Online monitoring method and platform for weld quality in sheet metal welding process

Abstract

The present invention discloses an online monitoring method and platform for weld quality in a sheet metal welding process, and relates to the field of welding technology. The method comprises: determining a welding call sequence through welding sequence analysis, extracting joint surface parameters of a reference sheet metal part and a first sheet metal part, performing material change synchronization analysis, and generating a welding parameter sequence. In the process of using welding parameters to control the welding of a battery-type welder, quality monitoring is performed based on a weld image and weld defect nodes are output. Welding control and defect identification of the remaining sheet metal parts are performed according to the welding call sequence, and a weld defect node set is output. The technical problem that the prior art cannot monitor weld defects in a sheet metal welding process in real time and accurately is solved, and the weld quality in a sheet metal welding process is monitored in real time, and the position and type of weld defects are accurately located, so that welding defects can be discovered and corrected in a timely manner, and the welding quality can be ensured to meet the technical effect of standards.

Description

Technical Field

[0001] The present application relates to the field of welding technology, and in particular to an online monitoring method and platform for weld quality in a sheet metal welding process. Background Art

[0002] Welding is an indispensable process in the manufacturing industry, especially in the production of sheet metal, body frames and other products. The quality of welds directly affects the strength, appearance and service life of the product, and weld defects (such as cracks, pores, over-welding, etc.) are common quality problems, which may cause safety hazards in serious cases. At present, quality monitoring in the welding process relies on manual inspection or basic welding control systems, but these methods have problems such as poor real-time performance, insufficient accuracy and frequent manual intervention, and cannot effectively cope with complex welding tasks. However, the existing technology still faces the problem that parameter changes during the welding process are difficult to monitor and optimize synchronously. There is an urgent need for an accurate, real-time and efficient weld quality monitoring system to improve the automation control of the welding process, reduce defects, reduce production costs, and improve overall efficiency.

[0003] In the current related technologies, there is a technical problem that weld defects in the sheet metal welding process cannot be monitored in real time and accurately. Summary of the invention

[0004] The present application solves the technical problem in the prior art that weld defects in the sheet metal welding process cannot be monitored in real time and accurately by providing an online monitoring method and platform for weld quality in the sheet metal welding process.

[0005] The present application provides an online monitoring method for weld quality in a sheet metal welding process, comprising:

[0006] A welding order analysis is performed on a workpiece to be welded to obtain a welding call sequence, wherein the workpiece to be welded is obtained by welding K sheet metal parts; after extracting a reference sheet metal part and a first sheet metal part based on the welding call sequence, joint surface parameters of the reference sheet metal part and the first sheet metal part are collected respectively to obtain a reference joint surface parameter sequence and a first joint surface parameter sequence; a material change synchronization analysis is performed on the reference joint surface parameter sequence and the first joint surface parameter sequence, and a welding parameter sequence is output based on the analysis result; in the process of using the welding parameter sequence to control a battery-type welder to perform joint surface welding on the reference sheet metal part and the first sheet metal part, a pre-built weld quality detection model is run to synchronously perform online monitoring of weld quality based on weld images, and multiple weld defect nodes are output; with the welding call sequence as a constraint, welding control and weld defect identification are performed on the remaining K-2 sheet metal parts, and a weld defect node set is output.

[0007] This application provides an online monitoring platform for weld quality in sheet metal welding processes, including:

[0008] A welding call sequence acquisition module, the welding call sequence acquisition module is used to perform welding sequence analysis on the workpiece to be welded to obtain a welding call sequence, wherein the workpiece to be welded is obtained by welding K sheet metal parts; a joint surface parameter acquisition module, the joint surface parameter acquisition module is used to extract a reference sheet metal part and a first sheet metal part based on the welding call sequence, and then respectively collect joint surface parameters of the reference sheet metal part and the first sheet metal part to obtain a reference joint surface parameter sequence and a first joint surface parameter sequence; a synchronization analysis module, the synchronization analysis module is used to analyze the reference joint surface parameter sequence and the first joint surface parameter sequence A material change synchronization analysis is performed, and a welding parameter sequence is output based on the analysis results; a weld quality monitoring module, wherein the weld quality monitoring module is used to run a pre-built weld quality detection model to synchronously perform online monitoring of weld quality based on a weld image during the process of using the welding parameter sequence to control a battery-type welder to perform joint surface welding on the reference sheet metal part and the first sheet metal part, and output a plurality of weld defect nodes; a weld defect node set output module, wherein the weld defect node set output module is used to perform welding control and weld defect identification on the remaining K-2 sheet metal parts with the welding call sequence as a constraint, and output a weld defect node set.

[0009] The method and platform for online monitoring of weld quality in the sheet metal welding process proposed in this application first determine the welding call sequence through welding sequence analysis, extract the joint surface parameters of the reference sheet metal and the first sheet metal, perform material change synchronization analysis, and generate a welding parameter sequence. In the process of using welding parameters to control the battery-type welder, quality monitoring is performed based on the weld image and weld defect nodes are output. Finally, welding control and defect identification of the remaining sheet metal parts are performed according to the welding call sequence, and a set of weld defect nodes is output, achieving the technical effect of real-time monitoring of weld quality in the sheet metal welding process and accurately locating the location and type of weld defects, so that welding defects can be discovered and corrected in a timely manner to ensure that the welding quality meets the standards. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in this application to illustrate the operations performed by the platform according to the embodiment of the present application. It should be understood that the previous or following operations are not necessarily performed precisely in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0011] Figure 1 A schematic flow chart of a method for online monitoring of weld quality in a sheet metal welding process provided in an embodiment of the present application.

[0012] Figure 2 A schematic diagram of the structure of an online weld quality monitoring platform for a sheet metal welding process provided in an embodiment of the present application.

[0013] Explanation of the reference numerals: welding call sequence acquisition module 10 , joint surface parameter acquisition module 20 , synchronization analysis module 30 , weld quality monitoring module 40 , weld defect node set output module 50 . DETAILED DESCRIPTION

[0014] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0015] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0016] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, platform, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0017] The present application embodiment provides a method for online monitoring of weld quality in a sheet metal welding process, such as Figure 1 As shown, the method includes:

[0018] Step S100, performing welding sequence analysis on the workpiece to be welded to obtain a welding call sequence, wherein the workpiece to be welded is obtained by welding K sheet metal parts. Specifically, analyzing the workpiece to be welded consisting of K sheet metal parts, collecting information such as the geometric shape, material properties, and surface quality of each sheet metal part, and constructing K sub-component models, analyzing the thickness and size change scale based on the sub-component model, determining the type of welding joint by splicing simulation, and assigning points for welding difficulty and clamping difficulty respectively, weighted fusion of relevant data to obtain K clamping stabilities, and then serializing the clamping stability and locating the reference sheet metal according to the sorting extreme value, and comprehensively analyzing the welding sequence of adjacent sheet metal parts based on the clamping stability, welding joint type, and overall structural requirements of the workpiece, thereby generating a welding call sequence that clearly defines the welding order of the K sheet metal parts, and providing precise guidance for subsequent welding operations.

[0019] In a possible implementation, a welding sequence analysis is performed on a workpiece to be welded to obtain a welding call sequence, wherein the workpiece to be welded is obtained by welding K sheet metal parts, and step S100 further includes step S110, performing CAD modeling on the K sheet metal parts to obtain K sub-component models. Specifically, the K sheet metal parts are modeled one by one using CAD software, and in the modeling process, they are accurately drawn according to the actual size and shape characteristics of the sheet metal parts. For sheet metal parts with complex surfaces or special shapes, the surface modeling and solid modeling tools of the CAD software are combined to ensure that the model can accurately reflect its geometric form, thereby obtaining K sub-component models. The model not only contains the appearance shape information of the sheet metal parts, but also provides basic data such as volume, surface area, and center of gravity position for subsequent analysis.

[0020] Step S120, interactively obtain K material information of the K sheet metal parts. Specifically, by interacting with the material database or directly reading the identification information of the sheet metal parts, obtain the material information corresponding to each of the K sheet metal parts, including the type of material (such as aluminum alloy, carbon steel, stainless steel, etc.), the grade of the material (such as 6061 aluminum alloy, Q235 carbon steel, etc.), the mechanical property parameters of the material (such as yield strength, tensile strength, elongation, etc.) and the physical property parameters of the material (such as thermal conductivity, thermal expansion coefficient, etc.). Such material information plays a key role in evaluating the behavior of sheet metal parts during welding and determining appropriate welding process parameters.

[0021] Step S130, based on the K sub-component models and K material information, the reliability of sheet metal welding clamping is evaluated, and K clamping stability is output. Specifically, based on the constructed K sub-component models, thickness feature analysis is performed. Using the measurement tool of the CAD software, the thickness data of different parts of each sheet metal part are obtained to determine the thickness variation scale, that is, the difference between the maximum thickness and the minimum thickness and the regional distribution of the thickness variation. For example, for some sheet metal parts with reinforcing ribs or stamping forming parts, the thickness variation may be more complicated. At the same time, the dimensional variation scale is analyzed, including the values ​​of the main dimensions such as the length, width, and height of the sheet metal part and the proportional relationship between them, and the irregularity of the shape, such as the degree of bending, the number and angle of corners, etc., needs to be considered. Thickness and dimensional characteristics directly affect the selection of the focus point and the distribution of the clamping force during welding clamping. The K sub-component models are spliced ​​and simulated according to the final assembly form of the workpiece to be welded, so as to determine the types of K welding joints. Common types of welding joints include butt joints, lap joints, corner joints, T-joints, etc. For different types of welding joints, welding difficulty is scored through a pre-established welding joint scoring table. This scoring table is based on a large amount of welding practice experience and welding process research, and takes into account factors such as the sensitivity of different joint types to welding parameters during welding, the difficulty of controlling welding deformation, and the difficulty of ensuring weld quality. For example, butt joints have higher requirements in ensuring weld flatness and uniformity of penetration, while lap joints need to pay special attention to the penetration of the welding current on the upper sheet metal and avoid excessive welding of the lower sheet metal. According to these factors, each welding joint is assigned a corresponding difficulty score to obtain K welding joint scores. Using the K material information obtained, the clamping difficulty is scored through the welding material scoring table. Different materials have different surface friction coefficients, hardness, toughness and other characteristics, which will affect the friction and clamping force transmission effect between the clamping tool and the sheet metal. For example, aluminum alloy materials are relatively soft and prone to indentation when clamped, requiring more sophisticated clamping force control, and their clamping difficulty scores are relatively high; while carbon steel materials are relatively hard and relatively stable when clamped, but may require greater clamping force, and also have corresponding clamping difficulty scores. Finally, the K thickness variation scales, K size variation scales, K welding joint scores, and K welding material scores are weighted and fused. The weights are determined based on the importance of each factor on the reliability of welding clamping, and are summarized through a large number of experiments and engineering practice experience. After weighted fusion calculations, K clamping stabilities are output, and these clamping stability values ​​can intuitively reflect the reliability of each sheet metal part during the welding clamping process.

[0022] Step S140, serialize the K clamping stabilities, and locate the reference joint surface based on the sorting extreme value. Specifically, the obtained K clamping stabilities are serialized and can be arranged in ascending or descending order according to the numerical value. In the serialized sequence, the reference joint surface is located based on the sorting extreme value (such as the maximum value or the minimum value). For example, if the joint surface corresponding to the sheet metal part with the highest clamping stability is selected as the reference joint surface, it is because such a joint surface is easier to ensure the stability and accuracy of welding at the initial stage of welding. This reference joint surface will serve as the starting reference point of the entire welding process, and the subsequent welding sequence will revolve around it.

[0023] Step S150, taking the reference sheet metal as the starting point, the welding order analysis of adjacent sheet metal parts is performed according to the K clamping stabilities, and the welding call sequence is generated. Specifically, taking the determined reference sheet metal as the starting point, the welding order analysis of adjacent sheet metal parts is performed according to multiple factors such as K clamping stabilities, welding joint types, and overall structural requirements of the workpiece to be welded. In the analysis process, priority is given to adjacent welding of sheet metal parts with similar clamping stability and better matching welding joint types. For example, for two sheet metal parts with high clamping stability and both of them are butt joints, welding can be performed first, which can reduce the number of adjustments and error accumulation during the welding process. At the same time, the structural integrity and force uniformity of the entire workpiece to be welded should also be considered to avoid welding deformation or stress concentration caused by unreasonable welding order. Through such a step-by-step analysis and decision-making process, a reasonable welding call sequence is finally generated, which clearly stipulates the order of K sheet metal parts in the welding process, provides accurate and scientific guidance for subsequent welding operations, and helps to improve welding quality and efficiency and reduce the generation of welding defects.

[0024] In a possible implementation, the reliability of sheet metal welding clamping is evaluated based on the K sub-component models and K material information, and K clamping stabilities are output. Step S130 further includes step S131, performing thickness feature analysis based on the K sub-component models to obtain K thickness variation scales and K size variation scales. Specifically, based on the constructed K sub-component models, a CAD analysis tool is used to perform thickness feature analysis on each sub-component model, and multiple measurement points are selected in various areas of the model. These measurement points need to cover different parts of the sheet metal, such as plane areas, curved areas, edge areas, and special structural areas where thickness variations may exist. Then, the thickness value at the measurement point in each sub-component model is calculated, and its maximum thickness value and minimum thickness value are determined. The difference between the two is the thickness variation scale of the sheet metal. For example, for a sheet metal part with a stamped groove, the thickness of the bottom of the groove may be thinner, while the surrounding area is relatively thicker. This large thickness variation will make it difficult to ensure that the entire sheet metal part is evenly stressed under the clamping force during welding and clamping, and it is easy to cause local deformation or loose clamping, thereby increasing the difficulty of clamping. By analyzing the K sub-component models one by one, K thickness variation scales are finally obtained. These data intuitively reflect the unevenness of the thickness of each sheet metal part. Use CAD analysis tools to analyze the dimensional characteristics of the K sub-component models. Determine the main dimensional parameters of each sheet metal part, including basic dimensional values ​​such as length, width, and height. Further analyze its dimensional changes. For example, for irregularly shaped sheet metal parts, measure the parameters such as the radius of curvature, bending angle, and irregularity of the edge of different parts, and determine its dimensional change scale based on the comprehensive parameters. For example, a sheet metal part with a complex curved surface has a large variation in the radius of curvature of different parts. This not only requires a special welding process to adapt to the welding of the curved surface, but also requires a suitable fixture to be designed according to its shape characteristics during the clamping process to ensure that the sheet metal part can be stably fixed on the welding table to prevent displacement or deformation during the welding process. Through a comprehensive analysis of K sub-component models, K dimensional variation scales are obtained. These data help to fully understand the geometric complexity and dimensional diversity of each sheet metal part.

[0025] Step S132, after splicing the K sub-component models with the workpiece to be welded as a constraint, collect and obtain the K welding joint types of the K sub-component models. Specifically, with the final assembly form of the workpiece to be welded as a constraint condition, the K sub-component models are spliced ​​and simulated in a virtual environment (such as the assembly module of the CAD software). During the splicing process, the connection method between each sheet metal part is analyzed to determine the K welding joint types of the K sub-component models. The welding joint types include common butt joints, corner joints, T-joints and other basic types, and the specific shape characteristics of the joints are analyzed, such as straight seam butt joints and bevel seam butt joints in butt joints, right-angle corner joints and obtuse corner joints in corner joints, etc. Different joint types and shapes have a significant impact on welding processes and clamping requirements. For example, it is relatively easy to control the straightness and uniformity of the weld depth of a straight butt joint during welding, but it is necessary to ensure that the butt edges of the two sheet metal parts are completely flush during clamping and that no relative displacement occurs during welding; and for a bevel butt joint, since the weld is in an inclined state, the change in welding angle needs to be considered in the setting of welding parameters, and the sheet metal parts also need to be prevented from sliding in the inclined direction during clamping. Through a detailed analysis of the splicing model, the type and shape characteristics of the welded joints of each sub-component model can be accurately identified.

[0026] Step S133, using the K welding joint types to traverse the welding joint scoring table to assign welding difficulty points, and obtain K welding joint scores. Specifically, the identified K welding joint types are used to compare the welding joint scoring table established in advance to assign welding difficulty points. The welding joint scoring table is constructed based on a large amount of welding engineering practical experience and welding process research results. In the scoring process, various factors are considered, such as the complexity of the welding process of the joint, the difficulty of weld quality control, and the sensitivity of welding deformation. For some joint types with high welding process requirements, such as T-joints, since they need to ensure the weld quality in three directions at the same time during welding, and are prone to deformation problems caused by welding stress concentration, they will be assigned a higher welding difficulty score; while for relatively simple butt joints, if they are straight seam butt joints and the thickness of the sheet metal is uniform, their welding difficulty scores are relatively low. By comparing each welding joint type with the various standards in the scoring table, the K welding joints are assigned corresponding welding joint scores, and the scores can quantitatively reflect the difficulty of each welding joint in the welding process.

[0027] Step S134, using the K material information to traverse the welding material scoring table to assign points for clamping difficulty, and obtain K welding material scores. Specifically, using the K material information previously obtained, including parameters such as material type, brand, mechanical properties and physical properties, traverse the welding material scoring table to assign points for clamping difficulty. The welding material scoring table is formulated according to the characteristics of different materials in the welding clamping process. For example, for materials with higher hardness, such as certain alloy steels, although they are not easy to deform during clamping, due to their high surface hardness, the friction between them and the clamping tool is large, and a larger clamping force is required to ensure stable clamping, and the clamping tool is easily worn during the clamping process, so it will be assigned a higher clamping difficulty score; and for some softer materials, such as pure aluminum, indentations and deformations are prone to occur during clamping, and more sophisticated clamping force control is required, and its clamping difficulty score is also higher. For some materials with special physical properties, such as copper alloys with high thermal conductivity, the heat transfer is fast during welding, which may affect the temperature distribution of the clamping part and thus affect the stability of the clamping. These factors will also be considered in the scoring process. By analyzing the information of K materials one by one and comparing them with the scoring table, K welding material scores are obtained. These scores reflect the special requirements and difficulties of different materials in welding clamping.

[0028] Step S135, weighted fusion of the K thickness change scales, K size change scales, K welding joint scores and K welding material scores, and output the K clamping stabilities. Specifically, the weighted fusion method is used to comprehensively calculate the K thickness change scales, K size change scales, K welding joint scores and K welding material scores to output K clamping stabilities. In the weighted fusion process, the weight coefficients of each factor are first determined. The determination of the weight coefficient is based on a large amount of experimental data and engineering practice experience, reflecting the relative importance of each factor on the welding clamping stability. For example, for the welding of precision sheet metal parts with extremely high welding accuracy requirements, the weights of the thickness change scale and the size change scale may be relatively high, because the factors directly affect the fitting accuracy during welding; and for the welding of some large structural parts, the weights of the welding joint type and material properties are more prominent, because they have a key impact on the overall quality and structural strength of the welding. The clamping stability of each sheet metal part is obtained by multiplying each factor by its corresponding weight coefficient and then summing the calculation. The higher the clamping stability, the easier it is to achieve stable and reliable fixation of the sheet metal during the welding clamping process. Conversely, it means that the clamping is more difficult and more attention and optimization are needed in the welding process and clamping scheme design.

[0029] Step S200, after extracting the reference sheet metal and the first sheet metal based on the welding call sequence, the joint surface parameters of the reference sheet metal and the first sheet metal are collected respectively to obtain the reference joint surface parameter sequence and the first joint surface parameter sequence. Specifically, the welding call sequence is analyzed, and the reference sheet metal and the first sheet metal are accurately positioned. For example, in the welding of the body frame, a larger and regular sheet metal located at the center bottom is selected as the reference sheet metal, and the adjacent side panel that is first welded with it is determined as the first sheet metal according to the sequence; the joint surface of the reference sheet metal is cleaned and pre-processed, and appropriate measuring equipment is selected according to its shape and size, such as a flatness measuring instrument for a flat surface and a three-coordinate measuring instrument for a curved surface, etc., to comprehensively collect its geometric shape, surface roughness, flatness and other parameters and organize them into a reference joint surface parameter sequence; similarly, the joint surface of the first sheet metal is also cleaned, equipment is selected, and parameters are measured, including parameters related to the edge of the joint with the reference part, so as to form a first joint surface parameter sequence, which lays a solid data foundation for subsequent welding parameter determination and quality control.

[0030] Step S300, perform material change synchronization analysis on the reference joint surface parameter sequence and the first joint surface parameter sequence, and output a welding parameter sequence based on the analysis results. Specifically, perform data preprocessing on the reference joint surface parameter sequence and the first joint surface parameter sequence, check and correct the erroneous data, and then integrate the relevant data to make them correspond, then identify the change nodes of the two respectively and merge them into an associated change node sequence, decompose the two parameter sequences processed by data alignment with the sequence to obtain multiple groups of joint surface associated parameters containing shape feature identifiers, and then extract the material information of the reference and the first sheet metal part from the K material information, and use this as a constraint to query and call the matching sample joint surface parameters and sample welding parameters in the welding data network calling system, use it to train the welding parameter fitting model, and finally load the joint surface associated parameters into the model for welding control analysis, so as to output a welding parameter sequence containing key parameters such as welding current and voltage.

[0031] In a possible implementation, the reference joint surface parameter sequence and the first joint surface parameter sequence are subjected to a material change synchronization analysis, and a welding parameter sequence is output based on the analysis result. Step S300 further includes step S310, in which the reference joint surface parameter sequence and the first joint surface parameter sequence are subjected to a data change synchronization analysis to obtain multiple sets of joint surface associated parameters. Specifically, the reference joint surface parameter sequence is analyzed by using analysis methods such as data difference calculation and slope change monitoring. For example, for the flatness parameter of the joint surface, the difference between adjacent measurement points is calculated. When the difference exceeds a preset threshold, it is determined as a potential change node; at the same time, the slope change of parameters such as surface roughness is analyzed. If the slope changes suddenly, the position is marked. In this way, the entire reference joint surface parameter sequence is traversed to determine multiple reference change nodes. The same method is applied to the first joint surface parameter sequence to find the corresponding first change node. Afterwards, the reference change node and the first change node are merged and sorted according to their positions in the sequence to construct an associated change node sequence. According to the sequence of associated change nodes, the reference joint surface parameter sequence and the first joint surface parameter sequence after precise data alignment are segmented. In each segment, multiple sets of joint surface associated parameters containing rich information are extracted. The parameters include the geometric shape details of the joint surface, such as accurate length measurement value, width ratio, arc size, curvature radius value, etc., and are closely combined with the material properties of the reference sheet metal and the first sheet metal. Taking the thermal expansion coefficient of the material as an example, if the thermal expansion coefficient of the reference sheet metal material is large, it may cause large deformation during the welding process, so the coefficient is recorded in the associated parameters for consideration in the subsequent welding parameter adjustment; for example, the hardness property of the material will affect the difficulty of fusion during welding, which is also included in the associated parameters. And each set of joint surface associated parameters is given a unique joint surface shape feature identifier, such as a quickly identifiable identification code for a joint surface with a specific arc and width ratio.

[0032] Step S320, extracting the reference material information of the reference sheet metal part and the first material information of the first sheet metal part from the K material information. Specifically, the repository of K material information is connected, and the reference material information of the reference sheet metal part is extracted through accurate data retrieval and screening. The information covers various characteristics of the reference sheet metal material, and clarifies its element composition and proportion from the chemical composition, such as determining the content of aluminum, magnesium, silicon and other elements in aluminum alloy; in terms of mechanical properties, the yield strength, tensile strength, elongation and other values ​​are recorded in detail; in the field of physical properties, the thermal conductivity, specific heat capacity, density and other parameters are mastered. Similarly, the first material information of the first sheet metal part is accurately extracted from the K material information, and organized into a standardized format that can be used for subsequent data screening.

[0033] Step S330, using the reference material information and the first material information as data screening constraints, performing welding data networking calls, and obtaining multiple sets of sample joint surface parameters and multiple sample welding parameters. Specifically, using the extracted reference material information and the first material information as core data screening constraints, access a professional and large welding data networking call system. The system stores a large amount of welding data from different welding scenarios, experimental research, and engineering practice. In the system, the reference material information and the first material information are input into the intelligent screening module, and the system performs high-speed and accurate matching searches in its database based on the constraints. Finally, multiple sets of sample joint surface parameters and multiple sample welding parameters are obtained, and the material properties of the sample data are highly matched with the current reference and the first sheet metal part, which can provide a solid and reliable data foundation for the training of subsequent welding parameter fitting models.

[0034] Step S340, adopt the said multiple groups of sample joint surface parameters and multiple sample welding parameters to train and obtain the welding parameter fitting model. Specifically, select a suitable machine learning algorithm, such as a neural network algorithm based on a multilayer perceptron to construct a welding parameter fitting model. The obtained multiple groups of sample joint surface parameters are used as the input layer data of the model, and the corresponding multiple sample welding parameters are used as the output layer data. During the training process, the model continuously adjusts the neuron connection weights and thresholds in the neural network through the back propagation algorithm. For example, initially, the model randomly assigns weights and thresholds, and after inputting the sample joint surface parameters, the predicted welding parameters are obtained, and the error with the real sample welding parameters is calculated. Then, the weights and thresholds are reversely adjusted according to the error. After hundreds or thousands of iterative trainings, the error is gradually reduced, so that the model can accurately learn the complex mapping relationship between the sample joint surface parameters and the sample welding parameters, thereby improving the prediction accuracy and generalization ability of the model.

[0035] Step S350, the multiple sets of joint surface associated parameters are loaded into the welding parameter fitting model for welding control analysis to obtain the welding parameter sequence. Specifically, when the welding parameter fitting model is fully trained to achieve the expected performance, the multiple sets of joint surface associated parameters carefully extracted before are loaded into the model for welding control analysis. After the model receives the input joint surface associated parameters, it carries out deep calculation and reasoning based on the mapping relationship that has been trained maturely inside it. Taking into account the geometric shape characteristics of the joint surface and the parameters related to the material properties, the model predicts the most appropriate welding current to ensure that the material can be fully fused during the welding process and will not be overheated and deformed due to excessive current or poor fusion due to too small current; determine the appropriate welding voltage to ensure that the arc burns stably during the welding process and avoid welding defects caused by voltage fluctuations; calculate the appropriate welding speed to make the weld beautiful and uniform and the welding strength meet the design requirements; at the same time, determine the accurate welding angle to ensure that the welding heat source can act evenly on the joint surface to maximize the welding quality. After this series of complex and precise calculations and analyses, the precise welding parameter sequence for the welding of the current reference sheet metal and the first sheet metal is finally obtained, providing scientific and reliable parameter guidance for the actual welding operation, ensuring that the welding process is completed efficiently and with high quality.

[0036] In one possible implementation, the reference joint surface parameter sequence and the first joint surface parameter sequence are subjected to data change synchronization analysis to obtain multiple groups of joint surface associated parameters, and step S310 further includes step S311, performing change node identification on the reference joint surface parameter sequence to obtain multiple reference change nodes. Specifically, a data fluctuation detection algorithm is used to analyze the reference joint surface parameter sequence. For example, for the thickness parameter of the joint surface, if the difference between adjacent data points exceeds the set thickness change threshold, this is marked as a potential change node. At the same time, check whether parameters such as surface roughness and hardness have similar abnormal fluctuations or trend changes, and determine multiple reference change nodes based on the above factors. The nodes represent the locations where the reference joint surface parameters have changed significantly.

[0037] Step S312, identify the change nodes of the first joint surface parameter sequence to obtain multiple first change nodes. Specifically, the first joint surface parameter sequence is processed using a data fluctuation detection algorithm. When the width parameter of the joint surface has a large numerical jump, that is, it is significantly different from the surrounding data points, or the material density parameter presents a discontinuous change, the position is determined as the first change node, thereby obtaining multiple key position information reflecting the change of the first joint surface parameters.

[0038] Step S313, merge the multiple baseline change nodes and the multiple first change nodes to obtain an associated change node sequence. Specifically, the multiple baseline change nodes and the multiple first change nodes identified above are merged and sorted according to their order and relative positions in their respective sequences. For example, if the 10th data point in the baseline joint surface sequence and the 12th data point in the first joint surface sequence both show changes that are significantly different from the surrounding data, these two nodes are matched and included in the associated change node sequence, so as to construct a sequence that can reflect the correlation between the changes in the two joint surface parameters.

[0039] Step S314, using the associated change node sequence to decompose the reference joint surface parameter sequence and the first joint surface parameter sequence after data alignment processing, to obtain multiple groups of joint surface associated parameters, wherein the multiple groups of joint surface associated parameters have multiple joint surface shape feature identifiers. Specifically, based on the constructed associated change node sequence as the segmentation basis, the reference joint surface parameter sequence and the first joint surface parameter sequence after data alignment processing (ensuring that the data of the reference joint surface and the first joint surface under the same position index can accurately correspond) are segmented. Parameters are extracted from each segment, and the parameters not only include joint surface geometric shape features such as length and angle, but also based on the synchronization of data changes, on the basis of the consistency of parameters in a small section of the welding joint surface, the identification information that can represent the characteristics of this section of the joint surface is summarized, such as "plane-equal thickness section" and "curved surface-gradient section", etc., to form multiple groups of joint surface associated parameters with multiple joint surface shape feature identifiers, so that different types of joint surface conditions can be quickly identified and processed later.

[0040] In a possible implementation, the multiple sets of joint surface associated parameters are loaded into the welding parameter fitting model for welding control analysis to obtain the welding parameter sequence, and step S350 further includes step S351, performing network call of weld images according to the welding parameter sequence, multiple sets of joint surface associated parameters and multiple joint surface shape features to obtain a standard weld image sequence and multiple defective weld image sets. Specifically, according to the welding parameter sequence (including multiple sets of welding current, voltage, speed, wire feeding speed), multiple sets of joint surface associated parameters and multiple joint surface shape features, the weld image database is accessed for network call. By matching the parameters and feature information, the standard weld image sequence corresponding to the current welding task under the ideal welding condition is screened out from the database, and the image presents an ideal welding quality, covering the geometric shape, smoothness, surface features and other aspects of the weld; and at the same time, multiple defective weld image sets with different quality defects (such as pores, cracks, over-welding or under-welding, etc.) are obtained.

[0041] Step S352, label the weld quality defects of the plurality of defective weld image sets to obtain a plurality of defective labelled image sets. Specifically, for the plurality of defective weld image sets obtained, the weld quality defects in each image are analyzed. For images with pores, the position, size, quantity and other information of the pores are clearly marked; for images with cracks, the direction, length, width and other characteristics of the cracks are marked; for images with over-welding or under-welding, the abnormal thickness area of ​​the weld is marked, etc., so as to obtain a plurality of defective labelled image sets with detailed quality defect labels.

[0042] Step S353, taking the standard weld image sequence as a positive sample, taking the multiple defect weld image sets and the multiple defect identification image sets as negative samples, and optimizing the parameters of the weld quality detection model until the defect recognition accuracy of the weld quality detection model meets the preset performance requirements, wherein the weld quality detection model includes multiple weld quality detection units, and the multiple weld quality detection units use the welding parameter sequence as an activation condition. Specifically, the standard weld image sequence is taken as a positive sample, and the multiple defect weld image sets and the multiple defect identification image sets are taken as negative samples, and are input into the weld quality detection model. The model includes multiple weld quality detection units, and the units use the welding parameter sequence as an activation condition, that is, according to the current welding parameters, the corresponding model submodules are activated. Using sample data, through multiple iterative training, the parameters of the model are continuously adjusted, such as the weights and thresholds in the neural network, the error between the model prediction result and the real label of the sample is calculated, and the model is optimized by back propagation and other algorithms until the defect recognition accuracy of the weld quality detection model meets the preset performance requirements, and various quality defect types in the weld can be accurately identified and the defect positions can be located.

[0043] Step S400, in the process of using the welding parameter sequence to control the battery-type welder to weld the joint surface of the reference sheet metal part and the first sheet metal part, the pre-constructed weld quality detection model is run to synchronously monitor the weld quality online based on the weld image, and multiple weld defect nodes are output. Specifically, during the welding operation, the welding parameter sequence including multiple sets of welding current, voltage, speed and wire feeding speed is first input into the battery-type welder control system, and the welding of the joint surface of the reference sheet metal part and the first sheet metal part is started after the welding specification is set. At the same time, the pre-constructed weld quality detection model is started, and the weld image is collected at a reasonable frequency using a high-definition industrial camera installed near the welding machine and transmitted to the model image receiving module via a high-speed data transmission line. The model first grays the image, reduces noise, and enhances the image, then extracts the weld geometry, texture, and grayscale features and compares them with the pre-stored standard features. If there is a deviation, the type and degree are determined to determine the defect type, and the defect position, that is, the weld defect node, is determined according to the image coordinates. Finally, multiple weld defect nodes containing defect position and type information are sorted and output for use in welding operations or quality control links.

[0044] In a possible implementation, in the process of using the welding parameter sequence to control the battery-type welder to weld the reference sheet metal part and the first sheet metal part to the joint surface, the pre-built weld quality detection model is run to synchronously monitor the weld quality online based on the weld image, and multiple weld defect nodes are output. Step S400 further includes step S410, in the process of using the welding parameter sequence to control the battery-type welder to weld the reference sheet metal part and the first sheet metal part to the joint surface, when the welding parameters change, the welding parameters before the change are called to obtain the real-time welding parameters, and the image acquisition unit at the front end of the battery-type welder is activated to acquire the weld image to obtain the real-time weld image. Specifically, during the process of using the welding parameter sequence to control the battery-type welder to weld the reference sheet metal part and the first sheet metal part to the joint surface, once the welding parameters change, the welding machine control system quickly captures the change, and immediately calls the welding parameters before the change, so as to accurately obtain the real-time welding parameters (such as the current actual welding current, voltage, speed, wire feeding speed, etc.). At the same time, the image acquisition unit at the front end of the battery-powered welding machine is activated. After being triggered, the unit acquires an image of the weld being welded to obtain a real-time weld image that can reflect the actual condition of the weld at the moment.

[0045] Step S420, the real-time welding parameters and the real-time weld image are sent to the weld quality detection model. Specifically, the acquired real-time welding parameters and the real-time weld image are stably sent to the pre-built weld quality detection model through a data transmission link (such as a data line or a wireless network). The transmission process ensures the integrity and timeliness of the data, and provides an accurate data basis for subsequent weld defect identification.

[0046] Step S430, after the weld quality detection model locates the real-time detection unit in the multiple weld quality detection units according to the real-time welding parameters, the real-time weld image is input into the real-time detection unit to perform weld defect recognition and output the real-time defect node. Specifically, after receiving the real-time welding parameters, the weld quality detection model matches and locates the multiple weld quality detection units contained therein according to the parameters, and finds the corresponding real-time detection unit. Each detection unit is optimized for a specific range or type of welding parameters to improve the accuracy and efficiency of defect detection. After the positioning is completed, the real-time weld image is input into the selected real-time detection unit. The unit uses an image analysis algorithm to extract features of the weld area in the image, such as the shape, texture, color and other features of the weld, and compares and analyzes with the pre-stored normal weld image feature library. If the feature difference is found to exceed the set threshold, it is determined that there is a weld defect, and the type of defect (such as pores, cracks, over-welding, under-welding, etc.) and the specific location in the weld are further determined, that is, the real-time defect node is output.

[0047] Step S440, and so on, in the process of using the welding parameter sequence to control the battery-type welder to weld the joint surface of the reference sheet metal part and the first sheet metal part, the weld quality detection model is run to synchronously monitor the weld quality online, and the multiple weld defect nodes are output. Specifically, in an iterative loop, during the entire process of using the welding parameter sequence to control the battery-type welder to weld the joint surface of the reference sheet metal part and the first sheet metal part, the weld quality detection model is continuously run, and the above-mentioned parameter monitoring, data collection, transmission, analysis and defect identification processes are continuously performed, thereby realizing synchronous online monitoring of weld quality. As welding progresses, multiple weld defect nodes are successively output, and these node information completely records the defects that occur during the welding process, including defect locations, types, etc., providing a key basis for subsequent welding quality assessment, repair and process optimization.

[0048] Step S500, with the welding call sequence as a constraint, the welding control and weld defect identification of the remaining K-2 sheet metal parts are performed, and a weld defect node set is output. Specifically, the welding call sequence is analyzed, and the welding order and connection relationship of the remaining K-2 sheet metal parts are clarified to formulate a welding plan. Then, for each remaining sheet metal part, the joint surface parameter collection process is repeated, and the welding parameter sequence is determined through steps such as material change synchronization analysis. Then, during welding, the sequence is used to control the welding machine. When the welding parameters change, the real-time parameters and weld images are obtained and sent to the weld quality detection model. The model locates the corresponding detection unit and identifies the defects and outputs the real-time nodes. Finally, the defect nodes in the welding of each sheet metal part are summarized and sorted, and a weld defect node set containing information such as the defect location, type and sheet metal part number to which it belongs is generated, providing a basis for welding quality assessment, etc.

[0049] The embodiment of the present application adopts the method of determining the welding call sequence through welding sequence analysis, extracting the joint surface parameters of the reference sheet metal part and the first sheet metal part, performing material change synchronization analysis, and generating a welding parameter sequence. When using welding parameters to control the welding process of the battery-type welder, quality monitoring is performed based on the weld image and weld defect nodes are output. Finally, welding control and defect identification of the remaining sheet metal parts are performed according to the welding call sequence, and a weld defect node set is output, so as to achieve the technical effect of real-time monitoring of the weld quality in the sheet metal welding process, accurately locating the position and type of weld defects, so as to be able to timely discover and correct welding defects and ensure that the welding quality meets the standards.

[0050] In the above, refer to Figure 1 The following describes in detail a method for online monitoring of weld quality in a sheet metal welding process according to an embodiment of the present invention. Figure 2 An online monitoring platform for weld quality in a sheet metal welding process according to an embodiment of the present invention is described.

[0051] The weld quality online monitoring platform for sheet metal welding process according to the embodiment of the present invention is used to solve the technical problem that the prior art cannot monitor the weld defects in the sheet metal welding process in real time and accurately, and achieves the technical effect of real-time monitoring of the weld quality in the sheet metal welding process, accurately locating the position and type of weld defects, so as to timely discover and correct welding defects and ensure that the welding quality meets the standards. The weld quality online monitoring platform for sheet metal welding process includes: a welding call sequence acquisition module 10, a joint surface parameter acquisition module 20, a synchronization analysis module 30, a weld quality monitoring module 40, and a weld defect node set output module 50.

[0052] The welding call sequence acquisition module 10 is used to perform welding sequence analysis on the workpiece to be welded to obtain a welding call sequence, wherein the workpiece to be welded is obtained by welding K sheet metal parts;

[0053] The joint surface parameter acquisition module 20 is used to extract the reference sheet metal part and the first sheet metal part based on the welding call sequence, and then respectively acquire the joint surface parameters of the reference sheet metal part and the first sheet metal part to obtain the reference joint surface parameter sequence and the first joint surface parameter sequence;

[0054] The synchronization analysis module 30 is used to perform material change synchronization analysis on the reference joint surface parameter sequence and the first joint surface parameter sequence, and output a welding parameter sequence based on the analysis result;

[0055] The weld quality monitoring module 40 is used to run a pre-built weld quality detection model to synchronously perform online monitoring of weld quality based on a weld image and output a plurality of weld defect nodes during the process of welding the joint surface of the reference sheet metal part and the first sheet metal part by controlling the battery-type welder using the welding parameter sequence;

[0056] The weld defect node set output module 50 is used to perform welding control and weld defect identification of the remaining K-2 sheet metal parts with the welding call sequence as a constraint, and output a weld defect node set.

[0057] The specific configuration of the welding call sequence acquisition module 10 will be described in detail below. As described above, the welding sequence analysis is performed on the workpiece to be welded to obtain the welding call sequence, wherein the workpiece to be welded is obtained by welding K sheet metal parts, and the welding call sequence acquisition module 10 further includes: a sub-component model acquisition unit, the sub-component model acquisition unit is used to perform CAD modeling on the K sheet metal parts to obtain K sub-component models; a material information acquisition unit, the material information acquisition unit is used to interactively obtain K material information of the K sheet metal parts; a reliability evaluation unit, the reliability evaluation unit is used to perform sheet metal welding clamping reliability evaluation based on the K sub-component models and K material information, and output K clamping stabilities; a reference joint surface positioning unit, the reference joint surface positioning unit is used to sequence the K clamping stabilities and locate the reference joint surface based on the sorting extreme value; a welding sequence analysis unit, the welding sequence analysis unit is used to take the reference sheet metal part as the starting point, perform welding sequence analysis of adjacent sheet metal parts according to the K clamping stabilities, and generate the welding call sequence.

[0058] Among them, based on the K sub-component models and K material information, the reliability evaluation of sheet metal welding clamping is performed, and K clamping stabilities are output. The reliability evaluation unit further includes: a change scale acquisition sub-unit, the change scale acquisition sub-unit is used to perform thickness feature analysis based on the K sub-component models, and obtain K thickness change scales and K size change scales; a constraint splicing sub-unit, the constraint splicing sub-unit is used to collect and obtain K welding joint types of the K sub-component models after splicing the K sub-component models with the workpiece to be welded as a constraint; a welding difficulty assignment sub-unit The welding difficulty assigning subunit is used to use the K welding joint types to traverse the welding joint scoring table to assign welding difficulty points, so as to obtain K welding joint scores; the clamping difficulty assigning subunit is used to use the K material information to traverse the welding material scoring table to assign clamping difficulty points, so as to obtain K welding material scores; the clamping stability output subunit is used to weightedly fuse the K thickness variation scales, the K size variation scales, the K welding joint scores and the K welding material scores, so as to output the K clamping stabilities.

[0059] Next, the specific configuration of the synchronization analysis module 30 will be described in detail. As described above, a material change synchronization analysis is performed on the reference joint surface parameter sequence and the first joint surface parameter sequence, and a welding parameter sequence is output based on the analysis results. The synchronization analysis module 30 further includes: a joint surface associated parameter acquisition unit, the joint surface associated parameter acquisition unit is used to perform data change synchronization analysis on the reference joint surface parameter sequence and the first joint surface parameter sequence to obtain multiple groups of joint surface associated parameters; a material information extraction unit, the material information extraction unit is used to extract the reference material information of the reference sheet metal and the first material information of the first sheet metal from the K material information; a data networking calling unit, the data networking calling unit is used to use the reference material information and the first material information as data screening constraints to perform welding data networking calling to obtain multiple groups of sample joint surface parameters and multiple sample welding parameters; a fitting model training unit, the fitting model training unit is used to use the multiple groups of sample joint surface parameters and multiple sample welding parameters to train and obtain a welding parameter fitting model; a welding control analysis unit, the welding control analysis unit is used to load the multiple groups of joint surface associated parameters into the welding parameter fitting model to perform welding control analysis and obtain the welding parameter sequence.

[0060] Wherein, the data change synchronization analysis is performed on the reference joint surface parameter sequence and the first joint surface parameter sequence to obtain multiple groups of joint surface associated parameters, and the joint surface associated parameter acquisition unit further includes: a change node identification subunit, the change node identification subunit is used to identify the change nodes of the reference joint surface parameter sequence to obtain multiple reference change nodes; a first change node acquisition subunit, the first change node acquisition subunit is used to identify the change nodes of the first joint surface parameter sequence to obtain multiple first change nodes; an associated change node sequence acquisition subunit, the associated change node sequence acquisition subunit is used to merge the multiple reference change nodes and the multiple first change nodes to obtain an associated change node sequence; a multiple group of joint surface associated parameter acquisition subunit, the multiple group of joint surface associated parameter acquisition subunit is used to use the associated change node sequence to decompose the reference joint surface parameter sequence and the first joint surface parameter sequence after data alignment processing to obtain multiple groups of joint surface associated parameters, wherein the multiple groups of joint surface associated parameters have multiple joint surface shape feature identifiers.

[0061] Wherein, the multiple groups of joint surface associated parameters are loaded into the welding parameter fitting model for welding control analysis to obtain the welding parameter sequence, and the welding control analysis unit further includes: an image networking calling subunit, the image networking calling subunit is used for welding seam image networking calling according to the welding parameter sequence, multiple groups of joint surface associated parameters and multiple joint surface shape features to obtain a standard weld image sequence and multiple defective weld image sets; a label identification subunit, the label identification subunit is used to identify the weld quality defect labels of the multiple defective weld image sets to obtain multiple defect identification image sets; a parameter optimization subunit, the parameter optimization subunit is used to use the standard weld image sequence as a positive sample, the multiple defective weld image sets and the multiple defect identification image sets as negative samples, and perform parameter optimization of the weld quality detection model until the defect recognition accuracy of the weld quality detection model meets the preset performance requirements, wherein the weld quality detection model includes multiple weld quality detection units, and the multiple weld quality detection units use the welding parameter sequence as an activation condition.

[0062] The specific configuration of the weld quality monitoring module 40 will be described in detail below. As described above, in the process of using the welding parameter sequence to control the battery-type welder to perform joint surface welding on the reference sheet metal part and the first sheet metal part, the pre-built weld quality detection model is run to synchronously perform online monitoring of the weld quality based on the weld image, and multiple weld defect nodes are output. The weld quality monitoring module 40 further includes: a real-time welding parameter acquisition unit, which is used to use the welding parameter sequence to control the battery-type welder to perform joint surface welding on the reference sheet metal part and the first sheet metal part. When the welding parameters change, the welding parameters before the change are called to obtain the real-time welding parameters, and the image acquisition unit at the front end of the battery-type welder is activated to perform weld image acquisition to obtain the real-time weld image; a data sending unit, which The data sending unit is used to send the real-time welding parameters and the real-time weld image to the weld quality detection model; the real-time defect node output unit, the real-time defect node output unit is used for the weld quality detection model to locate the real-time detection unit in the multiple weld quality detection units according to the real-time welding parameters, input the real-time weld image into the real-time detection unit to perform weld defect identification, and output the real-time defect node; the weld quality online monitoring unit, the weld quality online monitoring unit is used to, by analogy, run the weld quality detection model to synchronously perform weld quality online monitoring and output the multiple weld defect nodes during the process of using the welding parameter sequence to control the battery-type welding machine to perform joint surface welding on the reference sheet metal part and the first sheet metal part.

[0063] The online weld quality monitoring platform for sheet metal welding process provided by the embodiment of the present invention can execute the online weld quality monitoring method for sheet metal welding process provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0064] Although the present application makes various references to certain modules in the platform according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be implemented; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0065] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A method for online monitoring of weld quality in sheet metal welding process, characterized in that: The method comprises: Perform welding sequence analysis on a workpiece to be welded to obtain a welding call sequence, wherein the workpiece to be welded is obtained by welding K sheet metal parts; After extracting the reference sheet metal part and the first sheet metal part based on the welding call sequence, collecting joint surface parameters of the reference sheet metal part and the first sheet metal part respectively to obtain a reference joint surface parameter sequence and a first joint surface parameter sequence; Performing material change synchronization analysis on the reference joint surface parameter sequence and the first joint surface parameter sequence, and outputting a welding parameter sequence based on the analysis result; In the process of welding the joint surface of the reference sheet metal part and the first sheet metal part by controlling the battery-type welder using the welding parameter sequence, the pre-built weld quality detection model is run to synchronously perform online monitoring of the weld quality based on the weld image, and a plurality of weld defect nodes are output; Taking the welding call sequence as a constraint, welding control and weld defect identification are performed on the remaining K-2 sheet metal parts, and a weld defect node set is output; Performing welding sequence analysis on the workpiece to be welded to obtain a welding call sequence, the method comprising: Performing CAD modeling on the K sheet metal parts to obtain K sub-component models; Interactively obtaining K material information of the K sheet metal parts; Based on the K sub-component models and K material information, reliability evaluation of sheet metal welding clamping is performed, and K clamping stabilities are output; Sequencing the K clamping stabilities and locating the reference joint surface based on the sorted extreme values; Taking the reference sheet metal part as a starting point, the welding sequence of adjacent sheet metal parts is analyzed according to the K clamping stabilities to generate the welding call sequence.

2. The method for online monitoring of weld quality in sheet metal welding process according to claim 1, characterized in that: Performing material change synchronization analysis on the reference joint surface parameter sequence and the first joint surface parameter sequence, and outputting a welding parameter sequence based on the analysis result, the method comprising: Performing data change synchronization analysis on the reference joint surface parameter sequence and the first joint surface parameter sequence to obtain multiple groups of joint surface associated parameters; Extracting reference material information of the reference sheet metal part and first material information of the first sheet metal part from the K material information; Using the reference material information and the first material information as data screening constraints, performing a welding data networking call to obtain multiple groups of sample joint surface parameters and multiple sample welding parameters; The welding parameter fitting model is obtained by training the plurality of groups of sample joint surface parameters and the plurality of sample welding parameters; The multiple groups of joint surface associated parameters are loaded into the welding parameter fitting model to perform welding control analysis to obtain the welding parameter sequence.

3. The method for online monitoring of weld quality in sheet metal welding process according to claim 2, characterized in that: Performing data change synchronization analysis on the reference joint surface parameter sequence and the first joint surface parameter sequence to obtain multiple groups of joint surface associated parameters, the method comprising: Performing change node identification on the reference joint surface parameter sequence to obtain a plurality of reference change nodes; Performing change node identification on the first joint surface parameter sequence to obtain a plurality of first change nodes; Merging the plurality of baseline change nodes and the plurality of first change nodes to obtain an associated change node sequence; The reference joint surface parameter sequence and the first joint surface parameter sequence after data alignment are decomposed by using the associated change node sequence to obtain multiple groups of joint surface associated parameters, wherein the multiple groups of joint surface associated parameters have multiple joint surface shape feature identifiers.

4. The method for online monitoring of weld quality in sheet metal welding process according to claim 3, characterized in that: The method comprises: Performing a network call of weld images according to the welding parameter sequence, multiple sets of joint surface associated parameters and multiple joint surface shape features to obtain a standard weld image sequence and multiple defective weld image sets; Marking the multiple defective weld image sets with weld quality defect labels to obtain multiple defect marked image sets; The standard weld image sequence is used as a positive sample, and the multiple defective weld image sets and the multiple defect identification image sets are used as negative samples to perform parameter optimization of the weld quality detection model until the defect recognition accuracy of the weld quality detection model meets the preset performance requirements, wherein the weld quality detection model includes multiple weld quality detection units, and the multiple weld quality detection units use the welding parameter sequence as an activation condition.

5. The method for online monitoring of weld quality in sheet metal welding process according to claim 4, characterized in that: In the process of using the welding parameter sequence to control the battery-type welder to weld the joint surface of the reference sheet metal part and the first sheet metal part, running the pre-built weld quality detection model to synchronously perform online monitoring of the weld quality based on the weld image, and outputting multiple weld defect nodes, the method includes: In the process of welding the joint surface of the reference sheet metal part and the first sheet metal part by using the welding parameter sequence to control the battery-type welder, when the welding parameters change, the welding parameters before the change are called to obtain the real-time welding parameters, and the weld image is acquired by activating the image acquisition unit at the front end of the battery-type welder to acquire the real-time weld image; Sending the real-time welding parameters and the real-time weld image to the weld quality detection model; The weld quality detection model inputs the real-time weld image into the real-time detection unit to perform weld defect recognition and outputs a real-time defect node after locating the real-time detection unit in the multiple weld quality detection units according to the real-time welding parameters; By analogy, in the process of using the welding parameter sequence to control the battery-type welder to perform joint surface welding on the reference sheet metal part and the first sheet metal part, the weld quality detection model is run to synchronously perform online monitoring of the weld quality and output the multiple weld defect nodes.

6. The method for online monitoring of weld quality in sheet metal welding process according to claim 1, characterized in that: Based on the K sub-component models and K material information, reliability evaluation of sheet metal welding clamping is performed, and K clamping stabilities are output. The method includes: Performing thickness feature analysis based on the K subcomponent models to obtain K thickness variation scales and K size variation scales; After splicing the K sub-component models with the workpiece to be welded as a constraint, K welding joint types of the K sub-component models are collected and obtained; The K welding joint types are used to traverse the welding joint scoring table to assign welding difficulty scores, and K welding joint scores are obtained; Using the K material information to traverse the welding material scoring table to assign points to the clamping difficulty, and obtain K welding material scores; The K thickness variation scales, the K size variation scales, the K welding joint scores and the K welding material scores are weightedly fused to output the K clamping stabilities.

7. The online monitoring platform for weld quality in sheet metal welding process is characterized by: The platform is used to implement the online monitoring method for weld quality in a sheet metal welding process according to any one of claims 1 to 6, and the platform comprises: A welding call sequence acquisition module, wherein the welding call sequence acquisition module is used to perform welding sequence analysis on a workpiece to be welded to obtain a welding call sequence, wherein the workpiece to be welded is obtained by welding K sheet metal parts; A joint surface parameter acquisition module, wherein the joint surface parameter acquisition module is used to extract the reference sheet metal part and the first sheet metal part based on the welding call sequence, and then respectively acquire joint surface parameters of the reference sheet metal part and the first sheet metal part to obtain a reference joint surface parameter sequence and a first joint surface parameter sequence; A synchronization analysis module, the synchronization analysis module is used to perform a material change synchronization analysis on the reference joint surface parameter sequence and the first joint surface parameter sequence, and output a welding parameter sequence based on the analysis result; A weld quality monitoring module, wherein the weld quality monitoring module is used to run a pre-built weld quality detection model to synchronously perform online monitoring of weld quality based on a weld image and output a plurality of weld defect nodes during a process of performing joint surface welding on the reference sheet metal part and the first sheet metal part by controlling the battery-type welder using the welding parameter sequence; A weld defect node set output module is used to perform welding control and weld defect identification of the remaining K-2 sheet metal parts with the welding call sequence as a constraint, and output a weld defect node set.

Images