks (CNN) are more conducive to image data processing because they are highly invariant to data displacement, scaling, distortion, etc. The basic structure of a convolutional neural network includes: input layer, convolution layer, downsampling layer, connection layer, and output layer. The calculation formula for convolving an image using a convolution filter is as follows:
[0055]
[0056] Where f() represents the activation function, F conv2D Represents the two-dimensional convolution formula, bj represents the bias value, ωij represents the weight between neurons, and yj represents the jth neuron input.
[0057] A pooling layer is added after the convolution layer to subsample the feature map. By merging and pooling, the dimension of the image features and the amount of model calculation are reduced. The fully connected layer in the network structure is used to integrate local information, and finally a classifier is used for classification.
[0058] Step S202, acquiring an ultrasonic image of the surface of the wind turbine tower.
[0059] The ultrasonic image is obtained by scanning the surface of the wind turbine tower by the wall-climbing adsorption robot. Ultrasonic imaging mainly uses the strength of echoes for imaging. The propagation speed of sound waves in different media is different. Ultrasonic waves, like ordinary sound waves, are mechanical waves. Generally, sound waves with a frequency higher than 20,000 Hz are considered ultrasonic waves. When encountering defects such as pores or cracks with a diameter smaller than the wavelength, scattering will occur, which specifically appears as dark lines on the ultrasonic imaging, thereby realizing the state detection of solid media. The obtained image is a cross-section parallel to the direction of sound propagation and perpendicular to the measurement surface of the tower.
[0060] Optionally, a B-type line scan is used to measure the surface of the wind turbine tower to obtain an ultrasonic image.
[0061] Step S203: performing anisotropic diffusion denoising processing on the ultrasonic image to obtain a processed ultrasonic smoothed image.
[0062] Specifically, the above step S203 includes:
[0063] Step S2031, obtaining an anisotropic diffusion denoising model.
[0064] The main principle of anisotropic diffusion is to use the time variable t as the variance of the Gaussian kernel G to convolve the noise image, and use the image to be processed I0 as the medium. The image diffuses at a variable rate to obtain an enhanced smooth image.
[0065] Step S2032: input the ultrasound image into an anisotropic diffusion denoising model, iterate the ultrasound image through the diffusion coefficient, gradient operator and divergence operator of the anisotropic diffusion denoising model, and obtain a processed ultrasound smoothed image.
[0066] Exemplarily, the heat diffusion equation is as follows:
[0067]
[0068] Among them, I(x,y,t) represents the original ultrasound image, the time variable t represents the order parameter, and the diffusion process evolves over time, that is, the number of iterations, c(x,y,t), and div represent the diffusion coefficient, gradient operator and divergence operator respectively. Their sizes depend on the strength of the gradient and are generally taken as I(x,y,t)=g(I(x,y,t)).
[0069] Step S204: Process the ultrasonic smoothed image using the defect detection model to obtain a defect detection result.
[0070] Specifically, the above step S204 includes:
[0071] Step S2041 , performing sliding cropping on the ultrasonic smoothed image based on the target step length to generate various sub-images of the ultrasonic smoothed image.
[0072] The target step size can be selected according to actual needs, for example, set to 10.
[0073] Step S2042: Process each sub-image of the ultrasonic smoothed image using the defect detection model to obtain a probability value of each defect classification corresponding to each sub-image.
[0074] This defect classification includes lack of weld fusion, incomplete weld penetration, weld cracks, weld fatigue, bolt breakage, bolt loosening, and decreased bolt preload.
[0075] Exemplarily, each sub-image of the ultrasonic smoothed image is preprocessed and normalized and then input into the defect detection model for feature extraction and defect detection, and then the Softmax classifier is used to obtain the probability value of each defect classification corresponding to each sub-image, wherein each sub-image corresponds to a probability value of each defect classification.
[0076] Step S2043, using the voting method to obtain the mean of the probability values of each defect classification, and determining the defect classification with the largest mean as the defect detection result.
[0077] The probability values of each defect classification corresponding to each sub-image are comprehensively counted by the voting method, and the mean of the probability values of each defect classification is taken. The defect classification with the largest mean is determined as the defect detection result of the ultrasonic smooth image. Since there may be multiple defects in the welds and bolts on the surface of the wind turbine tower, the defect classification with the largest target number of means can also be used as the defect detection result of the ultrasonic smooth image, for example, the defect classifications with the top three means are determined as the defect detection results.
[0078] The wind turbine tower weld and bolt detection method provided in the present embodiment is executed by a data processing device of a wind turbine tower weld and bolt detection system. The wind turbine tower weld and bolt detection system also includes a wall-climbing adsorption robot. The wall-climbing adsorption robot scans the surface of the wind turbine tower to obtain an ultrasonic image of the surface of the wind turbine tower. The ultrasonic image is then subjected to anisotropic diffusion denoising processing by the data processing device to obtain a processed ultrasonic smooth image. The ultrasonic smooth image is processed using a defect detection model to obtain a defect detection result. The problems of difficulty, high cost, cumbersome operation and low accuracy of detection caused by manual work are avoided. The scheme is flexible, and the accuracy and efficiency of wind turbine tower weld and bolt defect detection are good.
[0079] In this embodiment, a wind turbine tower weld and bolt detection system is also provided, in which a data processing device is used to implement the above-mentioned embodiments and preferred implementation modes, which have been described and will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function.
[0080] This embodiment provides a wind turbine tower weld and bolt detection system. Figure 3 As shown, it includes a wall-climbing adsorption robot and a ground control platform.
[0081] Figure 4 The schematic diagram of the structure of the wall-climbing adsorption robot involved in an embodiment of the present invention is shown. The wall-climbing adsorption robot includes a walking adsorption mechanism 1, a driving mechanism 2, an ultrasonic detection transducer 4 and an ultrasonic signal transmission mechanism 5. The driving mechanism 2 is used to control the walking adsorption mechanism 1 so that the wall-climbing adsorption robot moves along the target route by adsorption on the surface of the wind turbine tower. The ultrasonic detection transducer 4 is used to collect ultrasonic signals on the surface of the wind turbine tower and transmit them to the driving mechanism 2 through the ultrasonic signal transmission mechanism 5. The driving mechanism 2 also includes a signal transmitting and receiving device for transmitting ultrasonic signals to a ground control platform.
[0082] The ground control platform includes a data receiving module and a data processing device. The data receiving module is used to receive the ultrasonic signal transmitted by the signal transmitting and receiving device. The data processing device includes a data acquisition module, a denoising module and a defect detection module. The data acquisition module is used to obtain an ultrasonic image of the surface of the wind turbine tower. The ultrasonic image is obtained by processing the ultrasonic signal. Exemplarily, the data acquisition module performs ultrasonic imaging processing on the ultrasonic signal to obtain an ultrasonic image. The denoising module is used to perform anisotropic diffusion denoising processing on the ultrasonic image to obtain a processed ultrasonic smooth image. The defect detection module is used to process the ultrasonic smooth image using a defect detection model to obtain a defect detection result. The defect detection model is a neural network model that has been pre-trained with sample images of the wind turbine tower surface.
[0083] In an optional embodiment, the denoising module is also used to obtain an anisotropic diffusion denoising model, input the ultrasound image into the anisotropic diffusion denoising model, iterate the ultrasound image through the diffusion coefficient, gradient operator and divergence operator of the anisotropic diffusion denoising model, and obtain a processed ultrasound smooth image.
[0084] In an optional embodiment, the defect detection module is also used to perform sliding cropping on the ultrasonic smooth image based on the target step size to generate sub-images of the ultrasonic smooth image, and use the defect detection model to process the sub-images of the ultrasonic smooth image to obtain probability values of each defect classification corresponding to each sub-image; wherein each sub-image corresponds to a probability value of each defect classification; the defect classification includes weld incomplete fusion, weld incomplete penetration, weld crack, weld fatigue, bolt breakage, bolt looseness, and decreased bolt preload. The voting method is used to take the mean of the probability values of each defect classification, and the defect classification with the largest mean is determined as the defect detection result.
[0085] In an optional embodiment, the walking adsorption mechanism also includes two symmetrical crawler modules for moving on the surface of the wind turbine tower. The walking adsorption mechanism 1 also includes a powerful permanent magnet adsorption component, which is an important functional part of the wall-climbing robot, especially when the robot needs to work at a height of tens of meters. The robot must have a strong adsorption force to cope with the wall attachment and load capacity during high-altitude work. The permanent magnet adsorption component is placed at the center of gravity of the wall-climbing adsorption robot to provide the same driving torque for the two crawler modules, ensuring that the adsorption force at the center of the wall-climbing adsorption robot is evenly distributed, and can be adsorbed on magnetic conductive walls such as wind turbine towers, and move on the surface of the tower through the crawler module, with stable movement and flexible steering. Optionally, the permanent magnet adsorption component uses neodymium iron boron magnets, which are high-powder, high-pressure, very high-density, high-magnetic storage, and strong adsorption. It has high temperature resistance, strong hardness, and can absorb objects hundreds of times its own weight, which is 30 to 50 times that of ordinary magnets of the same volume.
[0086] In an optional embodiment, the driving mechanism 2 is placed in the motor cabin inside the crawler module in the walking adsorption structure 1, which can control the walking adsorption mechanism 1 to move on the magnetic curved surface, and receive the signal of the ultrasonic detection transducer 4, and forward it to the ground control platform through wireless communication. The driving mechanism 2 also includes a controller. After the signal transmitting and receiving device receives the mobile signal wirelessly transmitted by the ground control platform, the controller can control the moving direction and distance of the wall-climbing adsorption robot according to the mobile signal. Specifically, a curved surface motion control algorithm is provided in the controller, and the controller can control the driving element motor in the driving mechanism 2 according to the algorithm to ensure the stable movement and precise positioning of the robot on the curved surface. Optionally, the controller calculates the output torque according to the inclination angle of the wind turbine tower surface, the friction and contact friction coefficient between the crawler module and the wind turbine tower surface, the gravity of the wall-climbing adsorption robot, the turning radius, the motor and gear transmission ratio, and the motor and gear transmission efficiency, and controls the walking adsorption mechanism through the output torque. Exemplarily, the output torque M of the driving element motor T for:
[0087]
[0088] Among them, μ is the contact friction coefficient between the track module and the surface of the wind turbine tower, F0 is the friction between the track module and the surface of the wind turbine tower, G is the gravity of the wall-climbing adsorption robot, R is the turning radius of the wall-climbing adsorption robot, β is the inclination angle of the wind turbine tower surface, i is the transmission ratio of the wall-climbing adsorption robot motor to the planetary gear (motor and gear transmission ratio), and η is the total efficiency of the wall-climbing adsorption robot from motor transmission to the final output of the motor to the driving wheel (motor and gear transmission efficiency).
[0089] Optionally, the wireless communication between the signal transmitting and receiving device of the wall-climbing adsorption robot and the data receiving module of the ground control platform adopts millimeter wave wireless communication technology, with a frequency range of 30 GHz to 300 GHz and a wavelength of about 1 mm to 10 mm. Compared with existing wireless communication technologies such as Wi-Fi, millimeter wave wireless communication technology has a larger bandwidth and higher data rate, and can stably transmit control signals and ultrasonic detection signals.
[0090] In an optional embodiment, the ultrasonic detection transducer is arranged at the front end of the wall-climbing adsorption robot and close to the surface of the wind turbine tower to transmit ultrasonic waves to the surface of the wind turbine tower and receive ultrasonic signals reflected by the surface of the wind turbine tower.
[0091] In an optional embodiment, the wall-climbing adsorption robot further includes a power supply mechanism 3 for providing a stable DC power supply to the walking adsorption mechanism 1 , the driving mechanism 2 , the ultrasonic detection transducer 4 , and the ultrasonic signal transmission mechanism 5 .
[0092] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.
[0093] The data processing equipment in the wind turbine tower weld and bolt detection system in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above-mentioned functions.
[0094] An embodiment of the present invention further provides a computer device having the data processing device in the above-mentioned wind turbine tower weld and bolt detection system.
[0095] See also Figure 5 , Figure 5 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Figure 5As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 5 A processor 10 is taken as an example.
[0096] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.
[0097] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.
[0098] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.
[0099] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.
[0100] The computer device also includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30 and the output device 40 may be connected via a bus or other means. Figure 5 The example of connecting through bus is taken in the following.
[0101] The input device 30 can receive input digital or character information, and generate key signal input related to the user settings and function control of the computer device, such as a touch screen, a keypad, a mouse, a track pad, a touch pad, an indicator bar, one or more mouse buttons, a trackball, a joystick, etc. The output device 40 may include a display device, an auxiliary lighting device (e.g., an LED) and a tactile feedback device (e.g., a vibration motor), etc. The above-mentioned display device includes but is not limited to a liquid crystal display, a light emitting diode, a display and a plasma display. In some optional embodiments, the display device can be a touch screen.
[0102] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.
[0103] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of the computer program instruction in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc., and accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.
[0104] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations shall all fall within the scope of protection of the present invention.
Claims
1. A method for detecting welds and bolts of a wind turbine tower, characterized in that: The method is executed by a data processing device of a wind turbine tower weld and bolt detection system; the wind turbine tower weld and bolt detection system also includes a wall-climbing adsorption robot; the wall-climbing adsorption robot is communicatively connected with the data processing device; the method includes: Acquire an ultrasonic image of the surface of the wind turbine tower; the ultrasonic image is obtained by the wall-climbing adsorption robot scanning the surface of the wind turbine tower; Performing anisotropic diffusion denoising processing on the ultrasonic image to obtain a processed ultrasonic smoothed image; The ultrasonic smoothed image is processed by using a defect detection model to obtain a defect detection result; the defect detection model is a neural network model that has been pre-trained with sample images of the wind turbine tower surface.
2. The method according to claim 1, characterized in that The performing anisotropic diffusion denoising processing on the ultrasonic image comprises: Get the anisotropic diffusion denoising model; The ultrasonic image is input into an anisotropic diffusion denoising model, and the ultrasonic image is iterated through the diffusion coefficient, gradient operator and divergence operator of the anisotropic diffusion denoising model to obtain a processed ultrasonic smoothed image.
3. The method according to claim 1 or 2, characterized in that: The method of processing the ultrasonic smoothed image using the defect detection model to obtain a defect detection result includes: Performing sliding cropping on the ultrasonic smoothed image based on a target step size to generate sub-images of the ultrasonic smoothed image; Using the defect detection model to process each sub-image of the ultrasonic smoothed image, respectively obtain the probability value of each defect classification corresponding to each sub-image; wherein each sub-image corresponds to the probability value of each defect classification; the defect classification includes weld lack of fusion, weld lack of penetration, weld crack, weld fatigue, bolt breakage, bolt loosening, and bolt preload reduction; The voting method is used to take the mean of the probability values of each defect classification, and the defect classification with the largest mean is determined as the defect detection result.
4. A wind turbine tower weld and bolt detection system, characterized in that: include: A wall-climbing adsorption robot comprises a walking adsorption mechanism, a driving mechanism, an ultrasonic detection transducer and an ultrasonic signal transmission mechanism; the driving mechanism is used to control the walking adsorption mechanism so that the wall-climbing adsorption robot is adsorbed on the surface of a wind turbine tower and moves along a target route; the ultrasonic detection transducer is used to collect ultrasonic signals from the surface of the wind turbine tower and transmit them to the driving mechanism through the ultrasonic signal transmission mechanism; the driving mechanism also comprises a signal transmitting and receiving device for transmitting ultrasonic signals to a ground control platform; A ground control platform, including a data receiving module and data processing equipment; The data receiving module is used to receive the ultrasonic signal transmitted by the signal transmitting and receiving device; the data processing device includes: A data acquisition module, used to acquire an ultrasonic image of the surface of the wind turbine tower; the ultrasonic image is obtained by processing the ultrasonic signal; A denoising module, used for performing anisotropic diffusion denoising on the ultrasonic image to obtain a processed ultrasonic smoothed image; The defect detection module is used to process the ultrasonic smoothed image using a defect detection model to obtain a defect detection result; the defect detection model is a neural network model that has been pre-trained with sample images of the wind turbine tower surface.
5. The system according to claim 4, characterized in that The walking adsorption mechanism also includes two symmetrical track modules for moving on the surface of the wind turbine tower; the walking adsorption mechanism also includes a permanent magnet adsorption assembly; the permanent magnet adsorption assembly is placed at the center of gravity of the wall-climbing adsorption robot to provide the same driving torque for the two track modules.
6. The system according to claim 5, characterized in that The driving mechanism also includes a controller, which is provided with a curved surface motion control algorithm for calculating the output torque according to the inclination angle of the wind turbine tower surface, the friction force and contact friction coefficient between the track module and the wind turbine tower surface, the gravity of the wall-climbing adsorption robot, the turning radius, the motor and gear transmission ratio, and the motor and gear transmission efficiency, and controlling the walking adsorption mechanism through the output torque.
7. The system according to any one of claims 4 to 6, characterized in that: The ultrasonic detection transducer is arranged at the front end of the wall-climbing adsorption robot and close to the surface of the wind turbine tower to transmit ultrasonic waves to the surface of the wind turbine tower and receive ultrasonic signals reflected by the surface of the wind turbine tower.
8. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the wind turbine tower weld and bolt detection method according to any one of claims 1 to 3 by executing the computer instructions.
9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the wind turbine tower weld and bolt detection method according to any one of claims 1 to 3.
10. A computer program product, characterized in that The method comprises computer instructions, wherein the computer instructions are used to enable a computer to execute the wind turbine tower weld and bolt detection method according to any one of claims 1 to 3.
Citation Information
Patent Citations
Automatic on-line detecting device for welding lines of mast of wind driven generator
CN101782552A
Control system of automatic detecting wall-climbing robot for fan tower cylinder welding seams
CN103853141A
Wind power tower welding seam defect detection device and method based on machine vision
CN111551565A
Wall surface nondestructive detection system and detection method of wall-climbing detection robot
CN115078540A
Metal weld defect detection method based on improved YOLOv5
CN116626166A
