Heater crack repairing method and device

Abstract

The invention provides a heater crack repairing method and device, relates to the technical field of crack repairing, and is used for solving the technical problems that a repairing mode in the related technology is low in efficiency and the spraying quality is difficult to guarantee. The heater crack repairing method comprises the steps that a crack image of the surface of a heater is obtained based on machine vision and transient thermal imaging; geometric features and thermal features of the crack are obtained, and contour information of the crack image is obtained based on an edge detection algorithm; a deep neural network model is trained, contour information, geometric features and thermal features are input into the deep neural network model, whether cracks need to be repaired or not is judged based on the deep neural network model, and the material amount needed for repairing the cracks is calculated; obtaining an optimal spraying path based on a graph theory or a genetic algorithm in combination with the contour information, the geometric features, the thermal features and the material quantity; the optimal spraying path is fed back to a self-adaptive spraying mechanical arm, and crack repairing is completed through the mechanical arm; and detecting whether the repaired crack area reaches a curing standard or not.

Description

Technical Field

[0001] This application relates to the technical field of crack repair, and particularly to a method and device for repairing cracks in a heater. Background Art

[0002] A bottom heater is a device used to heat the bottom of a heating device and is applied to various household and industrial devices such as electric kettles, rice cookers, coffee machines, ovens, and laboratory heating devices, etc. The main function of the bottom heater is to evenly transfer heat through bottom heating, thereby heating the container or the substances inside the container.

[0003] In a high-temperature environment, the heating area of the bottom heater will experience thermal stress cycles during the process of rapid heating and cooling, resulting in uneven expansion and contraction of the material, thereby causing cracks in the heating area of the bottom heater. In severe cases, it will also cause a sparking phenomenon, threatening the safety and energy efficiency of the device.

[0004] In the related art, manual inspection and manual spraying are used to repair the cracks generated in the heating area of the bottom heater. However, the above repair methods are inefficient and it is difficult to guarantee the spraying quality. Summary of the Invention

[0005] In view of the above problems, the embodiments of this application provide a method and device for repairing cracks in a heater to solve the problems of low efficiency of the repair method in the related art and difficult to guarantee the spraying quality.

[0006] To achieve the above object, the embodiments of this application provide the following technical solutions:

[0007] The embodiments of this application provide a method for repairing cracks in a heater, which includes: obtaining a crack image on the surface of the heater based on machine vision and transient thermal imaging technology; obtaining the geometric features and thermal features of the cracks in the crack image, and at the same time obtaining the contour information in the crack image based on an edge detection algorithm; training a deep neural network model using a historical crack sample data set; inputting the contour information, the geometric features, and the thermal features into the deep neural network model, and judging whether the crack needs to be repaired based on the deep neural network model and calculating the amount of material required to repair the crack; combining the contour information, the geometric features, the thermal features, and the amount of material, and obtaining an optimal spraying path based on graph theory or a genetic algorithm; feeding back the optimal spraying path to an adaptive spraying robotic arm, and completing the repair of the crack through the adaptive spraying robotic arm; detecting whether the repaired crack area reaches the curing standard.

[0008] In an embodiment of this application, the crack image is a two-dimensional or three-dimensional thermal image, and the thermal image includes a plurality of temperature data, and the plurality of temperature data are the temperature data of each material layer at different depths.

[0009] In an embodiment of the present application, the method for obtaining a crack image on the surface of a heater based on machine vision and transient thermal imaging technology includes: emitting a thermal pulse or a light pulse at the millisecond or microsecond level to the surface of the heater; capturing the temperature change on the surface of the heater through an infrared detector to obtain thermal imaging data; tracking the change of the thermal imaging data over time to obtain time series data; and converting the time series data into the crack image based on an image reconstruction method.

[0010] In an embodiment of the present application, before obtaining the contour information in the crack image based on an edge detection algorithm, it further includes: removing the noise in the crack image based on an image enhancement technology, and at the same time enhancing the contrast of the crack image to optimize the crack image.

[0011] In an embodiment of the present application, the geometric features include: crack length, crack width, and crack direction; the thermal features include: temperature distribution and heat flux change.

[0012] In an embodiment of the present application, after detecting whether the repaired crack area meets the curing standard, it further includes: detecting the actual spraying result, comparing the difference between the actual spraying result and the expected spraying result, and optimizing the deep neural network model.

[0013] In an embodiment of the present application, before feeding back the optimal spraying path to the adaptive spraying robot arm, it further includes: simulating the optimal spraying path based on computer simulation software, and modifying the optimal spraying path according to the simulation result.

[0014] In an embodiment of the present application, the adaptive spraying robot arm includes: a position sensor, a force sensor, and a vision sensor. The position sensor is used to monitor the position of the spray gun on the adaptive spraying robot arm in real time. The force sensor is used to monitor the pressure exerted by the spray gun on the surface of the heater in real time. The vision sensor is used to monitor the distance between the spray gun and the surface of the heater in real time.

[0015] In an embodiment of the present application, the adaptive spraying robotic arm includes an intelligent algorithm control unit, which is configured to receive the optimal spraying path and the real-time data fed back by the position sensor, the force sensor, and the vision sensor, and dynamically adjust the action parameters of the adaptive spraying robotic arm based on the optimal spraying path and the real-time data; the action parameters include: the position of the spray gun, the movement trajectory of the spray gun, the pressure exerted by the spray gun on the surface of the heater, the distance between the spray gun and the surface of the heater, the output rate of the spraying material, the spraying angle of the nozzle, etc.; the movement trajectory of the spray gun includes the limit movement trajectory, and the intelligent algorithm control unit is configured to control the movement trajectory of the spray gun within the range of the limit movement trajectory; the pressure exerted by the spray gun on the surface of the heater includes the limit pressure, and the intelligent algorithm control unit is configured to control the pressure exerted by the spray gun on the surface of the heater to be less than the limit pressure.

[0016] The embodiment of the present application further provides a heater crack repair device, which includes: a crack detection module, a data acquisition module, a model training module, an intelligent analysis and decision-making module, a path generation module, an adaptive spraying robotic arm, and a curing effect detection module; the crack detection module is configured to obtain the crack image on the surface of the heater through machine vision and transient thermal imaging technology; the data acquisition module is configured to obtain the contour information in the crack image through an edge detection algorithm, and simultaneously obtain the geometric features and thermal features of the crack in the crack image through thermal imaging technology; the model training module is configured to train a deep neural network model using a historical crack sample data set; the intelligent analysis and decision-making module is configured to input the contour information, the geometric features, and the thermal features into the deep neural network model, and is configured to determine whether the crack needs to be repaired through the deep neural network model and calculate the amount of material required to repair the crack; the path generation module is configured to combine the contour information, the geometric features, the thermal features, and the amount of material, and obtain an optimal spraying path based on graph theory or a genetic algorithm; the path generation module is further configured to feed back the optimal spraying path to the adaptive spraying robotic arm to complete the repair of the crack through the adaptive spraying robotic arm; the curing effect detection module is configured to detect whether the repaired crack area meets the curing standard.

[0017] The heater crack repair method provided by the embodiment of the present application has the following technical effects:

[0018] When machine vision and transient thermal imaging technology are used in combination, machine vision is used to identify the appearance features of the heater surface, while transient thermal imaging technology is used to detect the thermal characteristics of the heater surface. The crack image obtained by combining the two is a two-dimensional or three-dimensional thermal image, which can display the temperature differences of each material layer at different depths, helping to identify hidden defects, discontinuous cracks or other foreign objects, thereby providing more comprehensive quality control.

[0019] An optimal spraying path is obtained through a composite algorithm composed of an edge detection algorithm, a deep neural network model, and graph theory or genetic algorithms. This optimal spraying path can ensure that all areas are evenly sprayed, while minimizing the spraying time or minimizing material consumption, not only improving the spraying efficiency, but also significantly reducing costs and material waste. Brief Description of the Drawings

[0020] To more clearly illustrate the technical solutions of the embodiments of the present application, the accompanying drawings required for use in the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0021] Figure 1 It is a flowchart of the steps of the heater crack repair method provided by the embodiment of the present application. Detailed Embodiments

[0022] In order to make the above objects, features, and advantages of the embodiments of the present application more obvious and understandable, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0023] Reference Figure 1 , the heater crack repair method provided by the embodiment of the present application includes:

[0024] Step S1: Based on machine vision and transient thermal imaging technology, obtain a crack image of the heater surface.

[0025] Machine vision is a technology that uses a computer and a camera device to simulate the human visual system to acquire, process, and analyze visual information.

[0026] Transient thermal imaging technology is a technique that excites an object with a short pulse and uses an infrared thermal imaging device to capture the temperature distribution on the object's surface. It can detect tiny changes in the surface temperature of the object and generate a thermal image. Transient thermal imaging technology is different from traditional steady-state thermal imaging. It can reveal the deep structure or defects on the object's surface, improving the detection depth and the resolution of the thermal image.

[0027] When machine vision and transient thermal imaging technology are used in combination, machine vision is used to identify the appearance features on the surface of the heater, while transient thermal imaging technology is used to detect the thermal characteristics of the heater surface. The crack image obtained by combining the two is a two-dimensional or three-dimensional thermal image, which includes multiple temperature data. The multiple temperature data are the temperature data of each material layer at different depths. That is to say, this thermal image can display the temperature differences of each material layer at different depths, helping to identify hidden defects, discontinuous cracks or other foreign objects, thereby providing more comprehensive quality control.

[0028] In the embodiment of this application, step S1: Obtaining the crack image on the surface of the heater based on machine vision and transient thermal imaging technology includes:

[0029] Step S11: Emit a thermal pulse or a light pulse in milliseconds or microseconds to the surface of the heater.

[0030] Step S12: Capture the temperature change on the surface of the heater through an infrared detector to obtain thermal imaging data.

[0031] Step S13: Track the change of the thermal imaging data over time to obtain time series data.

[0032] Step S14: Convert the time series data into a crack image based on the image reconstruction method.

[0033] Among them, the thermal pulse or light pulse in milliseconds or microseconds is a short pulse, which is very fast. The purpose is to briefly change the temperature distribution on the surface of the heater without causing long-term temperature rise on the surface of the heater.

[0034] After the short pulse excitation, a high-sensitivity and high-time-resolution infrared detector is used to quickly capture the temperature change on the surface of the heater, record the transient thermal effect caused by the short pulse, and obtain thermal imaging data.

[0035] The infrared detector can convert the received thermal imaging data into digital signal data and send it to the software analysis tool. The software analysis tool can track the change of the thermal imaging data over time and analyze these changes to obtain time series data. The time series data includes data on the temperature rise, temperature drop or abnormal temperature fluctuation on the surface of the heater over time, so as to extract the internal structure of the surface of the heater and the crack information on the surface of the heater.

[0036] Among them, the software analysis tools can be: FLIR Tools, FLUKE SmartView, MATLAB, ThermoVision ExaminIR, ImageJ with Plugins, ThermoViewer, LabVIEW, or OpenCV, etc.

[0037] After that, an image reconstruction model is used to convert the time series data into a crack image.

[0038] Among them, the image reconstruction methods can be: Fourier transform, filtered back projection, iterative reconstruction algorithm, compressive sensing, deep learning method, convolutional neural network, generative adversarial network, autoencoder, regularization method, or model-driven method.

[0039] Step S2: Obtain the geometric features and thermal features of the cracks in the crack image, and at the same time obtain the contour information in the crack image based on the edge detection algorithm.

[0040] The geometric features include: crack length, crack width, and crack direction, and the thermal features include: temperature distribution and heat flux change.

[0041] Edge detection is a basic task in image processing and computer vision, aiming to identify regions with significant brightness changes in the image, that is, contours; contours usually correspond to object boundaries, texture changes, or other significant features in the image.

[0042] The edge detection algorithm can include: Canny algorithm, Sobel algorithm, Prewitt algorithm, or Laplacian algorithm.

[0043] In the embodiment of the present application, before step S2: obtaining the contour information in the crack image based on the edge detection algorithm, it further includes:

[0044] Removing the noise in the crack image based on the image enhancement technology, and at the same time enhancing the contrast of the crack image based on the image enhancement technology to optimize the crack image.

[0045] Among them, the image enhancement technology includes filtering technology and histogram equalization method. The noise in the crack image is removed based on the filtering technology, and the contrast of the crack image is enhanced based on the histogram equalization method.

[0046] The filtering technology enhances or suppresses specific image features by performing a convolution operation on the image. The filtering technology can include: low-pass filter, high-pass filter, and median filter.

[0047] A low-pass filter can remove high-frequency noise in an image and smooth the image; a high-pass filter can enhance the high-frequency components in an image and highlight edges and details; a median filter can remove salt-and-pepper noise while preserving edges and is used to process images with sharp noise.

[0048] Histogram equalization is a technique for enhancing image contrast. It is achieved by redistributing the gray levels of the image, evenly distributing the gray levels of the image across the entire gray range, thereby improving the global contrast of the image. It is simple and efficient and applicable to various types of images.

[0049] Step S3: Train a deep neural network model using the historical crack sample dataset.

[0050] The historical crack sample dataset includes various types of cracks, and the various types of cracks include linear cracks, branched cracks, and annular cracks.

[0051] The historical crack sample dataset also includes the severity levels corresponding to the various types of cracks, learning the complex relationships between the morphology, location of the cracks and their severity levels from historical cases to optimize the deep neural network model.

[0052] The deep neural network model can be a convolutional neural network model. A convolutional neural network is a deep learning model specifically designed to process data with a grid structure and can automatically extract and learn the spatial hierarchical features of an image.

[0053] Step S4: Input the contour information, geometric features, and thermal features into the deep neural network model, and based on the deep neural network model, determine whether a crack needs to be repaired and calculate the amount of material required for crack repair.

[0054] Take the contour information of the crack image, the geometric features of the crack, and the thermal features of the crack as feature vectors and input them into the deep neural network model. Through the deep neural network model, predict the severity level of the crack and whether it needs to be repaired immediately, and at the same time estimate the amount of material required to repair the crack and the optimal spraying strategy.

[0055] Step S5: Combine the contour information, geometric features, thermal features, and material amount to obtain the optimal spraying path based on graph theory or genetic algorithms.

[0056] Obtaining the optimal spraying path based on graph theory includes: first, dividing the surface to be sprayed into multiple small regions or points, and each region or point can be regarded as a vertex in the graph; establishing edges between adjacent vertices, and the weight of the edge can represent the cost of moving from one region to another, such as time, distance, or the consumption of spraying materials; finding the shortest path from the starting point to the ending point, ensuring that each vertex is visited at least once and the total path length is the shortest, so as to ensure that all regions are evenly sprayed while minimizing the spraying time or minimizing the material consumption, thereby finding the optimal spraying path. This not only improves the spraying efficiency but also significantly reduces costs and material waste.

[0057] The genetic algorithm is an optimization technique based on the principles of natural selection and genetics. Obtaining the optimal spraying path based on the genetic algorithm includes: representing the spraying path as an individual, defining a fitness function based on factors such as path length, spraying time, and material consumption, and the fitness function is used to evaluate the quality of the individual; randomly generating a set of feasible spraying paths as the initial population; selecting individuals for reproduction according to the fitness function; generating new offspring individuals by combining the genes of two parent individuals; randomly changing a part of the genes of the individual to increase the population diversity; replacing the individuals in the population with the newly generated offspring; continuously iterating through selection, crossover, mutation, and replacement until the termination condition is met. When the algorithm terminates, select the individual with the highest fitness as the optimal spraying path.

[0058] Combining the crack distribution information with the repair requirements, the optimal spraying path obtained by using graph theory or the genetic algorithm can ensure the spraying uniformity while minimizing the material consumption and spraying time; while ensuring that all crack regions are processed, it can also avoid repeated coverage.

[0059] It should be noted that when obtaining the optimal spraying path by using graph theory or the genetic algorithm, the motion efficiency and reachability constraints of the adaptive spraying robotic arm also need to be considered to ensure that the adaptive spraying robotic arm can completely implement the optimal spraying path.

[0060] In the embodiment of the present application, before step S5: feeding back the optimal spraying path to the adaptive spraying robotic arm, it further includes:

[0061] Simulating the optimal spraying path based on computer simulation software and modifying the optimal spraying path according to the simulation results.

[0062] Using computer simulation software to simulate the accuracy rate of the optimal spraying path under different crack conditions, testing and optimizing the spraying path before actual operation according to the simulated results, further ensuring the spraying quality, improving the spraying efficiency, reducing costs, and reducing material waste.

[0063] Among them, the computer simulation software may include: ANSYS Fluent, COMSOL Multiphysics, Autodesk CFD, Simul8, SprayView, RobotStudio, etc.

[0064] Step S6: Feed the optimal spraying path back to the adaptive spraying robotic arm, and complete the repair of the crack through the adaptive spraying robotic arm.

[0065] The adaptive spraying robotic arm is a highly flexible and intelligent device that can automatically adjust parameters such as the movement trajectory of the spray gun, the output rate of the spraying material, and the spraying angle of the nozzle according to the crack distribution on the heater surface and the optimal spraying path, so as to achieve a high-quality spraying effect.

[0066] Among them, the adaptive spraying robotic arm has at least six-axis degrees of freedom, and the six-axis degrees of freedom include: moving along the x-axis, moving along the y-axis, moving along the z-axis, rotating around the x-axis, rotating around the y-axis, and rotating around the z-axis. The above six-axis degrees of freedom are controlled by servo motors or stepper motors at the robotic arm joints, and can perform precise movement and attitude adjustment in three-dimensional space, enabling the adaptive spraying robotic arm to reach any position on the heater surface and achieve uniform spraying.

[0067] The adaptive spraying robotic arm further includes: a position sensor, a force sensor, and a vision sensor. The position sensor is used to monitor the position of the spray gun on the adaptive spraying robotic arm in real time, the force sensor is used to monitor the pressure exerted by the spray gun on the heater surface in real time, and the vision sensor is used to monitor the distance between the spray gun and the heater surface in real time; ensuring the accuracy of the spraying operation and preventing overspraying or under-spraying.

[0068] The adaptive spraying robotic arm further includes: a protective cover, which is arranged outside the position sensor, the force sensor, and the vision sensor. The protective cover can prevent water and dust, and is used to protect the sensors and prevent the sensors from being damaged due to environmental factors.

[0069] The adaptive spraying robotic arm further includes: an intelligent algorithm control unit, which is used to receive the optimal spraying path, and is also used to receive the real-time data fed back by the position sensor, the force sensor, and the vision sensor, and dynamically adjust the action parameters of the adaptive spraying robotic arm based on the optimal spraying path and the real-time data.

[0070] The intelligent algorithms in the intelligent algorithm control unit include: path planning algorithms, machine vision guidance algorithms, force control algorithms, etc.

[0071] The path planning algorithm can calculate the optimal spraying path based on the three-dimensional model of the heater surface and the distribution data of cracks; the machine vision guidance algorithm is used to track the position and shape changes of the heater surface in real time to ensure the continuity and consistency of spraying; the force control algorithm can ensure that the adaptive spraying robot arm can quickly adapt to the changes in different areas of the heater surface during the spraying task and apply appropriate force to avoid damaging the heater surface.

[0072] The action parameters of the adaptive spraying robot arm include: the position of the spray gun, the movement trajectory of the spray gun, the pressure exerted by the spray gun on the heater surface, the distance between the spray gun and the heater surface, the output rate of the spraying material, and the spraying angle of the nozzle, etc.

[0073] In the embodiment of the present application, the movement trajectory of the spray gun includes the extreme movement trajectory, and the intelligent algorithm control unit is also used to control the movement trajectory of the spray gun within the range of the extreme movement trajectory, so that the adaptive spraying robot arm has a stroke limit to avoid collision when the adaptive spraying robot arm operates in a complex or narrow space, damaging the mechanical structure or the heater itself.

[0074] The pressure exerted by the spray gun on the heater surface includes the extreme pressure, and the intelligent algorithm control unit is also used to control the pressure exerted by the spray gun on the heater surface to be less than the extreme pressure, so that the adaptive spraying robot arm has a force limit to avoid damaging the mechanical structure or the heater itself.

[0075] In the embodiment of the present application, the spraying modes of the nozzle of the adaptive spraying robot arm include fan-shaped spraying, circular spraying, conical spraying, direct current spraying, or multi-hole spraying, etc.; the range of the spraying angle of the nozzle includes 0 degrees - 180 degrees; the range of the spraying flow rate of the nozzle includes 1 m / s - 200 m / s.

[0076] The nozzle can adjust the spraying mode, spraying angle, and spraying flow rate according to requirements to ensure that the spraying material can achieve an ideal covering effect under different spraying conditions, while reducing material waste.

[0077] The adaptive spraying robot arm further includes: an integrated feedback control system, which is used to monitor the entire spraying process in real time, obtain the spraying effect, such as spraying thickness and adhesion of the spraying material, and automatically adjust the optimal spraying path, spraying angle of the nozzle, spraying flow rate of the nozzle, position of the spray gun, pressure exerted by the spray gun on the heater surface, distance between the spray gun and the heater surface, etc. according to the spraying effect. Ensure that every step from detection to execution can be precisely adjusted according to the feedback information to maintain the stability and consistency of the spraying quality.

[0078] In the embodiment of the present application, the adaptive spraying robot arm includes: a material supply monitoring system, which is used to monitor the stock of the spraying material and the conveying process of the spraying material in real time, and trigger an alarm when the stock of the spraying material is lower than the limit stock value. A low material warning mechanism is established to prevent the situation of interrupting the operation due to insufficient spraying material; and the monitoring of the spraying pipeline is established to ensure the uninterrupted supply of the spraying material and no blockage.

[0079] In the embodiment of the present application, the adaptive spraying robot arm further includes: an emergency stop button, which can quickly pause the system operation once an abnormality is found.

[0080] Step S7: Detect whether the repaired crack area meets the curing standard.

[0081] Among them, based on image processing algorithms, infrared thermal imaging analysis, ultrasonic detection, mechanical property testing algorithms or spectral analysis, it is detected whether the repaired crack area meets the curing standard to ensure complete curing and no damage to the surface of the heater.

[0082] The image processing algorithm uses machine vision technology to take pictures of the sprayed surface under specific wavelength light through a high-definition camera, analyzes the color uniformity, gloss and texture features of the image, and compares it with the image database of the standard cured surface to evaluate the uniformity and curing degree of the coating.

[0083] Infrared thermal imaging analysis uses an infrared thermal imager to monitor the temperature change of the coating during the curing process, identify the temperature rise curve during the coating curing process, and thus judge whether the material reaches the expected curing temperature and curing degree.

[0084] Ultrasonic detection can analyze the change of ultrasonic signals after penetrating the coating, evaluate whether there are voids, delamination or other defects inside the coating, indirectly reflect the curing quality, and is used for areas that require deep curing detection.

[0085] The mechanical property testing algorithm predicts the overall mechanical properties of the coating through the data obtained from tensile, hardness or impact tests, and compares the test results with the successful cases in the historical database to determine whether the predetermined curing standard is met.

[0086] Spectral analysis analyzes the reflection or absorption spectrum of the cured coating by using a spectrometer. Specific chemical bonds will change during the curing process, resulting in changes in spectral characteristics; by analyzing these changes in spectral characteristics, the curing degree and changes in material properties can be quantitatively evaluated.

[0087] In the embodiment of the present application, after step S7: detecting whether the repaired crack area meets the curing standard, it further includes:

[0088] Detect the actual spraying result, compare the difference between the actual spraying result and the expected spraying result, and optimize the deep neural network model.

[0089] That is to say, by comparing the difference between the expected and actual spraying results, the deep neural network model is fine-tuned, the decision-making process is continuously optimized, and a repair method with continuous learning ability is designed.

[0090] In the embodiments of the present application, the materials required for repairing cracks include: high-temperature epoxy resin, ceramic matrix composite material, high-temperature silicone sealant, metal repair agent, high-temperature polyurethane or high-temperature adhesive.

[0091] The above materials are all high-temperature resistant materials, which can adapt to the working temperature while maintaining good mechanical properties and sealing performance.

[0092] The embodiments of the present application also provide a heater crack repair device, which includes: a crack detection module, a data acquisition module, a model training module, an intelligent analysis and decision-making module, a path generation module, an adaptive spraying robotic arm, and a curing effect detection module.

[0093] Among them, the crack detection module is used to obtain the crack image on the surface of the heater through machine vision and transient thermal imaging technology; the data acquisition module is used to obtain the contour information in the crack image through an edge detection algorithm, and at the same time obtain the geometric and thermal characteristics of the crack in the crack image through thermal imaging technology; the model training module is used to train the deep neural network model with a historical crack sample data set; the intelligent analysis and decision-making module is used to input the contour information, geometric characteristics and thermal characteristics into the deep neural network model, and is used to judge whether the crack needs to be repaired through the deep neural network model and calculate the amount of material required to repair the crack; the path generation module is used to combine the contour information, geometric characteristics, thermal characteristics and material amount to obtain the optimal spraying path based on graph theory or genetic algorithm; the path generation module is also used to feedback the optimal spraying path to the adaptive spraying robotic arm to complete the repair of the crack through the adaptive spraying robotic arm; the curing effect detection module is used to detect whether the repaired crack area meets the curing standard.

[0094] In summary, the embodiments of the present application provide a method and device for repairing cracks in a heater. The method for repairing cracks in a heater includes: obtaining a crack image on the surface of the heater based on machine vision and transient thermal imaging technology; obtaining the geometric features and thermal features of the cracks in the crack image, and at the same time obtaining the contour information in the crack image based on an edge detection algorithm; training a deep neural network model using a historical crack sample data set; inputting the contour information, geometric features, and thermal features into the deep neural network model, and determining whether the cracks need to be repaired based on the deep neural network model and calculating the amount of material required for repairing the cracks; combining the contour information, geometric features, thermal features, and material amount to obtain an optimal spraying path based on graph theory or a genetic algorithm; feeding back the optimal spraying path to an adaptive spraying robotic arm to complete the repair of the cracks through the adaptive spraying robotic arm; and detecting whether the repaired crack area meets the curing standard.

[0095] When machine vision and transient thermal imaging technology are used in combination, machine vision is used to identify the appearance features on the surface of the heater, while transient thermal imaging technology is used to detect the thermal characteristics of the surface of the heater. The crack image obtained by combining the two is a two-dimensional or three-dimensional thermal image, which can display the temperature differences of each material layer at different depths, helping to identify hidden defects, discontinuous cracks, or other foreign objects, thus providing more comprehensive quality control.

[0096] An optimal spraying path is obtained through a composite algorithm composed of an edge detection algorithm, a deep neural network model, and graph theory or a genetic algorithm. This optimal spraying path can ensure that all areas are evenly sprayed, while minimizing the spraying time or minimizing material consumption, not only improving the spraying efficiency, but also significantly reducing costs and material waste.

[0097] The embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The same or similar parts among the various embodiments can be referred to each other.

[0098] It should be noted that the phrases such as "one embodiment", "embodiment", "exemplary embodiment", "some embodiments", etc. mentioned in the specification indicate that the described embodiments may include specific features, structures, or characteristics, but not necessarily every embodiment includes such specific features, structures, or characteristics. Moreover, such phrases do not necessarily refer to the same embodiment. In addition, when combining an embodiment to describe a specific feature, structure, or characteristic, it is within the knowledge scope of those skilled in the art to implement such a feature, structure, or characteristic in combination with other embodiments, whether explicitly or implicitly described.

[0099] In general, terms should be understood at least in part in light of their use in the context. For example, at least in part depending on the context, the term "one or more" as used herein can be used to describe any feature, structure, or property in the sense of a singular meaning, or can be used to describe a combination of features, structures, or properties in the sense of a plural meaning. Similarly, at least in part depending on the context, terms such as "a" or "the" can also be understood to convey a singular usage or a plural usage.

[0100] It should be readily understood that the terms "on", "above", and "over" in this disclosure should be interpreted in the broadest manner such that "on" not only means "directly on something", but also includes the meaning of "on something" with intermediate features or layers therebetween, and "above" or "over" not only includes the meaning of "above" or "over something", but can also include the meaning of "above" or "over something" with no intermediate features or layers therebetween (i.e., directly on something).

[0101] In addition, spatial relative terms such as "below", "beneath", "under", "above", "over", etc. may be used herein for ease of description to describe the relationship of one element or feature to another element or feature as shown in the figures. Spatial relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation shown in the figures. The device may have other orientations (rotated 90 degrees or at other orientations), and the spatial relative descriptors used herein may be interpreted accordingly.

[0102] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for repairing a heater crack, characterized in that: include: Based on machine vision and transient thermal imaging technology, the crack image on the heater surface is obtained; The geometric features and thermal features of the crack in the crack image are obtained, and the contour information in the crack image is obtained based on the edge detection algorithm; Use historical crack sample datasets to train deep neural network models; Inputting the contour information, the geometric features, and the thermal features into the deep neural network model, determining whether the crack needs to be repaired based on the deep neural network model, and calculating the amount of material required to repair the crack; Combining the profile information, the geometric features, the thermal features and the material quantity, obtaining an optimal spraying path based on graph theory or genetic algorithm; Feeding back the optimal spraying path to the adaptive spraying robot arm, and completing the repair of the cracks through the adaptive spraying robot arm; Check whether the repaired crack area meets the curing standards.

2. The heater crack repair method according to claim 1, characterized in that: The crack image is a two-dimensional or three-dimensional thermal image, and the thermal image includes a plurality of temperature data, and the plurality of temperature data are temperature data of various material layers at different depths.

3. The heater crack repair method according to claim 1, characterized in that: The method of obtaining a crack image on the heater surface based on machine vision and transient thermal imaging technology includes: emitting heat pulses or light pulses in the millisecond or microsecond range to the heater surface; Capturing the temperature change on the surface of the heater by an infrared detector to obtain thermal imaging data; Tracking changes in the thermal imaging data over time to obtain time series data; The time series data is converted into the crack image based on an image reconstruction method.

4. The heater crack repair method according to claim 1, characterized in that: Before acquiring the contour information in the crack image based on the edge detection algorithm, the method further includes: Based on the image enhancement technology, the noise in the crack image is removed, and the contrast of the crack image is enhanced, so as to optimize the crack image.

5. The heater crack repair method according to claim 1, characterized in that: The geometric features include: crack length, crack width and crack direction; The thermal characteristics include: temperature distribution and heat flow changes.

6. The heater crack repair method according to claim 1, characterized in that: After the detection of whether the repaired crack area meets the curing standard, the method further includes: Detecting actual spraying results, comparing the actual spraying results with expected spraying results, and optimizing the deep neural network model.

7. The heater crack repair method according to claim 1, characterized in that: Before feeding back the optimal spraying path to the adaptive spraying robot arm, the method further includes: The optimal spraying path is simulated based on computer simulation software, and the optimal spraying path is modified according to the simulation results.

8. The heater crack repair method according to claim 1, characterized in that: The adaptive spraying robot arm includes a position sensor, a force sensor and a visual sensor. The position sensor is used to monitor the position of the spray gun on the adaptive spraying robot arm in real time, the force sensor is used to monitor the pressure of the spray gun acting on the heater surface in real time, and the visual sensor is used to monitor the distance between the spray gun and the heater surface in real time.

9. The heater crack repair method according to claim 8, characterized in that: The adaptive spraying robot arm includes an intelligent algorithm control unit, which is used to receive the optimal spraying path and the real-time data fed back by the position sensor, the force sensor and the visual sensor, and dynamically adjust the motion parameters of the adaptive spraying robot arm based on the optimal spraying path and the real-time data; The action parameters include: the position of the spray gun, the movement trajectory of the spray gun, the pressure of the spray gun on the surface of the heater, the distance between the spray gun and the surface of the heater, the output rate of the spray material and the spray angle of the nozzle, etc.; The motion trajectory of the spray gun includes an extreme motion trajectory, and the intelligent algorithm control unit is used to control the motion trajectory of the spray gun to be within the range of the extreme motion trajectory; The pressure exerted by the spray gun on the surface of the heater includes a limit pressure, and the intelligent algorithm control unit is used to control the pressure exerted by the spray gun on the surface of the heater to be less than the limit pressure.

10. A heater crack repair device, characterized in that: include: Crack detection module, data acquisition module, model training module, intelligent analysis and decision module, path generation module, adaptive spraying robot arm and curing effect detection module; The crack detection module is used to obtain a crack image on the heater surface through machine vision and transient thermal imaging technology; The data acquisition module is used to acquire contour information in the crack image by using an edge detection algorithm, and to acquire geometric features and thermal features of the crack in the crack image by using thermal imaging technology; The model training module is used to train a deep neural network model using a historical crack sample data set; The intelligent analysis and decision module is used to input the contour information, the geometric features and the thermal features into the deep neural network model, and to determine whether the crack needs to be repaired through the deep neural network model, and calculate the amount of material required to repair the crack; The path generation module is used to obtain an optimal spraying path based on graph theory or genetic algorithm by combining the contour information, the geometric features, the thermal features and the material amount; The path generation module is also used to feed back the optimal spraying path to the adaptive spraying robot arm, so that the cracks are repaired by the adaptive spraying robot arm; The curing effect detection module is used to detect whether the repaired crack area meets the curing standard.

Images