An intelligent monitoring welding system for hot work

Abstract

The present invention relates to the field of welding technology, and specifically discloses an intelligent monitoring welding system for hot work operations. The present invention utilizes a high-speed industrial camera to collect the surface profile of the electrode head in real time, combines an image processing algorithm to automatically identify the flatness and surface defects of the electrode head, and triggers an early warning when it deviates from the set threshold. Multi-dimensional simultaneous monitoring can ensure that there are no blind spots, and a non-contact temperature sensor is arranged in the back area of ​​the electrode head to monitor the temperature hot spot distribution in real time, and the accumulated damage of the electrode head is jointly evaluated to determine whether the electrode head is recommended to continue to be used. When it is not recommended to continue to use, an early warning is triggered; the present invention realizes all-round perception and precise maintenance of the status of the resistance spot welding electrode head through dual-mode online monitoring of three-dimensional profile scanning and thermal infrared temperature measurement, combined with a data linkage early warning mechanism.

Description

Technical Field

[0001] The present invention relates to the field of welding technology, and more particularly to an intelligent monitoring welding system for hot work. Background Art

[0002] Resistance spot welding is a highly adaptable and productive material joining technology. Existing literature (Zhou You. Research on Signal Monitoring and Quality Assessment of Resistance Spot Welding Process [D]. Jiangxi University of Science and Technology, 2024. DOI: 10.27176 / d.cnki.gnfyc.2024.000408.) designed a resistance spot welding process signal monitoring system to acquire welding process signals, studied the mapping relationship between welding process signals and spot welding quality, and established a resistance spot welding quality assessment model. This model enables classification and identification of resistance spot welding quality and strength prediction, improving the efficiency of spot welding quality assessment. During the spot welding process, the electrode tip is subjected to high temperature, high pressure, and repeated mechanical friction. Due to the thermomechanical coupling, the material gradually undergoes plastic flow and structural softening, leading to two typical changes: first, the electrode tip undergoes irreversible changes in geometry, manifested by the gradual blunting of previously sharp corners, an increase in radius, and a loss of flatness in the stress-bearing surface; second, the microstructure and hardness of the electrode tip surface and near-surface layers undergo annealing and softening, resulting in a decrease in the material's yield strength, wear resistance, and thermal conductivity. These changes are small, but have a very sensitive impact on welding quality - the increase in local contact resistance makes the welding heat input unevenly distributed and the weld nugget generation unstable, which may eventually lead to a decrease in weld strength and defects.

[0003] Currently, most production lines rely on replacing or regrinding electrode tips based on "number of uses," "forming cycles," or "number of welds." This experience-based, regular maintenance method fails to reflect the actual degree of wear and deformation of the electrode tips. On the one hand, factors such as different batches of materials, welding parameters, and ambient temperature and humidity can cause significant differences in electrode tip wear rates. On the other hand, even with the same cumulative number of welds, individual electrode tips can be more prone to early failure due to load distribution or occasional overheating. As a result, some electrode tips continue to be used despite being severely deformed during their lifespan, leading to fluctuations in product quality. Meanwhile, some electrode tips remain in good condition within their maintenance cycle but are prematurely scrapped, wasting costs. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides an intelligent monitoring welding system for hot work operations. Through multi-modal fusion of high-speed industrial cameras and infrared temperature measurement, dual feature extraction of images and temperature gradients, and quantitative evaluation of cumulative damage, it is used to solve the problem that the experience-based regular maintenance method cannot reflect the actual wear and deformation degree of the electrode head, and realizes real-time, accurate, and traceable status monitoring and predictive maintenance of resistance spot welding electrode heads.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An intelligent monitoring welding system for hot work, including a 3D contour scanning module, a thermal infrared temperature measurement imaging module and a data linkage warning module; the 3D contour scanning module is used to use a high-speed industrial camera to collect the surface contour of the electrode head in real time, and automatically identify the flatness and surface defects of the electrode head in combination with the image processing algorithm, and trigger an early warning when it deviates from the set threshold; the 3D contour scanning module includes a high-speed industrial camera, an image data processing unit, an electrode head surface contour recognition unit, a contour re-recognition unit and an early warning trigger unit; the high-speed industrial camera is used to regularly obtain the surface contour image of the electrode head and use it as the first contour image data, and the surface contour image of the electrode head before use is used as the second contour image data; the image data processing unit is used to process the surface contour image of the electrode head and the second contour image data. A contour image data and a second contour image data are grayed, denoised and smoothed, and contour edges are extracted from the preprocessed first contour image data and the second contour image data to obtain corresponding contour points. Arc fitting is performed based on the contour points of the first contour image data to obtain a first contour curve, and arc fitting is performed based on the contour points of the second contour image data to obtain a second contour curve; an electrode head surface contour recognition unit is used to identify the flatness of the electrode head surface according to the first contour curve and the second contour curve; a contour re-recognition unit is used to identify whether there are defects on the electrode head surface based on changes in the flatness of the electrode head surface; and an early warning triggering unit is used to trigger an early warning when the contour re-recognition unit identifies that there are defects on the electrode head surface.

[0007] As a further solution of the present invention, the electrode tip surface contour recognition unit is used to identify the surface flatness of the electrode tip according to the first contour curve and the second contour curve, specifically:

[0008] Obtaining a first contour curve and a second contour curve, taking out the height values ​​corresponding to the first contour curve and the second contour curve at each identical lateral position, and calculating a local height difference between the two at the lateral position, where the local height difference represents a local concavity and convexity at the lateral position relative to the surface state of the electrode head;

[0009] The average value is calculated based on the local height difference, and the degree of dispersion of the local height difference around the average value at each lateral position is calculated to reflect the uniformity of the undulation of the entire electrode head end surface. If the degree of dispersion exceeds the preset discrete threshold, the surface flatness of the electrode head is poor;

[0010] The absolute value of the local height difference at each lateral position is calculated, and the maximum absolute value is selected to reflect the most serious single-point concave-convex situation on the end face of the electrode head. If the maximum absolute value exceeds the preset height difference threshold, the surface flatness of the electrode head is poor.

[0011] As a further solution of the present invention, the contour re-identification unit is used to identify whether there is a defect on the surface of the electrode head based on the change in the flatness of the electrode head surface, specifically:

[0012] When the surface flatness of the electrode head is detected to be poor, the horizontal coordinate position is marked as an abnormal position point to generate an abnormal index set, and the adjacent abnormal position points are divided into several abnormal segments according to the continuity within the abnormal index set, and the starting and ending horizontal coordinate ranges of each abnormal segment are recorded;

[0013] For each abnormal segment, return to the first contour image data, cut out the image area corresponding to the abnormal segment as the defect assessment area, and apply adaptive contrast enhancement and bilateral filtering image enhancement algorithms to the defect assessment area;

[0014] Continuous offset detection is used to identify pits or protrusions in the defect evaluation area after image enhancement to determine whether there is a defect.

[0015] As a further solution of the present invention, continuous offset detection is used to identify pits or protrusions in the defect evaluation area after image enhancement to determine whether a defect exists, specifically:

[0016] Compare the shape of the local height difference curve in the defect area to be evaluated and identify continuous unilateral deviations, that is, if the local height difference is continuously greater than the positive threshold or less than the negative threshold for more than k horizontal coordinate positions, then mark the defect area to be evaluated as defective;

[0017] Adaptive binarization and skeletonization are performed on the defect area to be evaluated to extract the slender linear structure. If the length of the extracted slender linear structure exceeds the preset minimum crack length and the width of the slender linear structure is less than the maximum crack width, the defect area to be evaluated is marked as defective.

[0018] As a further solution of the present invention, the thermal infrared temperature measurement and imaging module includes a non-contact temperature sensor, a temperature hotspot distribution monitoring unit, and a thermal cumulative damage assessment unit;

[0019] The non-contact temperature sensor is used to obtain the temperature distribution of the electrode head in real time;

[0020] The temperature hotspot distribution monitoring unit is used to extract temperature samples from each row according to the horizontal coordinate position range of the defect area to be evaluated, calculate the temperature differential gradient of adjacent points, and calculate the first temperature distribution data based on the temperature differential gradient; the first temperature distribution data includes an average gradient and a maximum gradient.

[0021] As a further solution of the present invention, the thermal cumulative damage assessment unit is used to obtain the first temperature distribution data of all rows, obtain the ratio of the average gradient and the maximum gradient, obtain the ratio of all rows at adjacent moments, and evaluate the cumulative damage of the electrode head by comparing the difference in the row ratios at adjacent moments. If the difference exceeds a preset threshold, it is determined that the cumulative damage of the electrode head is serious and it is not recommended to continue using it; if the difference does not exceed the preset threshold, it is continued to be used.

[0022] As a further solution of the present invention, the data linkage warning module is used to trigger a warning when the difference between the row ratios at adjacent moments exceeds a preset threshold.

[0023] The technical effects and advantages of the intelligent monitoring welding system for hot work of the present invention are as follows:

[0024] The present invention uses a high-speed industrial camera to collect the surface profile of the electrode head in real time, and combines it with an image processing algorithm to automatically identify the flatness and surface defects of the electrode head. When it deviates from the set threshold, it triggers an early warning. Multi-dimensional simultaneous monitoring can ensure that there are no blind spots. A non-contact temperature sensor is arranged in the back area of ​​the electrode head to monitor the temperature hot spot distribution in real time. The accumulated damage of the electrode head is jointly evaluated to determine whether the electrode head is recommended for continued use. If continued use is not recommended, an early warning is triggered.

[0025] The present invention realizes all-round perception and precise maintenance of the status of resistance spot welding electrode heads through dual-modal online monitoring of three-dimensional contour scanning and thermal infrared temperature measurement, combined with a data linkage early warning mechanism; the present invention realizes real-time, precise, and traceable status monitoring and predictive maintenance of resistance spot welding electrode heads through multi-modal fusion of high-speed industrial cameras and infrared temperature measurement, dual feature extraction of images and temperature gradients, and quantitative evaluation of cumulative damage, which significantly improves the stability of welding quality, reduces maintenance costs, and extends the life of electrodes. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A schematic structural diagram of an intelligent monitoring welding system for hot work provided by the present invention;

[0027] Figure 2 The thermal map of the electrode tip temperature hotspot distribution provided by the present invention;

[0028] Figure 3 This is a distribution diagram of the local height difference of the electrode head end face provided by the present invention. DETAILED DESCRIPTION

[0029] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the accompanying drawings. Obviously, the technical solutions described are only part of the present invention, not the entire invention. Based on the technical solutions of the present invention, all other technical solutions obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0030] Example 1

[0031] like Figure 1-Figure 3 FIG. 1 is a schematic diagram of a structure of an intelligent monitoring welding system for hot work, comprising a three-dimensional profile scanning module, a thermal infrared temperature measurement and imaging module, and a data linkage warning module; the three-dimensional profile scanning module is connected to the thermal infrared temperature measurement and imaging module, and the thermal infrared temperature measurement and imaging module is connected to the data linkage warning module;

[0032] The 3D contour scanning module is used to collect the surface contour of the electrode head in real time using a high-speed industrial camera, and automatically identify the flatness and surface defects of the electrode head in combination with the image processing algorithm, triggering an early warning when it deviates from the set threshold.

[0033] The thermal infrared temperature measurement and imaging module is used to arrange non-contact temperature sensors on the back area of ​​the electrode head to monitor the temperature hotspot distribution in real time, and to jointly evaluate the cumulative damage of the electrode head to determine whether the electrode head is recommended to continue to use.

[0034] When the electrode head is determined to be not recommended for continued use, the data linkage warning module triggers a warning.

[0035] Specifically, the three-dimensional contour scanning module includes a high-speed industrial camera, an image data processing unit, an electrode head surface contour recognition unit, a contour re-recognition unit and an early warning trigger unit; the high-speed industrial camera is connected to the image data processing unit, the image data processing unit is connected to the electrode head surface contour recognition unit, the electrode head surface contour recognition unit is connected to the contour re-recognition unit, and the contour re-recognition unit is connected to the early warning trigger unit.

[0036] The high-speed industrial camera is used to regularly obtain the surface profile image of the electrode head and use it as the first profile image data, and the surface profile image of the electrode head before use is used as the second profile image data.

[0037] The image data processing unit is used to grayscale, denoise and smooth the first contour image data and the second contour image data, and perform contour edge extraction on the preprocessed first contour image data and the second contour image data to obtain corresponding contour points, perform arc fitting based on the contour points of the first contour image data to obtain a first contour curve, and perform arc fitting based on the contour points of the second contour image data to obtain a second contour curve.

[0038] The electrode head surface contour recognition unit is used to recognize the surface flatness of the electrode head according to the first contour curve and the second contour curve.

[0039] The contour re-identification unit is used to identify whether there are defects on the surface of the electrode head based on the change in the flatness of the electrode head surface.

[0040] The early warning triggering unit is used to trigger an early warning when the contour re-identification unit identifies that there is a defect on the surface of the electrode head.

[0041] By collecting deformation and temperature information in parallel through 3D contour scanning and thermal infrared temperature measurement, early defects such as tiny passivation, pits or cracks on the electrode head can be immediately captured, avoiding the risk of batch rework caused by offline detection lag. Weighted fusion of multiple indicators such as geometric flatness, local height difference, temperature hot spots and gradients can not only reduce the false alarm rate of a single sensor, but also more accurately reflect the true state of the electrode head, ensuring the accuracy of the early warning. Based on the damage rate model of temperature field and gradient, the accumulation of thermal fatigue is converted into a quantifiable amount of damage, and compared with the threshold to determine whether to continue using it, so as to avoid the shutdown of the entire production line or rework of subsequent products due to excessive wear or sudden failure of the electrode head due to overheating.

[0042] Specifically, the electrode head surface contour recognition unit is used to identify the surface flatness of the electrode head according to the first contour curve and the second contour curve, specifically:

[0043] Obtaining a first contour curve and a second contour curve, taking out the height values ​​corresponding to the first contour curve and the second contour curve at each identical lateral position, and calculating a local height difference between the two at the lateral position, where the local height difference represents a local concavity and convexity at the lateral position relative to the surface state of the electrode head;

[0044] The average value is calculated based on the local height difference, and the degree of dispersion of the local height difference around the average value at each lateral position is calculated to reflect the uniformity of the undulation of the entire electrode head end surface. If the degree of dispersion exceeds the preset discrete threshold, the surface flatness of the electrode head is poor;

[0045] The absolute value of the local height difference at each lateral position is calculated, and the maximum absolute value is selected to reflect the most serious single-point concave-convex situation on the end face of the electrode head. If the maximum absolute value exceeds the preset height difference threshold, the surface flatness of the electrode head is poor.

[0046] Specifically, the contour re-identification unit is used to identify whether there are defects on the electrode head surface based on the change in the flatness of the electrode head surface, specifically:

[0047] When poor surface flatness of the electrode head is detected, the horizontal coordinate position is marked as an abnormal position point to generate an abnormal index set, and the adjacent abnormal position points are divided into several abnormal segments according to the continuity within the abnormal index set, and the starting and ending horizontal coordinate ranges of each abnormal segment are recorded.

[0048] For each abnormal segment, return to the first contour image data, cut out the image area corresponding to the abnormal segment (plus appropriate redundant margins) as the defect assessment area, and apply adaptive contrast enhancement and bilateral filtering image enhancement algorithms to the defect assessment area;

[0049] Continuous offset detection is used to identify pits or protrusions in the defect evaluation area after image enhancement to determine whether there is a defect.

[0050] Specifically, continuous offset detection is used to identify pits or protrusions in the defect evaluation area after image enhancement to determine whether a defect exists, specifically:

[0051] Compare the shape of the local height difference curve in the defect area to be evaluated and identify continuous unilateral deviations, that is, if the local height difference is continuously greater than the positive threshold or less than the negative threshold for more than k horizontal coordinate positions, then mark the defect area to be evaluated as defective;

[0052] Adaptive binarization and skeletonization are performed on the defect area to be evaluated to extract the slender linear structure. If the length of the extracted slender linear structure exceeds the preset minimum crack length and the width of the slender linear structure is less than the maximum crack width, the defect area to be evaluated is marked as defective.

[0053] The point-by-point calculation of the height difference between the first and second contour curves transforms "flatness" from an empirical judgment into a quantifiable, micron-level value. This includes both overall dispersion metrics and single-point maximum deviation, accurately characterizing the endface condition. Flatness deviation is detected as soon as the standard deviation or maximum height difference exceeds a threshold, significantly earlier than the human eye or indirect indicators such as weld force and current become ineffective. Anomaly index collection and abnormal segment location pinpoint early-stage micro-pits and protrusions, preventing subsequent more serious faults such as cracks and spatter. Image acquisition → difference calculation → segmentation → ROI extraction → image enhancement → defect detection enable millisecond-level online monitoring. This eliminates the need for machine downtime and disassembly inspection, nor does it rely on manual visual inspection, significantly improving inspection efficiency. Based on the difference between dual contour curves, the system is unaffected by external factors such as lighting and workpiece surface color. Abnormal segments are segmented based on continuity, discarding isolated noise points and reducing false alarm rates. The system can identify pits / protrusions corresponding to continuous unilateral offsets and capture long and narrow cracks through skeletonization. It can also be expanded to distinguish various surface defects such as ablation spots and spatter pits. Outputs the horizontal coordinate range of each abnormal section, allowing repairs to be performed only within that specific section, eliminating the need for complete head removal and grinding, saving time and cost. Flatness and defect results can be directly used as ROI coordinates for subsequent thermal infrared or acoustic emission data analysis, helping to precisely locate thermal fatigue hotspots and crack propagation, and building a multimodal, highly reliable fault diagnosis system.

[0054] Specifically, the thermal infrared temperature measurement and imaging module includes a non-contact temperature sensor, a temperature hotspot distribution monitoring unit, and a thermal cumulative damage assessment unit; the non-contact temperature sensor is connected to the temperature hotspot distribution monitoring unit, and the temperature hotspot distribution monitoring unit is connected to the thermal cumulative damage assessment unit;

[0055] The non-contact temperature sensor is used to obtain the temperature distribution of the electrode head in real time;

[0056] The temperature hotspot distribution monitoring unit is used to extract temperature samples from each row according to the horizontal coordinate position range of the defect area to be evaluated, calculate the temperature differential gradient of adjacent points, and calculate first temperature distribution data based on the temperature differential gradient; the first temperature distribution data includes an average gradient and a maximum gradient;

[0057] The thermal cumulative damage assessment unit is used to obtain the first temperature distribution data of all rows, obtain the ratio of the average gradient and the maximum gradient, obtain the ratio of all rows at adjacent moments, and evaluate the cumulative damage of the electrode head by comparing the difference in the row ratios at adjacent moments. If the difference exceeds the preset threshold, the cumulative damage to the electrode head is serious at this time and it is not recommended to continue using it; otherwise, it is recommended to continue using it.

[0058] By calculating the temperature differential gradient of adjacent points for each row of temperature samples and extracting the average gradient and the maximum gradient, it is possible to accurately capture tiny temperature mutations on the surface of the electrode head, much earlier than the peak temperature exceeds the limit, thereby discovering hidden damage earlier; the change of the gradient ratio of each row over time is included in the evaluation, and the difference of the ratio at adjacent moments is used to reflect the thermal fatigue accumulation rate, which can quantify the development trend of the cumulative damage of the electrode head and realize true predictive maintenance; sampling calculation is only performed within the horizontal section where the defect is to be evaluated, avoiding time-consuming processing of the entire thermal image field, significantly reducing the amount of calculation and improving the online response speed; the gradient ratio and its change amount correspond to the preset threshold one by one, the alarm logic is simple and clear, and easy to configure and adjust; the gradient ratio can be used to generate an alarm. In response to sudden increases, the cooling cycle or welding current-time parameters are dynamically adjusted, and the welding process is automatically optimized while monitoring, taking into account both quality and efficiency. Only a non-contact hotspot temperature sensor and a set of edge computing units are needed to complete the closed loop from hotspot capture → gradient analysis → damage assessment → early warning triggering, without the need for an additional complex sensor network. The gradient ratio is only a numerical result and can be directly coupled with the defect coordinates of the 3D contour scanning module to jointly improve the overall monitoring accuracy and reliability. By early and quantitatively evaluating thermal damage and issuing a "not recommended for further use" warning when the threshold is exceeded, sudden failure caused by overheating of the electrode head and the subsequent large-scale shutdown and rework can be avoided, reducing maintenance costs and production risks.

[0059] Figure 2 This is a heat map of the electrode tip temperature hotspot distribution provided by the present invention. The X-axis and Y-axis correspond to the horizontal and vertical position (in mm) of the electrode tip surface, respectively. The color gradient from blue (low temperature) to red (high temperature) represents the real-time temperature (in °C) at the corresponding point. The temperature of the point marked in the figure is 77.3°C, which is the local hotspot within the selected ROI (coinciding with the defect area).

[0060] Figure 3 This is a distribution diagram of the local height differences on the end face of the electrode tip provided by the present invention; the horizontal axis represents 20 equally spaced transverse locations on the end face within the range of 0mm–19mm, and the vertical axis represents the height difference (in mm) at each location relative to the reference profile of the new electrode. Positive values ​​are marked upward, indicating a "convexity" at that location; negative values ​​are marked downward, indicating a "concaveness." Clear protrusions (with a height difference of 0.7mm to 1.0mm) are clearly visible at transverse locations 7, 9, 10, 12, 16, and 19; depressions with depths of -0.8mm to -0.5mm occur at locations 3, 5, 14, 17, and 18, all exceeding the preset flatness tolerances, indicating that these areas require special inspection or refining.

[0061] Example 2

[0062] A car manufacturing plant currently has a body assembly line equipped with 12 resistance spot welders per line, used to weld door and side panels. Traditionally, each welding gun's electrode tip is manually replaced or sharpened after 3,000 welds. This often leads to premature failure of individual electrode tips, resulting in inconsistent weld quality and requiring production line rework, impacting production capacity.

[0063] An intelligent monitoring welding system for hot work was applied to the aforementioned electrode tip inspection: the welding operator turned on the machine at 8:00 AM, and the system simultaneously began data acquisition. After every 500 welds, the 3D contour module automatically captured a "first contour image." The image data processing unit calculated the contour curve at that point. At the 1500th point, the system detected that the standard deviation of the electrode tip flatness relative to the initial contour was 0.055>0.05, and the maximum deviation was 0.12mm>0.10mm, indicating a "degraded flatness." The contour re-identification unit marked the abnormal index set and divided it into two abnormal segments. After corresponding image enhancement, a slight pit was found in the second segment, confirming the presence of a defect and immediately triggering an early warning.

[0064] The system maps the defective area to be evaluated on the infrared screen based on the horizontal coordinates of the defective section reported. At that time, the infrared monitoring showed that the highest temperature in the defective area to be evaluated was 330 The average temperature is 280 Ten horizontal samples are taken as given in Table 1, and the average gradient is calculated to be 4.2 and the maximum gradient is 9.1.

[0065] Table 1 Horizontal sample example table

[0066]

[0067] The embodiment of the present invention utilizes a high-speed industrial camera to collect the surface contour of the electrode head in real time, combines the image processing algorithm to automatically identify the flatness and surface defects of the electrode head, and triggers an early warning when it deviates from the set threshold. Multi-dimensional simultaneous monitoring can ensure that there are no blind spots, and a non-contact temperature sensor is arranged in the back area of ​​the electrode head to monitor the temperature hotspot distribution in real time, and jointly evaluate whether the cumulative damage of the electrode head is recommended to continue using it. When it is not recommended to continue using it, a warning is triggered; the present invention realizes all-round perception and precise maintenance of the state of the resistance spot welding electrode head through dual-modal online monitoring of three-dimensional contour scanning and thermal infrared temperature measurement, combined with a data linkage early warning mechanism; the present invention realizes real-time, precise, and traceable state monitoring and predictive maintenance of the resistance spot welding electrode head through multi-modal fusion of high-speed industrial cameras and infrared temperature measurement, dual feature extraction of images and temperature gradients, and quantitative evaluation of cumulative damage, significantly improving welding quality stability, reducing maintenance costs, and extending electrode life.

[0068] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0069] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An intelligent monitoring welding system for hot work, comprising a three-dimensional contour scanning module, a thermal infrared temperature measurement and imaging module, and a data linkage warning module; characterized in that: The 3D profile scanning module uses a high-speed industrial camera to collect the surface profile of the electrode head in real time. It combines image processing algorithms to automatically identify the flatness and surface defects of the electrode head and triggers an early warning when it deviates from the set threshold. The 3D contour scanning module includes a high-speed industrial camera, an image data processing unit, an electrode head surface contour recognition unit, a contour re-recognition unit, and an early warning trigger unit; The high-speed industrial camera is used to periodically obtain a surface profile image of the electrode head and use it as first profile image data, and the surface profile image of the electrode head before use is used as second profile image data; The image data processing unit is used to perform grayscale, denoising, and smoothing processing on the first contour image data and the second contour image data, and perform contour edge extraction on the pre-processed first contour image data and the second contour image data to obtain corresponding contour points, perform arc fitting based on the contour points of the first contour image data to obtain a first contour curve, and perform arc fitting based on the contour points of the second contour image data to obtain a second contour curve; The electrode head surface contour recognition unit is used to recognize the surface flatness of the electrode head according to the first contour curve and the second contour curve; The contour re-identification unit is used to identify whether there are defects on the electrode head surface based on the change in the flatness of the electrode head surface; The early warning triggering unit is used to trigger an early warning when the contour re-identification unit identifies that there is a defect on the surface of the electrode head.

2. The intelligent monitoring welding system for hot work according to claim 1, characterized in that: The electrode head surface contour recognition unit is used to identify the surface flatness of the electrode head according to the first contour curve and the second contour curve, specifically: Obtaining a first contour curve and a second contour curve, taking out the height values ​​corresponding to the first contour curve and the second contour curve at each identical lateral position, and calculating a local height difference between the two at the lateral position, where the local height difference represents a local concavity and convexity at the lateral position relative to the surface state of the electrode head; The average value is calculated based on the local height difference, and the degree of dispersion of the local height difference around the average value at each lateral position is calculated to reflect the uniformity of the undulation of the entire electrode head end surface. If the degree of dispersion exceeds the preset discrete threshold, the surface flatness of the electrode head is poor; The absolute value of the local height difference at each lateral position is calculated, and the maximum absolute value is selected to reflect the most serious single-point concave-convex situation on the end face of the electrode head. If the maximum absolute value exceeds the preset height difference threshold, the surface flatness of the electrode head is poor.

3. The intelligent monitoring welding system for hot work according to claim 2, characterized in that: The contour re-identification unit is used to identify whether there are defects on the electrode head surface based on the change in the flatness of the electrode head surface. Specifically: When the surface flatness of the electrode head is detected to be poor, the horizontal coordinate position is marked as an abnormal position point to generate an abnormal index set, and the adjacent abnormal position points are divided into several abnormal segments according to the continuity within the abnormal index set, and the starting and ending horizontal coordinate ranges of each abnormal segment are recorded; For each abnormal segment, return to the first contour image data, cut out the image area corresponding to the abnormal segment as the defect assessment area, and apply adaptive contrast enhancement and bilateral filtering image enhancement algorithms to the defect assessment area; Continuous offset detection is used to identify pits or protrusions in the defect evaluation area after image enhancement to determine whether there is a defect.

4. The intelligent monitoring welding system for hot work according to claim 3, characterized in that: Continuous offset detection is used to identify pits or protrusions in the defect evaluation area after image enhancement to determine whether there is a defect. Specifically: Compare the shape of the local height difference curve in the defect area to be evaluated and identify continuous unilateral deviations, that is, if the local height difference is continuously greater than the positive threshold or less than the negative threshold for more than k horizontal coordinate positions, then mark the defect area to be evaluated as defective; Adaptive binarization and skeletonization are performed on the defect area to be evaluated to extract the slender linear structure. If the length of the extracted slender linear structure exceeds the preset minimum crack length and the width of the slender linear structure is less than the maximum crack width, the defect area to be evaluated is marked as defective.

5. The intelligent monitoring welding system for hot work according to claim 3, characterized in that: The thermal infrared temperature measurement and imaging module includes a non-contact temperature sensor, a temperature hotspot distribution monitoring unit, and a thermal cumulative damage assessment unit; The non-contact temperature sensor is used to obtain the temperature distribution of the electrode head in real time; The temperature hotspot distribution monitoring unit is used to extract temperature samples from each row according to the horizontal coordinate position range of the defect area to be evaluated, calculate the temperature differential gradient of adjacent points, and calculate the first temperature distribution data based on the temperature differential gradient; The first temperature distribution data includes an average gradient and a maximum gradient.

6. The intelligent monitoring welding system for hot work according to claim 5, characterized in that: The thermal cumulative damage assessment unit is used to obtain the first temperature distribution data of all rows, obtain the ratio of the average gradient to the maximum gradient, obtain the ratio of all rows at adjacent moments, and evaluate the cumulative damage of the electrode head by judging whether the difference between the row ratios at adjacent moments exceeds a preset threshold.

7. The intelligent monitoring welding system for hot work according to claim 6, characterized in that: The data linkage warning module is used to trigger a warning when the difference between the row ratios at adjacent moments exceeds a preset threshold.

Images