Laser welding system control method for metal fittings

Abstract

The invention relates to the technical field of laser welding, in particular to a laser welding system control method for metal accessories. The method comprises the following steps that automobile industry metal fitting data are obtained, welding demand analysis is conducted according to the automobile industry metal fitting data, and therefore welding demand data are obtained; laser welding feature extraction and resistance spot welding feature extraction are carried out according to the welding demand data, so that laser welding data and resistance spot welding data are obtained; carrying out welding simulation according to the laser welding data and the resistance spot welding data so as to obtain welding simulation data; welding form feature extraction and welding strength feature extraction are carried out according to the welding simulation data, so that welding form data and welding strength data are obtained; and carrying out welding defect detection according to the welding form data and the welding strength data so as to obtain welding defect data. Automatic adjustment and accurate control of the laser welding process are achieved based on the laser welding technology.

Description

Technical Field

[0001] The invention relates to the technical field of laser welding, and in particular to a laser welding system control method for metal accessories. Background Art

[0002] Traditional control methods mostly rely on static welding parameter settings, such as laser power, welding speed, focus position, etc. Although these parameters can meet the needs of conventional welding to a certain extent, it is difficult to dynamically adjust according to real-time feedback when facing complex welding tasks or different metal materials. Especially when dealing with complex geometric shapes or irregular metal accessories, traditional methods fail to fully consider the changes in the welding environment, such as welding position, material surface state and other factors, resulting in poor stability of the welding process and easy welding defects such as pores, cracks or overmelting. Traditional methods have great limitations in real-time monitoring and feedback adjustment of the molten pool during welding. The morphology, size, temperature distribution, etc. of the molten pool are important factors affecting the welding quality, while most of the existing control methods only focus on the basic parameters in the welding process (such as laser power and welding speed). The lack of accurate monitoring and adjustment means for the changes in the morphology of the molten pool makes it impossible to timely detect and correct welding defects such as uneven molten pool and insufficient penetration, resulting in unstable welding quality and failure to meet high-precision requirements. Summary of the invention

[0003] Based on this, it is necessary for the present invention to provide a laser welding system control method for metal accessories to solve at least one of the above technical problems.

[0004] To achieve the above object, a laser welding system control method for metal accessories includes the following steps:

[0005] Step S1: Obtaining automobile industry metal parts data, and performing welding demand analysis based on the automobile industry metal parts data, thereby obtaining welding demand data;

[0006] Step S2: performing laser welding feature extraction and resistance spot welding feature extraction according to the welding requirement data, thereby obtaining laser welding data and resistance spot welding data; performing welding simulation according to the laser welding data and the resistance spot welding data, thereby obtaining welding simulation data;

[0007] Step S3: extracting welding morphology features and welding strength features according to the welding simulation data, thereby obtaining welding morphology data and welding strength data; performing welding defect detection according to the welding morphology data and welding strength data, thereby obtaining welding defect data;

[0008] Step S4: acquiring laser welding system data; optimizing laser welding parameters of the laser welding system data according to the welding defect data, thereby obtaining laser welding optimization parameters;

[0009] Step S5: adjusting the laser welding process of the welding simulation data according to the laser welding optimization parameters, thereby obtaining laser welding process adjustment data, and uploading the data to the laser welding system to execute the laser welding process control task.

[0010] The present invention can provide customized welding solutions according to different welding tasks and the characteristics of metal materials by comprehensively analyzing the welding requirements of metal accessories in the automotive industry. By acquiring metal accessories data and performing welding requirements analysis, the specific welding requirements of each metal accessory can be accurately understood, thereby providing data support for the design and optimization of subsequent welding processes. On this basis, by extracting the characteristics of laser welding and resistance spot welding, combined with welding simulation, specific welding simulation data can be obtained, providing a scientific basis for the comparison and selection of different welding processes. By extracting the welding morphology and welding strength characteristics of the welding simulation data, the defects generated in the welding process can be fully grasped, including but not limited to pores, cracks, overmelting and other problems. Combined with welding defect detection technology, the defect area can be identified and located in real time during the welding process, thereby providing real-time data support for the control and optimization of welding quality. This link ensures that welding defects can be discovered in time, and provides accurate guidance for subsequent welding parameter optimization, effectively reducing quality problems caused by welding defects. After welding defect detection, accurate welding parameter optimization can be performed according to welding defect data through analysis and optimization of laser welding system data. The acquisition of laser welding optimization parameters makes the welding process more accurate and efficient. Compared with traditional static parameter settings, the optimized welding parameters can be dynamically adjusted to adapt to changes in the actual welding environment, avoiding defects caused by unstable welding environment, thereby greatly improving the stability and consistency of the welding process. The welding process is adjusted using optimized laser welding parameters, and the adjusted process is uploaded to the laser welding system for control, which not only improves the welding quality, but also reduces production problems caused by unstable processes. Through process adjustment, various changes in the welding process can be adapted in real time, avoiding the problems caused by the lack of flexibility and real-time feedback in traditional methods. Overall, the implementation of this method can significantly improve the stability, accuracy and efficiency of the welding process, and provide a practical solution for high-quality welding of metal accessories.

[0011] Optionally, step S1 specifically includes:

[0012] Step S11: acquiring automobile industry metal parts data, and performing welding area feature extraction based on the automobile industry metal parts data, thereby obtaining automobile industry metal parts welding area data;

[0013] Step S12: performing narrow area identification according to the welding area data of automobile industry metal parts, thereby obtaining the welding narrow area data of automobile industry metal parts;

[0014] Step S13: performing multi-curvature surface analysis on the welding narrow area data of automobile industry metal accessories, thereby obtaining multi-curvature surface data of automobile industry metal accessories;

[0015] Step S14: performing spatial complexity evaluation according to the multi-curvature surface data of automobile industry metal accessories, thereby obtaining the spatial complexity of automobile industry metal accessories;

[0016] Step S15: Perform welding method requirement adaptation evaluation according to the spatial complexity of automobile industry metal parts, so as to obtain welding requirement data.

[0017] The present invention can accurately identify and define the welding area by acquiring the data of metal accessories in the automotive industry and extracting the welding area features based on these data, thus laying a foundation for the subsequent optimization of the welding process. Through the processing of the welding area data, the key areas of the welding task can be effectively defined, ensuring that the welding process can accurately cover the areas that need to be welded, and avoiding the problems of missed welding or excessive welding. In the process of identifying the narrow area, the method can further screen out the complex areas or the parts that are particularly difficult to weld, and customize special welding schemes for these areas, thereby improving the welding accuracy and stability, and reducing the defects caused by irregular shapes. For the narrow welding area, after performing multi-curvature surface analysis, the geometric characteristics of the area can be more comprehensively grasped, and the influence of the curvature change of the surface on the welding process can be understood. This analysis can effectively predict and avoid welding problems caused by surface geometric changes, such as uneven molten pool, insufficient penetration, etc. Through further spatial complexity evaluation, the value of the surface complexity of the accessories can be quantified, providing data support for the selection of welding technology, ensuring that the welding parameters match the geometric characteristics of the accessories, and avoiding the influence of inappropriate welding methods on the welding quality. According to the spatial complexity data, the welding method can be accurately evaluated, and the most suitable welding method can be selected, thereby optimizing the entire welding process. Through this precise demand adaptation, the welding process can be dynamically adjusted according to the actual conditions of metal accessories, improving the efficiency, stability and quality of the welding process.

[0018] Optionally, step S13 is specifically:

[0019] Step S131: performing point cloud neighborhood calculation on the narrow area data of welding of automobile industry metal parts, thereby obtaining point cloud neighborhood data;

[0020] Step S132: performing weighted averaging according to the point cloud neighborhood data to obtain point cloud neighborhood weighted average data, and updating the point cloud neighborhood data according to the point cloud neighborhood weighted average data to obtain point cloud neighborhood update data;

[0021] Step S133: constructing a smooth surface model of the metal accessory according to the point cloud neighborhood update data, thereby obtaining a smooth surface model of the metal accessory;

[0022] Step S134: performing Gaussian curvature calculation on the smooth surface model of the metal accessory to obtain Gaussian curvature data;

[0023] Step S135: dividing the smooth surface model of the metal accessory into curvature regions according to the Gaussian curvature data, thereby obtaining flat curvature region data and curved curvature region data;

[0024] Step S136: performing curvature gradient analysis on the flat curvature region data, thereby obtaining flat curvature region gradient data;

[0025] Step S137: performing wavelet transformation on the curved curvature region data, thereby obtaining the curved curvature region wavelet scale data;

[0026] Step S138: integrating the multi-curvature surfaces of automobile industry metal parts according to the gradient data of the flat curvature region and the wavelet scale data of the curved curvature region, thereby obtaining the multi-curvature surface data of automobile industry metal parts.

[0027] The present invention can accurately capture the subtle features of the accessory surface by performing point cloud neighborhood calculation on the data of the narrow welding area of ​​the automobile industry metal accessories, and provide data support for subsequent surface processing. By performing weighted averaging on the point cloud neighborhood data, the interference of noise data can be effectively eliminated, making the point cloud data more accurate and stable, thereby improving the quality of the data. This operation not only makes the update of the point cloud data smoother, but also lays the foundation for building a more accurate metal accessory surface model. After updating the data by weighted averaging of the point cloud neighborhood, a smooth surface model of the metal accessory is constructed, which can better express the overall morphology of the accessory surface and help the subsequent geometric analysis and welding process formulation. By performing Gaussian curvature calculation on the smooth surface model, the curvature changes at different positions on the accessory surface can be accurately revealed, providing a basis for the subsequent curvature area division and welding process optimization. Gaussian curvature data provides a mathematical basis for judging whether the accessory surface is flat or curved, and then through the division of the curvature area, the data of the flat curvature area and the curved curvature area are obtained, which can better understand the geometric morphology differences of metal accessories. After performing curvature gradient analysis on the flat curvature area, we can deeply explore the geometric changes in the flat area, which is helpful to further optimize the welding process, ensure that the heat-affected zone during welding is effectively controlled, and avoid the generation of uneven molten pools. By performing wavelet transform on the data of the curved curvature area, curvature information at different scales can be extracted. This analysis not only helps to identify small-scale deformations on the surface of the accessory, but also helps to accurately control local heating and cooling during the welding process to avoid excessive melting or lack of penetration. By integrating the gradient data of the flat curvature area with the wavelet scale data of the curved curvature area, multi-curvature surface data of metal accessories for the automotive industry is formed. This data can more accurately reflect the complex geometric characteristics of the accessory surface. This integration method provides comprehensive support for the precise design of the welding process, ensures the adaptability and accuracy of the welding process, and thus improves the quality and efficiency of welding of metal accessories for the automotive industry.

[0028] Optionally, step S2 specifically includes:

[0029] Step S21: performing laser welding feature extraction and resistance spot welding feature extraction according to welding requirement data, thereby obtaining laser welding data and resistance spot welding data;

[0030] Step S22: performing laser welding simulation on the laser welding data to obtain laser welding simulation data;

[0031] Step S23: performing resistance spot welding simulation on the resistance spot welding data, thereby obtaining resistance spot welding simulation data;

[0032] Step S24: performing welding simulation integration according to the laser welding simulation data and the resistance spot welding simulation data, so as to obtain welding simulation data.

[0033] The present invention can accurately analyze the key parameters and conditions required for each welding method in the welding process by extracting laser welding features and resistance spot welding features according to welding demand data. This operation makes the characteristics of laser welding and resistance spot welding clearly divided, and provides clear basic data for subsequent welding simulation and process adjustment. The obtained laser welding data and resistance spot welding data provide a reliable basis for parameter setting in the simulation process, ensuring the independence and applicability of each welding method. By simulating the laser welding data, key factors such as the formation of the molten pool, the distribution of the heat-affected zone, and welding defects in the laser welding process can be understood in detail. This simulation not only helps to predict the welding effect, but also provides guidance for further optimizing the welding process, thereby ensuring that the welding quality meets the requirements. Similarly, by simulating the resistance spot welding data, the thermal cycle, joint strength, and formation process of the welding joint of the resistance spot welding can be deeply analyzed to ensure the stability and reliability of the welding quality. Through the simulation of these two welding methods, the advantages and disadvantages of different welding methods can be comprehensively evaluated, and the process can be adjusted in a targeted manner. Integrating the laser welding simulation data and the resistance spot welding simulation data can combine the simulation results of the two welding methods to provide a comprehensive welding effect analysis. Through integration, we can better understand the mutual influence and synergy of different welding methods, provide a basis for formulating the optimal welding plan, and ensure that the most appropriate welding method can be selected in specific applications.

[0034] Optionally, step S22 is specifically:

[0035] Step S221: extracting laser power characteristics and laser beam characteristics from laser welding data, thereby obtaining laser power data and laser beam data;

[0036] Step S222: constructing a laser beam distribution model according to the laser beam data, thereby obtaining a laser beam distribution model;

[0037] Step S223: identifying the maximum energy density of the focus according to the laser beam distribution model, thereby obtaining the maximum energy density data of the focus;

[0038] Step S224: performing laser heat source identification according to the focus maximum energy density data and the laser power data, thereby obtaining laser heat source data;

[0039] Step S225: drawing a temperature field distribution map according to the laser heat source data, thereby obtaining a temperature field distribution map of the laser heat source;

[0040] Step S226: Identify the irregular temperature gradient region according to the temperature field distribution diagram of the laser heat source, thereby obtaining the irregular temperature gradient region data;

[0041] Step S227: performing welding pool unevenness detection on the automotive industry metal parts data according to the irregular temperature gradient area data, thereby obtaining welding pool unevenness data;

[0042] Step S228: performing point source laser beam adjustment optimization according to the welding molten pool unevenness data, thereby obtaining point source laser beam adjustment data;

[0043] Step S229: Perform laser welding simulation according to the point source laser beam adjustment data to obtain laser welding simulation data.

[0044] The present invention can accurately extract the key laser parameters that affect the welding process by extracting the laser power characteristics and the laser beam characteristics of the laser welding data. These characteristics provide basic data for the subsequent laser beam distribution model and laser heat source analysis, ensuring that the welding process can be carried out under reasonable laser power and beam characteristics, thereby improving the welding quality. Then, by constructing a laser beam distribution model based on the laser beam data, the propagation characteristics of the laser beam and its distribution state during the welding process can be accurately predicted, providing a scientific basis for controlling the welding heat input. Through the identification of the maximum energy density of the focus, the energy distribution at the focus of the laser beam can be found, which is crucial for optimizing the welding process. After obtaining the maximum energy density data of the focus, the laser heat source is further identified, and the heat distribution released by the laser during the welding process can be accurately determined, providing a basis for subsequent temperature field control. By drawing the temperature field distribution map, the influence of the laser heat source on the metal accessories can be described in detail, and the temperature distribution during the welding process can be further understood, thereby helping to identify welding defects such as overheating or local overcooling. The identification of irregular temperature gradient areas helps to discover the temperature unevenness that occurs during the welding process, which often leads to uneven molten pools, thereby affecting the quality of welded joints. By identifying these irregular temperature gradient areas, timely remedial measures can be taken to avoid welding defects caused by temperature gradients. Detection based on uneven welding molten pool data helps to identify the uneven distribution of the molten pool, thereby guiding real-time adjustments during the welding process to ensure the uniformity of the molten pool and improve welding quality. Further optimization of point source laser beam adjustment helps to improve beam distribution and ensure that laser heat is more evenly distributed in the welding area, thereby effectively avoiding the negative impact of temperature gradients and optimizing the quality of welded joints.

[0045] Optionally, step S228 is specifically:

[0046] Mark the insufficient penetration area according to the uneven data of the welding pool, so as to obtain the insufficient penetration area data;

[0047] Mark the over-melting area according to the uneven data of the welding molten pool, so as to obtain the over-melting area data;

[0048] The laser power is increased for the insufficient melting depth area data to obtain the laser power data;

[0049] The laser beam scanning speed is increased for the over-melting area data, thereby obtaining the laser beam scanning speed data;

[0050] The point source laser beam adjustment optimization integration is performed according to the laser power data and the laser beam scanning speed data, so as to obtain the point source laser beam adjustment data.

[0051] The present invention can accurately identify the areas with insufficient penetration during welding by marking the areas with insufficient penetration according to the data of uneven welding pool. These areas usually affect the strength and quality of the welded joint. Early marking and identification of these areas provide clear targets for subsequent process adjustments to ensure that the welding quality meets the requirements. By marking the overmelting area, the areas where overmelting occurs during welding can be accurately located, which lead to material waste, weld defects and even damage to welded parts, so that these problems can be avoided in time during welding. Then, by increasing the laser power in the areas with insufficient penetration, the energy output of the laser can be enhanced to ensure that the penetration of the welding area is sufficient, thereby improving the welding quality. The adjustment of the laser power increase enables welding to better adapt to complex materials or geometric shapes and improve the strength and reliability of the welded joint. For the overmelting area, by increasing the laser beam scanning speed, the contact time between the laser beam and the metal during welding can be effectively reduced, thereby reducing excessive heat input, avoiding the occurrence of overmelting, and ensuring the stability and strength of the welded joint. The increase in the laser beam scanning speed helps to better control the heat input during welding and avoid material damage and unnecessary welding defects caused by excessive heat. Finally, by integrating the laser power data with the laser beam scanning speed data and optimizing the adjustment of the point source laser beam, the characteristics of insufficient melting depth and overmelting area can be comprehensively considered, the adjustment accuracy of the laser beam can be effectively improved, and the uniformity of energy distribution during welding can be ensured, thereby optimizing the welding quality. This optimization and integration process provides accurate process adjustment data for actual welding operations, reduces the occurrence of welding defects, and improves welding quality and efficiency.

[0052] Optionally, step S23 is specifically:

[0053] Step S231: extracting welding current characteristics from resistance spot welding data to obtain welding current data;

[0054] Step S232: drawing a heat generation distribution map according to the welding current data, thereby obtaining a heat generation distribution map;

[0055] Step S233: performing temperature field calculation on the heat generation distribution diagram to obtain welding current temperature field data;

[0056] Step S234: performing a molten pool formation simulation according to the welding current temperature field data, thereby obtaining welding current molten pool data;

[0057] Step S235: Acquire phase change temperature threshold data of automobile industry metal parts;

[0058] Step S236: performing heat-affected zone identification on the welding current molten pool data according to the phase change temperature threshold data of the automobile industry metal parts, thereby obtaining heat-affected zone data;

[0059] Step S237: adjusting the welding current speed of the heat affected zone data to obtain welding current speed data;

[0060] Step S238: Perform resistance spot welding simulation according to the welding current speed data to obtain resistance spot welding simulation.

[0061] The present invention can accurately obtain the characteristic data of the current by extracting the welding current characteristics of the resistance spot welding data, provide key parameters for the subsequent welding process control and optimization, and help identify the current changes in the welding process and their influence on the welding quality. Further, the heat generation distribution diagram is generated according to the welding current data, which can intuitively display the heat distribution in the welding process, provide effective support for the precise control of the heat source in the welding process, ensure that the temperature of the welding area is controlled within a reasonable range, and avoid problems such as overheating or uneven cooling. By calculating the temperature field of the heat generation distribution diagram, the welding current temperature field data can be obtained, which helps to analyze the temperature changes in different areas during the welding process and provide data support for further welding quality control. Next, the molten pool formation simulation is performed according to the welding current temperature field data, which can accurately simulate the formation of the molten pool during the welding process, and then evaluate the morphology, size and distribution of the molten pool, help adjust the energy input during the welding process, optimize the welding quality, and prevent the molten pool from being too large or too small. Obtaining the phase change temperature threshold data of metal accessories in the automotive industry is a key step in the welding process. Through these data, the temperature range in which the metal material undergoes phase change during the welding process can be determined, thereby accurately identifying the heat affected zone during the welding process and helping to determine the heat treatment requirements of the welding area. By identifying the heat-affected zone of the welding current pool data based on the phase change temperature threshold data, the areas affected by heat during the welding process can be effectively identified, and then the welding parameters can be optimized and adjusted in these areas to ensure that the welding quality meets the expected standards. By adjusting the welding current speed of the heat-affected zone data, the heat input and welding speed can be accurately controlled by adjusting the speed of the welding current, avoiding excessive or insufficient heat input and reducing the occurrence of welding defects. Finally, by simulating resistance spot welding with welding current speed data, the impact of the adjusted welding parameters on the actual welding effect can be verified, ensuring that the various parameters in the welding process are consistent with the expected goals, thereby achieving high-quality welding results.

[0062] Optionally, step S3 specifically includes:

[0063] Step S31: extracting welding morphology features and welding strength features according to welding simulation data, thereby obtaining welding morphology data and welding strength data;

[0064] Step S32: performing surface crack detection on the welding morphology data to obtain welding surface crack data;

[0065] Step S33: performing weld porosity detection on the welding strength data to obtain weld porosity data;

[0066] Step S34: integrating welding defects according to the welding surface crack data and the weld porosity data, thereby obtaining welding defect data.

[0067] The present invention extracts welding morphological features and welding strength features through welding simulation data, and can obtain detailed morphological data and strength data from the welding process, which provides a key technical basis for subsequent welding quality analysis and optimization. The morphological data can reflect the appearance, size, shape, etc. of the welding, while the welding strength data helps to evaluate the durability and resistance of the welding joint, and provides a guarantee for ensuring that the welding quality meets the structural requirements. Further surface crack detection of the welding morphological data can timely discover crack problems that occur during the welding process. These cracks often affect the strength and service life of the welding joint. Timely detection and identification help to take effective repair measures and reduce the risk of structural damage. By performing weld porosity detection on the welding strength data, the porosity defects generated during the welding process can be identified. The porosity not only affects the welding strength, but also causes stress concentration and reduces the overall performance of the welded part. By accurately identifying the weld porosity, the welding process parameters can be adjusted in time to avoid the occurrence or expansion of defects and ensure that the welding quality meets the standards. Finally, integrating welding defects based on welding surface crack data and weld porosity data helps to comprehensively analyze defect problems in the welding process, summarize and provide targeted improvement measures, thereby optimizing the entire welding process, improving the quality of welded parts, and enhancing their stability and reliability.

[0068] Optionally, step S32 is specifically:

[0069] Step S321: extracting weld contour features from welding shape data to obtain weld contour data;

[0070] Step S322: dividing the weld direction according to the weld contour data, thereby obtaining vertical direction weld data and parallel direction weld data;

[0071] Step S323: performing longitudinal magnetization on the vertical weld data to obtain longitudinal magnetization data;

[0072] Step S324: performing transverse magnetization on the parallel direction weld data to obtain transverse magnetization data;

[0073] Step S325: performing dry magnetic powder spraying according to the longitudinal magnetization data, thereby obtaining dry magnetic powder spraying data;

[0074] Step S326: performing wet magnetic powder spraying according to the transverse magnetization data, thereby obtaining wet magnetic powder spraying data;

[0075] Step S327: performing magnetic powder spray fusion according to the dry magnetic powder spray data and the wet magnetic powder spray data, thereby obtaining magnetic powder spray data;

[0076] Step S328: Perform surface crack defect visualization on the magnetic powder spraying data to obtain welding surface crack data.

[0077] The present invention can obtain the geometric feature information of the weld by extracting the weld contour feature of the welding morphological data, which helps to further analyze the morphology and structure of the weld joint to ensure the controllability of the welding quality. On this basis, the weld direction is divided according to the weld contour data, and the weld can be divided into the vertical direction and the horizontal direction, and the stress and mechanical properties of the welding in different directions are further clarified, which is crucial for subsequent defect detection and performance evaluation. By longitudinally magnetizing the vertical weld data, potential defects in this direction, such as cracks and pores, can be concentratedly detected, thereby improving the accuracy of defect detection. Similarly, by transversely magnetizing the parallel weld data, it is helpful to find surface defects in the parallel direction and ensure the comprehensive quality inspection of the weld joint. After longitudinal magnetization and transverse magnetization, dry and wet magnetic powder spraying can be used respectively to perform more accurate defect detection on the weld areas of different magnetization directions. Dry magnetic powder spraying helps to show defects such as longitudinal cracks on the surface of the weld joint, while wet magnetic powder spraying can enhance the defect appearance effect of the parallel weld. By integrating the dry and wet magnetic powder spraying data, comprehensive and accurate welding defect detection data can be obtained, providing multi-dimensional information for welding quality assessment. Finally, surface crack defect visualization based on these magnetic powder spraying data can effectively reveal potential crack defects on the surface of the welded joint, timely discover and deal with welding quality problems, and ensure the performance and safety of the welded parts.

[0078] Optionally, step S33 is specifically:

[0079] Step S331: performing strength statistics on the welding strength data to obtain low-strength welding data;

[0080] Step S332: performing region identification on the welding simulation data according to the low-strength welding data, thereby obtaining low-strength welding region data;

[0081] Step S333: performing X-ray imaging according to the low-intensity welding area data, thereby obtaining an X-ray low-intensity welding area image;

[0082] Step S334: performing a binary conversion on the X-ray low-intensity welding area image to obtain a binary image of the X-ray low-intensity welding area;

[0083] Step S335: performing gradient calculation on the binary image of the low-intensity X-ray welding area to obtain binary image gradient data, and performing pore gradient statistics based on the binary image gradient data to obtain a high-gradient binary pore image;

[0084] Step S336: performing corrosion according to the high gradient binary pore image, so as to obtain a high gradient binary pore corrosion image;

[0085] Step S337: dilating the high gradient binary pore corrosion image to obtain a high gradient binary pore dilation image;

[0086] Step S338: Identify the pore positions based on the high-gradient binary pore expansion image to obtain weld pore data.

[0087] The present invention can effectively identify the low-intensity area appearing in the welding process by performing strength statistics on the welding strength data, which is of great significance for evaluating the quality of the welded joint. By performing regional identification based on the low-intensity welding data, the specific location of the low-intensity in the welding process can be accurately located, so that subsequent detection and repair can be carried out in a targeted manner. By performing X-ray imaging on the low-intensity welding area data, a detailed image of the area can be obtained, providing a non-destructive testing method to evaluate the internal structure and defects of the low-intensity area, ensuring that no potential problems are missed. After the X-ray low-intensity welding area image is binarized, it helps to simplify image processing, improve the efficiency of subsequent analysis, and highlight the difference between the low-intensity area and the surrounding structure. Gradient calculation is performed on the X-ray low-intensity welding area binarized image, which can more clearly show the edge and morphology of the welding area, thereby helping to identify the details of the welding defects. Porosity gradient statistics provide a high-gradient binarized pore image, which helps to more accurately locate the pores formed in the welding process and provide data support for subsequent repair and improvement. By performing corrosion treatment on the high-gradient binarized pore image, image noise can be reduced, the true morphology of the welding defect can be further highlighted, and subsequent analysis and processing are convenient. The expansion process can restore the pore information lost during the corrosion process, ensure the integrity of the pores is preserved, and provide more accurate welding defect analysis. Finally, the pore position identification through high-gradient binary pore expansion image can accurately locate the position of welding pores, providing a reliable basis for welding quality control and defect repair. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0089] Figure 1 It is a schematic flow chart of the steps of the laser welding system control method for metal accessories of the present invention;

[0090] Figure 2 Detailed step flow diagram of step S1 in the present invention;

[0091] Figure 3 Detailed step flow diagram of step S2 in the present invention;

[0092] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0093] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. 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 technicians in this field without creative work are within the scope of protection of the present invention.

[0094] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0095] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0096] To achieve this, please refer to Figures 1 to 3 The present invention provides a laser welding system control method for metal accessories, the method comprising the following steps:

[0097] Step S1: Obtaining automobile industry metal parts data, and performing welding demand analysis based on the automobile industry metal parts data, thereby obtaining welding demand data;

[0098] In this embodiment, it is necessary to extract detailed data of metal accessories from the material management system in the automobile manufacturing process. The data includes information such as the size, shape, material, thickness, surface treatment, etc. of the accessories. The specific data can be collected by scanners, computer-aided design (CAD) systems, and sensors in the production workshop. The extracted data needs to be cleaned to remove irrelevant or redundant information. Then, based on the acquired metal accessories data, engineering analysis tools such as SolidWorks or CATIA are used to analyze the welding requirements of each accessory. The welding demand analysis is mainly aimed at the structural and connection strength requirements of the accessories, combined with the applicability of the welding process (such as welding position, welding method and process parameters, etc.). For each accessory, the welding process manual and technical specifications are used to check whether a specific welding process (such as laser welding, spot welding, etc.) is required, and this information is recorded as welding demand data, including the welding method, welding position, welding process requirements, material type and thickness of each component.

[0099] Step S2: performing laser welding feature extraction and resistance spot welding feature extraction according to the welding requirement data, thereby obtaining laser welding data and resistance spot welding data; performing welding simulation according to the laser welding data and the resistance spot welding data, thereby obtaining welding simulation data;

[0100] In the present embodiment, when the laser welding feature is extracted, it is necessary to input the operating parameters of the laser welding system, such as laser power, welding speed, focal position and beam diameter, etc., and in addition, it is necessary to consider the thickness of the metal accessories, surface finish and other factors. After these parameters are input, laser welding simulation software (such as ANSYS or Comsol Multiphysics) can be used for simulation, and the physical fields such as heat source distribution, temperature field, stress field, etc. in the welding process are calculated to obtain the key features of laser welding, such as heat input distribution, molten pool shape, cooling rate, etc. Similarly, for the resistance spot welding feature extraction, it is necessary to input the core parameters of resistance welding such as electrode pressure, welding current, time, etc. Then, resistance welding simulation software (such as Abaqus) is used to simulate current and heat distribution, and data such as heat conduction, indentation depth and solder joint characteristics in the welding process are extracted. The laser welding data and resistance spot welding data obtained by these two steps will provide necessary input for subsequent welding simulation.

[0101] Step S3: extracting welding morphology features and welding strength features according to the welding simulation data, thereby obtaining welding morphology data and welding strength data; performing welding defect detection according to the welding morphology data and welding strength data, thereby obtaining welding defect data;

[0102] In this embodiment, welding morphological features, including the geometry, depth, width and molten pool flow of the weld, are extracted from the welding simulation data. A high-resolution imaging system (such as a microscope or X-ray CT scan) is used to perform physical inspection of the weld joint to capture the surface and internal morphological features of the weld. Then, these image data are input into the welding defect detection algorithm, and analyzed using an automated detection tool based on image processing. Through edge detection algorithms (such as the Canny algorithm) and morphological analysis (such as the corrosion expansion method), defect areas such as pores and cracks are identified. At the same time, the welding strength feature extraction uses a stress test method, and the maximum tensile strength, fracture position, and strain distribution data are recorded by performing a tensile test on the weld joint. In addition, a finite element analysis software (such as Abaqus) is used to further calculate the strength performance of the weld joint according to the thermal stress data of the welding process, and extract relevant strength data. Combined with the welding morphological data and the welding strength data, it is possible to evaluate whether there are welding defects and generate welding defect data, mainly including information such as defect type, defect location, and defect severity.

[0103] Step S4: acquiring laser welding system data; optimizing laser welding parameters of the laser welding system data according to the welding defect data, thereby obtaining laser welding optimization parameters;

[0104] In this embodiment, it is necessary to obtain real-time working data from the laser welding system, including welding process parameters such as laser power, welding speed, focal position, pulse width, etc. These parameters can be monitored and recorded in real time from the sensor through a welding control system (such as a PLC or SCADA system). After acquiring the data, the laser welding parameters are optimized, mainly adjusting the welding parameters according to the welding defect data to reduce the occurrence of welding defects. First, an initial welding parameter range is determined, and a suitable optimization target is set, such as optimizing the width and depth of the weld to reduce the probability of pores or cracks. In this process, an automated tool based on a genetic algorithm or a particle swarm optimization algorithm can be used to search parameters, and the combination of optimized parameters can be analyzed by simulation. For example, during the welding process, the welding power and welding speed are optimized for pore defects, and their combination is adjusted until the minimum pore incidence rate is obtained. According to the specific defect type and frequency extracted from the welding defect data, the optimized welding process parameters are obtained by adjusting the parameters.

[0105] Step S5: adjusting the laser welding process of the welding simulation data according to the laser welding optimization parameters, thereby obtaining laser welding process adjustment data, and uploading the data to the laser welding system to execute the laser welding process control task.

[0106] In this embodiment, the welding simulation data is adjusted according to the laser welding optimization parameters to ensure that the welding process can meet the quality requirements in actual production. By inputting the optimization parameters (such as power, welding speed, focal position, etc.) into the laser welding control system, and performing welding process adjustment. The adjusted data can be uploaded to the control system of the welding machine through the welding process management system (such as MES system). In order to ensure the real-time and accuracy of the data, the process adjustment parameters can be transmitted to the laser welding equipment in real time by setting a data transmission protocol such as Modbus protocol or Ethernet / IP protocol. In this process, process verification is also required to ensure that the welding machine can perform these adjusted processes. The quality and performance of the welded joint are detected through actual welding tests, and the adjustment results are fed back to continuously optimize the welding process, which is uploaded to the laser welding system for precise control to ensure that the welding tasks on the production line can be performed efficiently and accurately.

[0107] Optionally, step S1 specifically includes:

[0108] Step S11: acquiring automobile industry metal parts data, and performing welding area feature extraction based on the automobile industry metal parts data, thereby obtaining automobile industry metal parts welding area data;

[0109] In the present embodiment, the CAD drawing data of metal accessories is obtained from the parts database in the automobile industry production, including the material type, thickness, shape, and contact area information of each accessory. The extraction process is carried out by professional CAD tools (such as AutoCAD or CATIA), and these tools are used to analyze the geometric data of the parts to obtain the welding area of ​​each metal accessory. The definition of the welding area is based on the contact area at the connection surface of the accessory, which is usually the contact surface of two or more metal parts. By inputting the geometric data of the accessory into the welding area extraction algorithm, the algorithm identifies the welding area according to the welding standard (such as the minimum size of the contact surface is 10mm). These welding areas can be plane or curved areas, and the extracted welding area data include welding position, shape, size and specific requirements of each area (such as welding strength, tensile strength, etc.).

[0110] Step S12: performing narrow area identification according to the welding area data of automobile industry metal parts, thereby obtaining the welding narrow area data of automobile industry metal parts;

[0111] In this embodiment, a geometric analysis algorithm (such as a boundary recognition algorithm) is used to calculate the boundary of the welding area and identify its narrow part in two-dimensional or three-dimensional space. The specific operation includes inputting the coordinate data of the welding area, scanning the area segment by segment using a scanline algorithm, calculating the width of each segment, and setting a threshold value for the narrow area (for example, an area with a width less than 5 mm is considered a narrow area). For each identified narrow area, its characteristic data such as position, size and shape are recorded to finally form the welding narrow area data. These data are further used to determine the appropriate welding method and welding parameters.

[0112] Step S13: performing multi-curvature surface analysis on the welding narrow area data of automobile industry metal accessories, thereby obtaining multi-curvature surface data of automobile industry metal accessories;

[0113] In this embodiment, multi-curvature analysis is performed on the surfaces of these areas. The goal of multi-curvature analysis is to identify the curvature changes of the surface in multiple directions, especially for parts with complex geometric shapes. Each narrow welding area is analyzed using a curvature analysis tool (such as the CurvatureMap tool or the surface analysis function in Matlab). The process first inputs the surface geometry data of the narrow area (such as point cloud data obtained by a three-dimensional scanner), and then determines the curvature distribution of the surface by calculating parameters such as the surface normal vector, the principal curvature, and the Gaussian curvature. During the analysis, an appropriate threshold is set, and a curvature value with a standard value of the principal curvature greater than 5 is usually selected to define a multi-curvature area. The generated multi-curvature surface data includes a curvature distribution map of each surface and the direction of surface change, forming welding data containing curvature information.

[0114] Step S14: performing spatial complexity evaluation according to the multi-curvature surface data of automobile industry metal accessories, thereby obtaining the spatial complexity of automobile industry metal accessories;

[0115] In this embodiment, the spatial complexity of the multi-curvature surface data is evaluated to evaluate the geometric complexity of the welding area. The evaluation is achieved by calculating the surface area, volume and complexity of the spatial geometry of the welding area. In the specific implementation, the multi-curvature surface data is first input using a three-dimensional modeling software (such as SolidWorks or AutoCAD), and the surface area and volume of the welding area are calculated using a geometric volume calculation tool. According to the geometric shape of the welding area (such as a circle, an ellipse, a curved surface, etc.), the spatial complexity is calculated using a spatial complexity algorithm (such as an entropy algorithm or a boundary complexity algorithm). For each welding area, the welding area is classified by setting a specific spatial complexity threshold (such as an area with a complexity greater than 10 as a high complexity area). The evaluation result is output as the spatial complexity data of each area, including the complexity level of the area, shape characteristics, etc.

[0116] Step S15: Perform welding method requirement adaptation evaluation according to the spatial complexity of automobile industry metal parts, so as to obtain welding requirement data.

[0117] In this embodiment, based on the spatial complexity data and the characteristics of the welding area, the welding method required for each area (such as laser welding, spot welding or TIG welding, etc.) is evaluated. In the specific operation, the adaptation rules of the welding method are first set. For example, for areas with high spatial complexity (such as complex surfaces or multi-curvature areas), laser welding or robot welding technology is preferably used; for areas with low spatial complexity, traditional resistance spot welding and other processes can be used. By consulting the welding process database and combining the value of the spatial complexity, the system will automatically match the appropriate welding method and generate welding requirement data. The welding requirement data includes the welding method, welding parameter requirements, welding speed, etc. for each area. Based on these data, the subsequent welding process can be optimized and scheduled to ensure the accurate implementation of the process.

[0118] Optionally, step S13 is specifically:

[0119] Step S131: performing point cloud neighborhood calculation on the narrow area data of welding of automobile industry metal parts, thereby obtaining point cloud neighborhood data;

[0120] In this embodiment, the data of the narrow welding area of ​​metal accessories in the automotive industry is obtained, and the three-dimensional point cloud data of the welding area is obtained by a point cloud data acquisition tool (such as a 3D scanner or a lidar). The goal of point cloud neighborhood calculation is to extract the spatial relationship of each point through neighborhood analysis. In actual operation, the point cloud data is processed using a toolkit such as PCL (Point Cloud Library) to calculate the neighborhood area of ​​each point in space. The neighborhood area is defined as other points whose distance from the current point is less than a set threshold. The distance threshold can be set to 1mm, and the K nearest neighbors of each point are calculated using the K-nearest neighbor algorithm (KNN algorithm) based on Euclidean distance. The obtained point cloud neighborhood data contains the neighborhood point set of each point, as well as the geometric characteristics of the neighborhood area, such as average distance, density, etc. These point cloud neighborhood data provide the basis for subsequent weighted averaging and model building.

[0121] Step S132: performing weighted averaging according to the point cloud neighborhood data to obtain point cloud neighborhood weighted average data, and updating the point cloud neighborhood data according to the point cloud neighborhood weighted average data to obtain point cloud neighborhood update data;

[0122] In this embodiment, the point cloud neighborhood data is used for weighted averaging. A weighting function is defined based on the neighborhood data of each point. The weighting function is weighted according to the distance of the neighborhood points. The neighborhood points that are closer are given a greater weight, and the neighborhood points that are farther away are given a smaller weight. In specific operations, the weight can be calculated by a Gaussian function (for example, weight = exp(-d 2 / 2σ2)), where d is the distance between the current point and the neighboring point, and σ is the standard deviation for controlling the weighting degree, which is usually set to 2mm. According to this weighting function, the coordinates of the neighboring points are weighted averaged to obtain the new coordinates of each point. Through this operation, the noise in the point cloud data is smoothed and the smoothness of the data is enhanced. The point cloud neighborhood data is updated according to the weighted averaged coordinates, and the new point cloud coordinate values ​​are updated back to the neighborhood data set, providing more accurate point cloud data for the subsequent smooth surface model construction.

[0123] Step S133: constructing a smooth surface model of the metal accessory according to the point cloud neighborhood update data, thereby obtaining a smooth surface model of the metal accessory;

[0124] In this embodiment, a three-dimensional reconstruction algorithm (such as the Poisson reconstruction algorithm) is used to reconstruct the surface of the updated point cloud data to generate a continuous three-dimensional surface. During the reconstruction process, the surface smoothness parameters are set to ensure that the reconstructed surface is smooth and conforms to the actual physical form. For the surface of metal accessories, the default settings of the Poisson algorithm are usually used to ensure that the surface is as smooth as possible and without overfitting under the constraints of the point cloud data. The result of the reconstruction is a smooth surface model of the metal accessory, which contains the coordinate data, normal direction and related geometric information of each surface point. This smooth surface model provides the necessary input data for the subsequent Gaussian curvature calculation.

[0125] Step S134: performing Gaussian curvature calculation on the smooth surface model of the metal accessory to obtain Gaussian curvature data;

[0126] In this embodiment, the Gaussian curvature calculation method is used to analyze the smooth surface model of the metal accessories obtained in step S133. The calculation of Gaussian curvature is achieved by multiplying the two principal curvatures on the surface. The specific calculation formula is: K=κ1*κ2, where κ1 and κ2 are the principal curvatures of any point on the surface. In the implementation process, the normal vector and principal curvature of each surface point are first extracted from the smooth surface model, and the principal curvature value of each point is calculated by numerical methods (such as least squares method or finite difference method). Then, using these two principal curvature values, the Gaussian curvature of each point is calculated. When calculating the Gaussian curvature, a reasonable calculation accuracy is set according to the spatial position, normal vector and direction of the principal curvature of the surface point, and the accuracy is usually set to 0.1°. The generated Gaussian curvature data includes the Gaussian curvature value of each point, which will be used for subsequent curvature area division.

[0127] Step S135: dividing the smooth surface model of the metal accessory into curvature regions according to the Gaussian curvature data, thereby obtaining flat curvature region data and curved curvature region data;

[0128] In this embodiment, a Gaussian curvature threshold (for example, 0.5) is set, and the area greater than the threshold is divided into a curved curvature area, and the area less than the threshold is divided into a flat curvature area. In specific implementation, the Gaussian curvature value of each point in the smooth surface model is traversed, and each point is classified into a corresponding curvature area. In this process, the setting of the threshold depends on the specific welding requirements of the metal accessories. For ordinary welding areas, 0.5 is usually used as the threshold to ensure a clear division of the curved area and the flat area. After the division is completed, the flat curvature area data and the curved curvature area data are generated respectively, containing their respective geometric shapes, positions, areas and other information.

[0129] Step S136: performing curvature gradient analysis on the flat curvature region data, thereby obtaining flat curvature region gradient data;

[0130] In this embodiment, the curvature gradient calculation method is selected, and the curvature gradient is calculated by the second-order difference method. The curvature gradient value is obtained by calculating the Gaussian curvature difference between adjacent points. Each point in the flat curvature area is traversed, and the curvature difference between it and the surrounding neighborhood points is calculated, and then the gradient value is calculated using the difference formula. The unit of the gradient value is curvature unit / distance unit, and the calculated gradient data can reflect the curvature change rate of the flat area. The generated curvature gradient data contains the gradient value of each flat area and serves as the input data for subsequent wavelet transform analysis.

[0131] Step S137: performing wavelet transformation on the curved curvature region data, thereby obtaining the curved curvature region wavelet scale data;

[0132] In this embodiment, the goal of wavelet transform is to decompose the complex information of the curved curvature area into different scales to extract the curvature features of different frequencies. In the specific implementation, the discrete wavelet transform (DWT) method is used to transform the curved area data through the selected wavelet basis function (such as Daubechies wavelet or Haar wavelet). The transformation process first decomposes the Gaussian curvature data of each curved area into low-frequency and high-frequency parts. Through multi-layer wavelet decomposition, data of different scales can be obtained to describe the local characteristics of the surface at different scales. By setting the threshold of the wavelet scale (for example, the part with a frequency higher than 5Hz is a high frequency), the generated wavelet scale data will be further used for subsequent multi-curvature surface integration.

[0133] Step S138: integrating the multi-curvature surfaces of automobile industry metal parts according to the gradient data of the flat curvature region and the wavelet scale data of the curved curvature region, thereby obtaining the multi-curvature surface data of automobile industry metal parts.

[0134] In this embodiment, the flat area gradient data and the curved area wavelet scale data are matched and fused. In the implementation process, the two data sets are integrated according to the complexity and importance of each area through a data fusion algorithm (such as a weighted average method or a Kalman filter method). The integrated data will contain complete curvature information, including the curvature change of the flat area and the scale change of the curved area. The integrated multi-curvature surface data provides a detailed description of the geometry of the metal fittings and serves as the basis for subsequent welding process optimization.

[0135] Optionally, step S2 specifically includes:

[0136] Step S21: performing laser welding feature extraction and resistance spot welding feature extraction according to welding requirement data, thereby obtaining laser welding data and resistance spot welding data;

[0137] In the present embodiment, laser welding feature extraction is performed according to welding requirement data. Welding requirement data generally includes information such as material type, thickness, welding method and required welding accuracy of metal accessories. According to these requirements, suitable laser welding parameters are selected, such as laser power (set to between 500W and 2000W), welding speed (set to between 1 and 3mm / s) and focal length (set to focal length 8-10mm). Through these parameters, the laser welding feature extraction algorithm is used to calculate the radiation intensity distribution of the laser beam and the heat input of the laser action zone. The result of this step is laser welding data, which contains information such as the size of the heat affected zone, welding speed and heat transfer characteristics in the laser welding process. Subsequently, the resistance spot welding feature is extracted, and the same welding requirement data is used to define the welding parameters of the resistance spot welding, including current intensity (set to 5kA to 15kA), voltage (set to 2V to 5V), welding time (set to 5ms to 20ms), and the heat input between the electrodes, the molten pool formation and its dynamic changes are calculated by the spot welding feature extraction algorithm. This process ultimately yields resistance spot welding data, including heat distribution in the weld area, molten pool formation, and electrode pressure.

[0138] Step S22: performing laser welding simulation on the laser welding data to obtain laser welding simulation data;

[0139] In this embodiment, a suitable heat conduction simulation method (such as finite element method FEA) is selected to perform numerical simulation on the laser welding process. The simulation process uses the welding parameters (such as laser power, welding speed, focal length, etc.) in the laser welding data as input, and calculates the distribution of heat flow during welding through the heat conduction equation. At this time, the material properties (such as thermal conductivity, specific heat capacity, etc.) and geometric shape of the welding workpiece are also processed as input parameters. The heat conduction model usually sets the boundary condition as the initial temperature value of the welding area to room temperature, and the radiation intensity of the laser beam is distributed according to the welding power distribution. Through the calculation of this model, the temperature field distribution and thermal stress conditions during the welding process can be obtained. In order to improve the accuracy of the simulation, a smaller grid size (such as 1mm×1mm) is used for calculation to ensure that the changes in heat flow are fully simulated. The simulation result is the laser welding simulation data, which includes the temperature changes, heat conduction paths and thermal stress distribution of each point during the welding process.

[0140] Step S23: performing resistance spot welding simulation on the resistance spot welding data, thereby obtaining resistance spot welding simulation data;

[0141] In the present embodiment, the resistance spot welding process is simulated by using the parameters such as current, voltage, welding time in the resistance spot welding characteristic data. The current density distribution model is used to calculate the heat input between the electrodes according to the contact resistance of the electrode, the distribution of the current through the electrodes and the welding time. During simulation, the heat conduction process between the electrode contact surface, the welding point area and the metal workpiece is analyzed using the finite element method (FEA). For the simulation of resistance spot welding, factors such as the thermal conductivity of the electrode material, the electrode pressure, and the welding gap need to be considered to ensure the accuracy of the simulation. During the simulation process, the change of the temperature field and the formation process of the molten pool are dynamically simulated by the change of the current, voltage and welding time. The initial current is set to 10kA, and simulation is performed according to the actual working conditions. Finally, the simulation result is the resistance spot welding simulation data, including the temperature change of the electrode and the workpiece contact area during the welding process, the molten pool size and the electrode pressure and other information.

[0142] Step S24: performing welding simulation integration according to the laser welding simulation data and the resistance spot welding simulation data, so as to obtain welding simulation data.

[0143] In this embodiment, the simulation results of laser welding and resistance spot welding are fused by a data integration method (such as a weighted average method or a fitting algorithm). When integrating, a suitable weight is selected. The weight coefficient is usually set according to the application scenario and importance of each welding method. For example, laser welding has a greater impact in higher temperature areas, so a higher weight can be set. The integrated data contains information such as heat input, temperature field distribution, thermal stress, and molten pool formation of the two welding methods. In specific operations, the temperature, heat flow, stress and other parameters in the two simulation data sets are aligned using a data fusion algorithm, and global optimization is performed. The integrated welding simulation data is the final welding process simulation result, which provides a comprehensive welding process prediction, including the advantages and disadvantages of the two welding methods and the prediction information of welding quality. This simulation result provides a basis for subsequent welding process optimization and parameter adjustment.

[0144] Optionally, step S22 is specifically:

[0145] Step S221: extracting laser power characteristics and laser beam characteristics from laser welding data, thereby obtaining laser power data and laser beam data;

[0146] In this embodiment, the laser power in the laser welding process is collected in real time by a laser power monitoring instrument. The power data of the laser output is collected by a power sensor, and the range is set to between 100W and 3000W, and the power threshold is set according to the welding process requirements. For example, for a certain metal material, the power threshold is set to 1500W. The collected power data is used to describe the energy input characteristics of the laser during the welding process. Next, a high-precision laser measurement system (such as a laser beam analyzer) is used to measure the laser beam to extract characteristic data such as its diameter, spot size, eccentricity and fluctuation. The collection range of the beam data is generally between 5mm and 20mm, and the measurement error of the spot size is controlled within ±0.1mm. According to the measured beam diameter and power distribution data, the relevant characteristic data of the laser power and the beam are obtained as input for subsequent analysis.

[0147] Step S222: constructing a laser beam distribution model according to the laser beam data, thereby obtaining a laser beam distribution model;

[0148] In this embodiment, the laser beam data is used to construct a laser beam distribution model. The light intensity data of multiple measurement points are combined with the measurement of the distance sensor to describe the changing characteristics of the laser beam at different distances. Several different measurement positions are selected in the welding area (for example: 1mm, 2mm, 3mm, etc. from the focal position), and the intensity distribution of the laser beam is measured at these positions. According to the collected data, a mathematical method (such as the least squares method) is used to fit the intensity distribution curve of the laser beam to determine the radius and energy distribution of the beam. The distribution model usually includes the radiation radius of the beam, the spot shape (for example, circular or elliptical), and the attenuation characteristics of the light intensity. The radiation radius of the beam can be set according to the requirements of the welding process, and is usually controlled between 3mm and 10mm. Finally, a distribution model of the laser beam is obtained, which reflects the intensity changes of the laser beam at different positions in space.

[0149] Step S223: identifying the maximum energy density of the focus according to the laser beam distribution model, thereby obtaining the maximum energy density data of the focus;

[0150] In this embodiment, the energy density at the center of the beam in the beam distribution model is analyzed to identify the focal position of the beam. The focal position is usually the area where the energy of the laser beam is most concentrated, and the energy density at this position is the highest. By measuring the energy distribution before and after the focus layer by layer, the exact position of the focus is determined, and the maximum energy density at this position is calculated. During the identification of the focal position, the accuracy of the focal position is ensured by adjusting the beam detection area of ​​the measuring device. The data of the maximum energy density is usually provided by an energy sensor, and the sensitivity of the sensor is set to 1W / cm 2 , ensuring that the highest energy density at the laser beam focus can be accurately measured.

[0151] Step S224: performing laser heat source identification according to the focus maximum energy density data and the laser power data, thereby obtaining laser heat source data;

[0152] In this embodiment, the maximum energy density data of the focus is matched with the laser power data to confirm the location and size of the heat source. The laser power data directly reflects the energy of the laser input, while the maximum energy density data reflects the energy concentration at the focus. Based on these data, the heat distribution of the welding area is estimated by a heat source identification algorithm (such as heat source calculation based on the integral method). The parameters of the process, such as the intensity, expansion radius and shape of the heat source, will depend on the laser power and the focus energy density of the beam. Usually, the calculation of these parameters is based on the parameters within the power range set by the laser, such as the power range of 1000W to 2500W.

[0153] Step S225: drawing a temperature field distribution map according to the laser heat source data, thereby obtaining a temperature field distribution map of the laser heat source;

[0154] In this embodiment, the temperature distribution of the welding area is calculated using a heat conduction model based on the size, shape and distribution of the heat source. In this process, it is necessary to input the energy distribution of the heat source, the thermal conductivity and specific heat capacity of the workpiece material and other physical properties. The conduction process of the heat source on the metal surface is simulated by numerical simulation (such as finite element analysis) to obtain temperature values ​​at different positions. The temperature field distribution diagram shows the temperature changes at each point in the welding area, and the area with the highest temperature is usually near the laser focus. The temperature accuracy is usually set to ±5°C to ensure that the data is accurate enough for subsequent analysis.

[0155] Step S226: Identify the irregular temperature gradient region according to the temperature field distribution diagram of the laser heat source, thereby obtaining the irregular temperature gradient region data;

[0156] In this embodiment, irregular temperature gradient areas are identified based on the temperature field distribution map. By analyzing the temperature change rate in the temperature field, the area with abnormal temperature gradient changes is located. The specific operation is to identify the area with excessive temperature gradient by setting the temperature gradient threshold (for example, the area where the temperature change per millimeter exceeds 20°C). Image processing software is used to further confirm the location and range of the irregular temperature gradient area in combination with temperature data and graphic recognition algorithms. The key to the recognition process is to accurately set the threshold to ensure that the area with too fast temperature changes can be correctly marked.

[0157] Step S227: performing welding pool unevenness detection on the automotive industry metal parts data according to the irregular temperature gradient area data, thereby obtaining welding pool unevenness data;

[0158] In this embodiment, the data of irregular temperature gradient area is used to detect the unevenness of the welding molten pool. The uniformity of the molten pool during welding is detected by analyzing the overlap between the abnormal temperature gradient area and the molten pool. In this process, thermal imaging technology is used to obtain the temperature data of the molten pool, and the shape and temperature distribution of the welding molten pool are analyzed in combination with the welding path. The uniformity detection of the molten pool is usually based on the temperature fluctuation in the molten pool area, and a threshold value such as a fluctuation range of 10°C is set to determine whether there is an uneven temperature area in the molten pool. For uneven areas, mark them and provide data support for subsequent adjustments.

[0159] Step S228: performing point source laser beam adjustment optimization according to the welding molten pool unevenness data, thereby obtaining point source laser beam adjustment data;

[0160] In this embodiment, the focal position, power and spot size of the laser beam are adjusted by analyzing the abnormal area of ​​temperature distribution in the molten pool. The adjustment process of the laser beam is based on the data of the uneven welding pool, and uses a real-time feedback mechanism to make adjustments according to the temperature changes monitored during the welding process. For example, when it is found that the temperature in a certain area is too high, the laser power can be appropriately reduced, or the diameter of the beam can be changed to make it cover the welding area more evenly. The core of the adjustment process is to ensure the optimization of the beam, usually by controlling the focal length and power distribution of the beam through software.

[0161] Step S229: Perform laser welding simulation according to the point source laser beam adjustment data to obtain laser welding simulation data.

[0162] In this embodiment, the adjusted laser beam data is input into the simulation software to simulate the thermal effect of the laser beam on the metal surface during welding. This process involves inputting the power, focal position and heat source distribution data of the beam into the simulation model to simulate its heat conduction and material behavior during the actual welding process. The simulation results will provide data such as the temperature field and molten pool morphology during the welding process, further verifying whether the optimized laser beam can improve the welding quality and provide a basis for actual operation.

[0163] Optionally, step S228 is specifically:

[0164] Mark the insufficient penetration area according to the uneven data of the welding pool, so as to obtain the insufficient penetration area data;

[0165] In this embodiment, an infrared thermal imager is used to obtain a thermal image of the welding process and extract temperature distribution information. A threshold for insufficient penetration is set, for example, an area below 1800°C (a value determined by the melting point of the welding material and the operating parameters) is marked as insufficient penetration. An image processing algorithm (such as Canny edge detection) is used to locate the boundary of the low temperature zone, and the position and area of ​​insufficient penetration are determined in combination with the weld geometry model. By vectorizing the marked area, accurate data of the insufficient penetration area is output for subsequent optimization.

[0166] Mark the over-melting area according to the uneven data of the welding molten pool, so as to obtain the over-melting area data;

[0167] In this embodiment, the high temperature point distribution statistical analysis is performed on the welding pool data, and the over-melting threshold is set. For example, the area above 2000°C is used as the over-melting area. The multi-peak analysis is performed through the distribution curve of the thermal imaging data to find abnormal high temperature points. The regional growth algorithm is combined with the neighborhood temperature gradient to expand the range of the over-melting area. The maximum temperature standard in the laser welding process specification is used to determine the specific boundary of the over-melting area. The over-melting area data finally generated is in the form of polygonal vectors, recording the specific over-melting position and range.

[0168] The laser power is increased for the insufficient melting depth area data to obtain the laser power data;

[0169] In this embodiment, the laser power requirement of the corresponding area is calculated based on the data of the insufficient penetration area. Combined with the thermal conductivity of the welding material (for example, 40W / m·K) and the welding speed (such as 2mm / s), the reverse heat transfer method is used to estimate the required power increase value. Through the laser power control module, the power output is gradually increased, and the initial power increase is set to 5%. The power increment is adjusted by real-time monitoring of the temperature gradient change in the welding area until the temperature reaches the set penetration temperature threshold. The value and time of each power adjustment are recorded to generate the laser power data after the increase.

[0170] The laser beam scanning speed is increased for the over-melting area data, thereby obtaining the laser beam scanning speed data;

[0171] In this embodiment, the over-melting area often causes heat accumulation due to the low scanning speed. A dynamic adjustment device for the laser head is used to set the initial scanning speed increase to 10%, and the scanning trajectory and speed change are adjusted in real time. A temperature feedback control system is used to record the regional temperature distribution after each speed adjustment to ensure that the maximum temperature is reduced to the target temperature range (for example, below 2000°C). The final laser beam scanning speed data is generated based on the recorded adjustment trajectory and speed change.

[0172] The point source laser beam adjustment optimization integration is performed according to the laser power data and the laser beam scanning speed data, so as to obtain the point source laser beam adjustment data.

[0173] In this embodiment, the laser power data and the laser beam scanning speed data are combined to adopt a multi-dimensional control strategy to optimize the energy distribution and movement path of the point source laser beam. Through the laser optical path adjustment module, the beam divergence angle and the focus point position are precisely controlled to evenly distribute the power density of the laser beam in different areas. The movement path of the laser head is adjusted, the overlap rate of the scanning trajectory is optimized, and the scanning overlap ratio is controlled between 10% and 15% to improve energy utilization. Real-time feedback is used to monitor the changes in the welding temperature field and the shape of the molten pool after the beam adjustment, and finally the point source laser beam adjustment data is generated for subsequent welding operations.

[0174] Optionally, step S23 is specifically:

[0175] Step S231: extracting welding current characteristics from resistance spot welding data to obtain welding current data;

[0176] In this embodiment, real-time current waveform data is collected from the resistance spot welding equipment, and the current signal is recorded using a high-precision current sensor. The current signal is digitized by a data acquisition card, and the sampling frequency is set to 10kHz to capture subtle current changes during the welding process. Wavelet transform is used to perform spectral analysis on the signal to extract key features such as current amplitude, peak current, and duration. Based on the conductive properties of the welding material, such as steel with a resistance of 20μΩ·m, the collected current characteristics are calibrated, the noise signal is filtered out, and complete welding current characteristic data is generated.

[0177] Step S232: drawing a heat generation distribution map according to the welding current data, thereby obtaining a heat generation distribution map;

[0178] In this embodiment, the welding current data is combined with the resistance spot welding electrode contact resistance data to calculate the distribution of heat generation at each time point. The electrode contact area is set to 10mm 2 , calculate the heat generation value according to Joule's law. Use the 2D meshing tool to divide the welding area into equally spaced mesh units with a unit of 0.1 mm. 2 The heat generation value of each unit is counted, and the calculation results are plotted as a heat generation distribution map through visualization tools (such as MATLAB or Python Matplotlib library), and the high and low heat areas are reflected through color gradients.

[0179] Step S233: performing temperature field calculation on the heat generation distribution diagram to obtain welding current temperature field data;

[0180] In this embodiment, the temperature field is calculated using the heat conduction equation based on the heat generation distribution diagram. The thermal conductivity of the material is set to 50 W / (m·K), the specific heat capacity is 500 J / (kg·K), and the density is 7800 kg / m 3 , the temperature distribution in each grid unit is calculated by the finite difference method. The initial temperature is set to the ambient temperature (25°C), and the boundary condition is an adiabatic boundary. The calculation time step is set to 0.01s, and the temperature change during the welding process is iteratively simulated to obtain the welding current temperature field data. The output data is represented in matrix form, including the temperature value of each grid node.

[0181] Step S234: performing a molten pool formation simulation according to the welding current temperature field data, thereby obtaining welding current molten pool data;

[0182] In this embodiment, the melting temperature threshold of the material is set by using the welding current temperature field data. For example, the melting point of steel is 1538°C, and the molten pool is identified in the area where the temperature exceeds the threshold. The region growing algorithm is used to mark the adjacent grids that meet the melting point conditions as the molten pool area starting from the initial position of the melting point. A spatial model of the molten pool shape is constructed through a three-dimensional visualization tool (such as ANSYS or COMSOL Multiphysics), and the welding current molten pool data is generated to record the volume, shape and position of the molten pool.

[0183] Step S235: Acquire phase change temperature threshold data of automobile industry metal parts;

[0184] In this embodiment, the phase change temperature range of the metal is determined according to the material composition of the metal parts of the automobile industry, and the relevant technical data or laboratory measurement data are queried. For example, for dual-phase steel, the phase change temperature range is determined to be 720°C to 900°C. The phase change point of the material during heating and cooling is tested by differential scanning calorimetry (DSC), and the phase change start temperature and end temperature are recorded to form complete phase change temperature threshold data.

[0185] Step S236: performing heat-affected zone identification on the welding current molten pool data according to the phase change temperature threshold data of the automobile industry metal parts, thereby obtaining heat-affected zone data;

[0186] In this embodiment, the welding current molten pool data is used to identify the area around the molten pool where the temperature is within the phase change temperature range. By calculating the temperature value of each grid node, the nodes with a temperature between 720°C and 900°C are screened and connected to the boundary of the molten pool area to form a heat-affected zone. The geometric analysis tool is used to calculate the area and volume of the heat-affected zone, and the heat-affected zone data is generated and stored in the form of a polygonal mesh.

[0187] Step S237: adjusting the welding current speed of the heat affected zone data to obtain welding current speed data;

[0188] In this embodiment, the heat affected zone data is analyzed, and the dynamic relationship between the welding speed and the current is calculated to avoid an excessively large heat affected zone. The initial welding speed is set to 2 mm / s, and the speed is gradually increased by 0.5 mm / s each time, and the change in the area of ​​the heat affected zone is monitored. When the area of ​​the heat affected zone shrinks to the target range (for example, reduced to 80% of the initial area), the corresponding welding speed data is recorded. During the adjustment of the welding speed, the current output is synchronously optimized through the real-time current control module to generate the final welding current speed data.

[0189] Step S238: Perform resistance spot welding simulation according to the welding current speed data to obtain resistance spot welding simulation.

[0190] In this embodiment, the multi-physics coupling simulation of the welding process is constructed using the welding current speed data. The welding current, speed, heat distribution and other data are imported into the welding simulation software (such as Simufact or ABAQUS), and the boundary conditions and material properties of the welding area are set. The simulation is run step by step to verify the influence of the dynamic adjustment of the current speed on the welding quality. The output simulation results include the morphology of the welding pool, the distribution of the heat-affected zone and the mechanical properties parameters of the metal joint, and finally form a complete resistance spot welding simulation data.

[0191] Optionally, step S3 specifically includes:

[0192] Step S31: extracting welding morphology features and welding strength features according to welding simulation data, thereby obtaining welding morphology data and welding strength data;

[0193] In this embodiment, parameters such as the molten pool shape, welding heat affected zone and penetration depth are imported from the welding simulation data to extract the morphological features of the molten pool boundary. An image processing algorithm, such as the Canny edge detection algorithm, is used to identify the welding area boundary and extract the width, length and shape changes of the welding morphology. For welding strength feature extraction, a loading simulation is applied to the weld area through finite element analysis software. The loading method is vertical stretching, and the loading range is 10MPa to 200MPa, which is gradually increased to measure the maximum bearing strength of the weld. The maximum stress value before the weld breaks and the corresponding plastic deformation are recorded as welding strength feature data, and the output welding morphology data and strength data are stored in a table form.

[0194] Step S32: performing surface crack detection on the welding morphology data to obtain welding surface crack data;

[0195] In this embodiment, an ultrasonic detector is used, and a probe with a frequency of 20MHz is set to scan the welding surface to ensure that the detection depth covers the surface area of ​​the welding heat affected zone. The waveform data generated by the scan is analyzed, and the abnormal signal area is extracted by the short-time Fourier transform algorithm, and the position where the signal amplitude exceeds the set threshold (for example, the signal reflection amplitude is 5 times higher than the noise baseline) is marked as a potential crack point. Combined with the microscopic image acquisition equipment to obtain a microscopic image of the welding surface, the potential crack points marked by the ultrasonic wave are further verified. Finally, the confirmed crack position, length and width are generated into the welding surface crack data, and output in the form of a two-dimensional plane map and a crack data table.

[0196] Step S33: performing weld porosity detection on the welding strength data to obtain weld porosity data;

[0197] In this embodiment, X-ray imaging technology is used to perform nondestructive testing on the inside of the weld. The X-ray tube voltage is set to 120kV, the exposure time is 5ms, and high-resolution images of the weld area are collected. Through the image contrast algorithm, the uneven brightness area inside the weld is marked as a potential pore area. The volume of the marked area is reconstructed in three dimensions, and the pore diameter and volume are calculated through boundary detection. If the volume exceeds 0.1mm 3 Or if the diameter is greater than 0.5mm, it is marked as weld pore. Output weld pore data, including location, quantity and size parameters, and store them in the form of table and 3D point cloud.

[0198] Step S34: integrating welding defects according to the welding surface crack data and the weld porosity data, thereby obtaining welding defect data.

[0199] In this embodiment, the welding surface crack data and weld porosity data are input into the welding defect integration algorithm. Using the spatial coordinate matching technology, the two types of data are superimposed and analyzed to check whether the porosity area intersects or approaches the crack area, and the threshold of the proximity distance is set to 0.5mm. Through data screening, the interactive area between the porosity and the crack is marked as a composite defect and assigned a unique number. All defect data are classified and counted, and defect distribution maps and statistical tables are generated according to the position (weld center or edge), size (length or volume) and type (crack, porosity or composite defect). The final welding defect data includes the spatial coordinates, geometric features and type information of the defect, which are stored in the form of a database and support subsequent data calls and analysis.

[0200] Optionally, step S32 is specifically:

[0201] Step S321: extracting weld contour features from welding shape data to obtain weld contour data;

[0202] In this embodiment, the boundary point set of the weld is located by an edge detection algorithm (such as Canny edge detection), and the boundary points are connected according to the spatial coordinates to form a weld contour line. The weld contour line data is curve fitted, and the cubic B-spline algorithm is used to ensure the smoothness of the fitting. A threshold is set to define the contour feature, for example, the maximum allowable deviation of the weld width is 0.2 mm, and the height fluctuation does not exceed 0.5 mm. The cross-sectional contour area, symmetry, and geometric parameters of the weld are calculated, all feature data are saved as a contour data file, and the weld contour is visualized in a three-dimensional rendering manner.

[0203] Step S322: dividing the weld direction according to the weld contour data, thereby obtaining vertical direction weld data and parallel direction weld data;

[0204] In this embodiment, the main axis direction of the weld is first calculated, and the main direction vector of the weld morphology is extracted by the PCA (principal component analysis) method. According to the plane coordinate system where the weld is located, the weld direction is divided into a vertical direction and a parallel direction according to the angle with the reference coordinate system. Specifically defined as: weld segments with an angle less than 10° are classified as parallel direction welds, and weld segments with an angle close to 90° and an allowable deviation range of ±5° are classified as vertical direction welds. The classified data are saved as two independent sets of weld direction data files, and statistical information of the starting point, end point and length of each weld segment is attached.

[0205] Step S323: performing longitudinal magnetization on the vertical weld data to obtain longitudinal magnetization data;

[0206] In this embodiment, a longitudinal magnetization coil is selected, the coil length is 1.2 times the length of the weld section, and the current intensity is set to 400A. The magnetization device is evenly arranged along the main direction of the weld section, and the magnetic field intensity of the magnetization area is controlled to be not less than 800A / m. The magnetization coil is excited by segmented pulse current to ensure that the magnetic field covers the entire weld area. A Hall effect magnetic field sensor is used to detect the magnetization intensity of the weld area to ensure that the magnetization effect reaches the preset range. After completion, the detected magnetization intensity data is recorded as a longitudinal magnetization data file.

[0207] Step S324: performing transverse magnetization on the parallel direction weld data to obtain transverse magnetization data;

[0208] In this embodiment, a transverse magnetization fixture is selected, and the fixture opening width is adjusted to 1.5 times the weld width to ensure full coverage of the weld area. The fixture is energized with an alternating current (frequency of 50 Hz, current intensity of 600 A) to generate a transverse magnetic field acting on the weld area. The magnetized weld area is subjected to real-time magnetic field distribution detection, and a two-dimensional magnetic induction scanning device is used to record the uniformity of the magnetic field inside the weld. If the magnetic field uniformity deviation exceeds 10%, the current intensity is adjusted for compensation until the uniform distribution requirement is met, and the final detection data is saved as a transverse magnetization data file.

[0209] Step S325: performing dry magnetic powder spraying according to the longitudinal magnetization data, thereby obtaining dry magnetic powder spraying data;

[0210] In this embodiment, a dry magnetic powder spraying device is selected according to the longitudinal magnetization data, and dry magnetic powder spraying is performed on the weld area. The nozzle diameter of the spraying device is 2mm, the spraying pressure is set to 0.2MPa, and the spraying distance is 25mm. During the spraying process, the magnetic powder is evenly applied, and the flow rate is manually controlled to ensure that the magnetic powder coverage is not less than 95%. The weld surface after spraying is detected using a microscope, and the uniformity of the spraying thickness is recorded, and the thickness deviation is controlled within ±0.1mm. After completion, the spraying effect data is stored as a dry magnetic powder spraying data file in the form of an image and a parameter table.

[0211] Step S326: performing wet magnetic powder spraying according to the transverse magnetization data, thereby obtaining wet magnetic powder spraying data;

[0212] In this embodiment, a wet magnetic powder spraying device is used according to the transverse magnetization data to configure a magnetic powder suspension (magnetic powder concentration is 10g / L, dispersant content is 0.1%). The suspension is evenly sprayed on the weld area through a spray gun, the spraying pressure is set to 0.3MPa, the spraying distance is 20mm, and the spraying speed is 10cm / s. After the spraying is completed, an ultraviolet lamp is used to irradiate the weld area to observe the distribution of the magnetic powder. If the distribution is broken or concentrated, adjust the suspension concentration and repeat the spraying until the distribution uniformity reaches the standard. All records in the spraying process are stored as a wet magnetic powder spraying data file.

[0213] Step S327: performing magnetic powder spray fusion according to the dry magnetic powder spray data and the wet magnetic powder spray data, thereby obtaining magnetic powder spray data;

[0214] In this embodiment, the weld surface data generated by the two spraying methods are superimposed in the same coordinate system using image overlay technology to analyze the consistency of spray thickness and uniformity. The judgment threshold is set to a spray thickness difference of no more than 0.2 mm. If any position exceeds the threshold, it is marked as a spray abnormality point. Combined with the weld direction data, the magnetic powder distribution of the abnormal point area is recalculated and the data is corrected. The final output magnetic powder spraying data includes a spray thickness distribution map, an abnormal point statistics table, and a corrected weld spray distribution map.

[0215] Step S328: Perform surface crack defect visualization on the magnetic powder spraying data to obtain welding surface crack data.

[0216] In this embodiment, the sprayed surface is scanned by a fluorescence detection device, and the excitation wavelength is set to 365nm to observe the magnetic powder aggregation phenomenon at the crack position. The crack features in the fluorescent image are segmented by an image processing algorithm to extract the crack length, width and depth information. The crack length threshold is set to 0.5mm, the width threshold is set to 0.1mm, and the data exceeding the threshold is recorded as a valid crack. The generated welding surface crack data includes the spatial coordinates, geometric parameters and appearance images of the crack, and is stored in the form of a standardized file.

[0217] Optionally, step S33 is specifically:

[0218] Step S331: performing strength statistics on the welding strength data to obtain low-strength welding data;

[0219] In this embodiment, it is necessary to use a precisely calibrated welding strength testing device to collect welding strength data. Incremental tensile force is applied to the welding area, and the force value at the breaking point is recorded in Newton (N) and stored as a strength data array. Classification standards are used in the statistical process to mark welding strength values ​​below a set threshold (e.g., 200N, based on material properties and welding requirements) as low-strength welding data. Point-by-point scanning method is used in the statistical analysis process to distinguish the data, and the low-strength welding data is saved in a dedicated database for use in subsequent steps.

[0220] Step S332: performing region identification on the welding simulation data according to the low-strength welding data, thereby obtaining low-strength welding region data;

[0221] In this embodiment, a high-resolution laser scanner (with an accuracy of 0.01 mm) and a welding simulation image are combined for superposition comparison, and the coordinate point range marked in the low-intensity data is used to automatically identify and extract the welding area through an edge recognition algorithm (such as the Canny algorithm). The identified low-intensity welding area data is output in a matrix form, including the starting and ending spatial coordinates (such as [0,0,0] to [10,5,2]).

[0222] Step S333: performing X-ray imaging according to the low-intensity welding area data, thereby obtaining an X-ray low-intensity welding area image;

[0223] In this embodiment, an industrial X-ray imaging device (with a radiation source energy set to 120 kV and a focal spot size of 0.4 mm) is used to image the area marked by the low-intensity welding area data. The device must first adjust the angle between the source and the detector according to the three-dimensional coordinates of the welding area to ensure complete coverage of the welding area. The imaging process collects data in real time and generates clear X-ray images. The image resolution is set to 1024×1024 pixels, and the image data is stored in a dedicated server in TIFF format.

[0224] Step S334: performing a binary conversion on the X-ray low-intensity welding area image to obtain a binary image of the X-ray low-intensity welding area;

[0225] In this embodiment, an image processing tool (such as OpenCV library) is used to grayscale the obtained X-ray image, and the image pixel intensity value is standardized to a range of 0 to 255. Then, the binarization threshold is set to 127 (set according to the X-ray absorption characteristics of the weld material), and the area with a pixel value higher than the threshold is marked as 1, and the area with a pixel value lower than the threshold is marked as 0, and a binarized image is generated and stored as matrix data.

[0226] Step S335: performing gradient calculation on the binary image of the low-intensity X-ray welding area to obtain binary image gradient data, and performing pore gradient statistics based on the binary image gradient data to obtain a high-gradient binary pore image;

[0227] In this embodiment, the Sobel operator is used to calculate the gradient of the binary image, and the window size is set to 3×3 to calculate the gradient strength and direction of each pixel. According to the gradient strength threshold set to 50 (the boundary value measured in the experiment), the area with high gradient strength is counted, and a marked image of the high gradient area is generated (the resolution is the same as the original image). Figure 1 The marked high gradient area reflects the distribution of welding pores and is stored in the form of a binary image.

[0228] Step S336: performing corrosion according to the high gradient binary pore image, so as to obtain a high gradient binary pore corrosion image;

[0229] In this embodiment, the high gradient binary pore image is processed by morphological corrosion operation, the corrosion kernel size is set to 3×3 (set according to the experimental results of pore size), and the number of corrosion is 1. The corrosion operation reduces the boundary of the high gradient area by taking the minimum value of the neighborhood pixel value. The eroded image is stored as a high gradient binary pore corrosion image.

[0230] Step S337: dilating the high gradient binary pore corrosion image to obtain a high gradient binary pore dilation image;

[0231] In this embodiment, the eroded image is processed by a morphological dilation operation, the dilation kernel size is set to 3×3, and the dilation times are 1. The dilation operation restores the smaller pore area after the erosion to its original size and retains its shape characteristics by taking the maximum value of the neighborhood pixel value. The dilated image is stored as a high-gradient binary pore dilation image.

[0232] Step S338: Identify the pore positions based on the high-gradient binary pore expansion image to obtain weld pore data.

[0233] In this embodiment, the pore position is identified based on the expansion image, and the center coordinates and area of ​​the pores are determined using the connected domain analysis algorithm (8-neighborhood connectivity). The specific operation includes traversing the area where the image pixel value is 1, and calculating the centroid (expressed in pixel coordinates) and area (expressed in number of pixels) for each connected area. The identified pore position data is output in a table form, including the pore number, center coordinates (such as [15,30]) and area information.

[0234] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0235] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A laser welding system control method for metal accessories, characterized in that: The following steps are involved: Step S1: Obtaining automobile industry metal parts data, and performing welding demand analysis based on the automobile industry metal parts data, thereby obtaining welding demand data; Step S2: performing laser welding feature extraction and resistance spot welding feature extraction according to the welding requirement data, thereby obtaining laser welding data and resistance spot welding data; performing welding simulation according to the laser welding data and the resistance spot welding data, thereby obtaining welding simulation data; Step S3: extracting welding morphology features and welding strength features according to the welding simulation data, thereby obtaining welding morphology data and welding strength data; performing welding defect detection according to the welding morphology data and welding strength data, thereby obtaining welding defect data; Step S4: acquiring laser welding system data; optimizing laser welding parameters of the laser welding system data according to the welding defect data, thereby obtaining laser welding optimization parameters; Step S5: adjusting the laser welding process of the welding simulation data according to the laser welding optimization parameters, thereby obtaining laser welding process adjustment data, and uploading the data to the laser welding system to execute the laser welding process control task.

2. The laser welding system control method for metal accessories according to claim 1, characterized in that: Step S1 is specifically as follows: Step S11: acquiring automobile industry metal parts data, and performing welding area feature extraction based on the automobile industry metal parts data, thereby obtaining automobile industry metal parts welding area data; Step S12: performing narrow area identification according to the welding area data of automobile industry metal parts, thereby obtaining the welding narrow area data of automobile industry metal parts; Step S13: performing multi-curvature surface analysis on the welding narrow area data of automobile industry metal accessories, thereby obtaining multi-curvature surface data of automobile industry metal accessories; Step S14: performing spatial complexity evaluation according to the multi-curvature surface data of automobile industry metal accessories, thereby obtaining the spatial complexity of automobile industry metal accessories; Step S15: Perform welding method requirement adaptation evaluation according to the spatial complexity of automobile industry metal parts, so as to obtain welding requirement data.

3. The laser welding system control method for metal accessories according to claim 2, characterized in that: Step S13 is specifically as follows: Step S131: performing point cloud neighborhood calculation on the narrow area data of welding of automobile industry metal parts, thereby obtaining point cloud neighborhood data; Step S132: performing weighted averaging according to the point cloud neighborhood data to obtain point cloud neighborhood weighted average data, and updating the point cloud neighborhood data according to the point cloud neighborhood weighted average data to obtain point cloud neighborhood update data; Step S133: constructing a smooth surface model of the metal accessory according to the point cloud neighborhood update data, thereby obtaining a smooth surface model of the metal accessory; Step S134: performing Gaussian curvature calculation on the smooth surface model of the metal accessory to obtain Gaussian curvature data; Step S135: dividing the smooth surface model of the metal accessory into curvature regions according to the Gaussian curvature data, thereby obtaining flat curvature region data and curved curvature region data; Step S136: performing curvature gradient analysis on the flat curvature region data, thereby obtaining flat curvature region gradient data; Step S137: performing wavelet transformation on the curved curvature region data, thereby obtaining the curved curvature region wavelet scale data; Step S138: integrating the multi-curvature surfaces of automobile industry metal parts according to the gradient data of the flat curvature region and the wavelet scale data of the curved curvature region, thereby obtaining the multi-curvature surface data of automobile industry metal parts.

4. The laser welding system control method for metal accessories according to claim 1, characterized in that: Step S2 is specifically as follows: Step S21: performing laser welding feature extraction and resistance spot welding feature extraction according to welding requirement data, thereby obtaining laser welding data and resistance spot welding data; Step S22: performing laser welding simulation on the laser welding data to obtain laser welding simulation data; Step S23: performing resistance spot welding simulation on the resistance spot welding data, thereby obtaining resistance spot welding simulation data; Step S24: performing welding simulation integration according to the laser welding simulation data and the resistance spot welding simulation data, so as to obtain welding simulation data.

5. The laser welding system control method for metal accessories according to claim 4, characterized in that: Step S22 is specifically as follows: Step S221: extracting laser power characteristics and laser beam characteristics from laser welding data, thereby obtaining laser power data and laser beam data; Step S222: constructing a laser beam distribution model according to the laser beam data, thereby obtaining a laser beam distribution model; Step S223: identifying the maximum energy density of the focus according to the laser beam distribution model, thereby obtaining the maximum energy density data of the focus; Step S224: performing laser heat source identification according to the focus maximum energy density data and the laser power data, thereby obtaining laser heat source data; Step S225: drawing a temperature field distribution map according to the laser heat source data, thereby obtaining a temperature field distribution map of the laser heat source; Step S226: Identify the irregular temperature gradient region according to the temperature field distribution diagram of the laser heat source, thereby obtaining the irregular temperature gradient region data; Step S227: performing welding pool unevenness detection on the automotive industry metal parts data according to the irregular temperature gradient area data, thereby obtaining welding pool unevenness data; Step S228: performing point source laser beam adjustment optimization according to the welding molten pool unevenness data, thereby obtaining point source laser beam adjustment data; Step S229: Perform laser welding simulation according to the point source laser beam adjustment data to obtain laser welding simulation data.

6. The laser welding system control method for metal accessories according to claim 5, characterized in that: Step S228 is specifically as follows: Mark the insufficient penetration area according to the uneven data of the welding pool, so as to obtain the insufficient penetration area data; Mark the over-melting area according to the uneven data of the welding molten pool, so as to obtain the over-melting area data; The laser power is increased for the insufficient melting depth area data to obtain the laser power data; The laser beam scanning speed is increased for the over-melting area data, thereby obtaining the laser beam scanning speed data; The point source laser beam adjustment optimization integration is performed according to the laser power data and the laser beam scanning speed data, so as to obtain the point source laser beam adjustment data.

7. The laser welding system control method for metal accessories according to claim 4, characterized in that: Step S23 is specifically as follows: Step S231: extracting welding current characteristics from resistance spot welding data to obtain welding current data; Step S232: drawing a heat generation distribution map according to the welding current data, thereby obtaining a heat generation distribution map; Step S233: performing temperature field calculation on the heat generation distribution diagram to obtain welding current temperature field data; Step S234: performing a molten pool formation simulation according to the welding current temperature field data, thereby obtaining welding current molten pool data; Step S235: Acquire phase change temperature threshold data of automobile industry metal parts; Step S236: performing heat-affected zone identification on the welding current molten pool data according to the phase change temperature threshold data of the automobile industry metal parts, thereby obtaining heat-affected zone data; Step S237: adjusting the welding current speed of the heat affected zone data to obtain welding current speed data; Step S238: Perform resistance spot welding simulation according to the welding current speed data to obtain resistance spot welding simulation.

8. The laser welding system control method for metal accessories according to claim 1, characterized in that: Step S3 is specifically as follows: Step S31: extracting welding morphology features and welding strength features according to welding simulation data, thereby obtaining welding morphology data and welding strength data; Step S32: performing surface crack detection on the welding morphology data to obtain welding surface crack data; Step S33: performing weld porosity detection on the welding strength data to obtain weld porosity data; Step S34: integrating welding defects according to the welding surface crack data and the weld porosity data, thereby obtaining welding defect data.

9. The laser welding system control method for metal accessories according to claim 8, characterized in that: Step S32 is specifically as follows: Step S321: extracting weld contour features from welding shape data to obtain weld contour data; Step S322: dividing the weld direction according to the weld contour data, thereby obtaining vertical direction weld data and parallel direction weld data; Step S323: performing longitudinal magnetization on the vertical weld data to obtain longitudinal magnetization data; Step S324: performing transverse magnetization on the parallel direction weld data to obtain transverse magnetization data; Step S325: performing dry magnetic powder spraying according to the longitudinal magnetization data, thereby obtaining dry magnetic powder spraying data; Step S326: performing wet magnetic powder spraying according to the transverse magnetization data, thereby obtaining wet magnetic powder spraying data; Step S327: performing magnetic powder spray fusion according to the dry magnetic powder spray data and the wet magnetic powder spray data, thereby obtaining magnetic powder spray data; Step S328: Perform surface crack defect visualization on the magnetic powder spraying data to obtain welding surface crack data.

10. The laser welding system control method for metal accessories according to claim 8, characterized in that: Step S33 is specifically as follows: Step S331: performing strength statistics on the welding strength data to obtain low-strength welding data; Step S332: performing region identification on the welding simulation data according to the low-strength welding data, thereby obtaining low-strength welding region data; Step S333: performing X-ray imaging according to the low-intensity welding area data, thereby obtaining an X-ray low-intensity welding area image; Step S334: performing a binary conversion on the X-ray low-intensity welding area image to obtain a binary image of the X-ray low-intensity welding area; Step S335: performing gradient calculation on the binary image of the low-intensity X-ray welding area to obtain binary image gradient data, and performing pore gradient statistics based on the binary image gradient data to obtain a high-gradient binary pore image; Step S336: performing corrosion according to the high gradient binary pore image, so as to obtain a high gradient binary pore corrosion image; Step S337: dilating the high gradient binary pore corrosion image to obtain a high gradient binary pore dilation image; Step S338: Identify the pore positions based on the high-gradient binary pore expansion image to obtain weld pore data.

Images