Mattress frame forming precision control method and system

Abstract

The invention provides a mattress frame forming precision control method and system, relates to the field of automatic production of mattress frames, and aims to predict macroscopic deformation tendency based on optical characteristics of heat affected zone images of test weldments and adjust welding parameters according to the macroscopic deformation tendency so as to actively control and compensate deformation before production or in an early stage. The method has the advantages that the macroscopic deformation tendency can be predicted based on the heat affected zone characteristics in the welding process, and welding parameters are adjusted accordingly, so that deformation is actively controlled and compensated before production or in the early stage, the forming precision is improved, the rejection rate is reduced, and the defect that the prior art depends on lagging detection and experience adjustment is overcome.

Description

Technical Field

[0001] The present application relates to the field of automated production of mattress frames, and more specifically, to a method and system for controlling the molding precision of mattress frames. Background Art

[0002] The mattress frame is a critical step in the automated mattress production line. When production tasks change, such as changing the frame size or using wire of a different grade, diameter, or surface treatment, the production system adjusts the corresponding parameters. After bending and cutting the wire, a welding robot automatically welds the wire to form the frame structure.

[0003] Improper heat input may not only cause defects in solder joints, but more importantly, it will generate uneven thermal stresses in and around the welding area. These thermal stresses are released during cooling after welding and are converted into shrinkage, bending or torsional deformation of the frame components. Since the mattress frame is a structure formed by multiple sections of rods connected by multiple welds, the slight deformation generated by a single weld will be transmitted and accumulated through the constraints between the rods, eventually leading to macroscopic geometric accuracy deviations of the entire frame, such as flatness warping, diagonal length deviation, corner angle deviation, etc. This problem is particularly prominent when dealing with complex and special-shaped frames or frames with non-standard joints, because their geometric configuration leads to different heat dissipation and constraint conditions for each joint. If the welding parameters cannot be adjusted in a targeted manner, the superposition of uneven local deformations will cause the final formed frame to deviate seriously from the design requirements.

[0004] Therefore, under the task of frequent specification and material switching, how to overcome the welding heat input mismatch caused by traditional general database matching or experience adjustment, effectively suppress the macro welding deformation caused by improper heat input, and establish a method that can quickly and accurately predict and actively compensate welding parameters before mass production to ensure the molding accuracy and production consistency of the mattress frame, is a technical problem that needs to be solved urgently.

[0005] In view of the above problems, the existing technology is in urgent need of improvement. Summary of the Invention

[0006] The purpose of this application is to provide a mattress frame molding precision control method and system, which has the advantages of being able to predict the macro deformation tendency based on the heat-affected zone characteristics during the welding process and adjust the welding parameters accordingly, thereby actively controlling and compensating for deformation before production or in the early stages, improving molding precision, and reducing scrap rate, overcoming the shortcomings of the existing technology that relies on lagging detection and experience-based adjustment.

[0007] This application provides a mattress frame molding precision control method, comprising: Before formal mass production, a test weld is performed using a preset set of initial welding parameters, an image of the heat-affected zone located in a preset observation area of the test weld is collected, and optical characteristic parameters of the heat-affected zone image are extracted; the test weld is a mattress frame sample that is tested for test welding before mass production; Determining the macro deformation tendency of the mattress frame under the initial welding parameters according to a pre-established mapping relationship between the optical characteristic parameters and the macro deformation tendency of the mattress frame; According to the macro deformation tendency of the mattress frame under the initial welding parameters and the welding parameter compensation rules, the initial welding parameters are adjusted to obtain welding accuracy optimization parameters; Welding precision optimization parameters are used for batch welding production of mattress frames.

[0008] Furthermore, the present application also proposes a mattress frame molding precision control method, which determines the macro deformation tendency of the mattress frame under initial welding parameters based on a pre-established mapping relationship between optical characteristic parameters and the macro deformation tendency of the mattress frame, as well as the optical characteristic parameters, including: Obtaining the joint types of the mattress frame and the correspondence between the pre-established joint types and a set of specific optical characteristic parameters used to predict macro deformation tendencies; the joint types include corner joints, butt joints, and T-joints; Determining a specific optical characteristic parameter set from the corresponding relationship according to the connector type, and selecting parameters corresponding to the specific optical characteristic parameter set from the extracted optical characteristic parameters as optical characteristic parameters used in the current prediction; Based on the optical characteristic parameters currently used in the prediction and the pre-established mapping relationship between the optical characteristic parameters and the macro deformation tendency of the mattress frame, the macro deformation tendency of the mattress frame under the initial welding parameters is determined.

[0009] Furthermore, the present application also proposes, according to a mattress frame molding precision control method, the macro deformation tendency includes at least multiple deformation indicators including length, width, angle and / or flatness; the compensation rule includes the influence relationship between welding parameter adjustment and multiple deformation indicators; According to the macro deformation tendency and welding parameter compensation rules, adjust the initial welding parameters to obtain the welding accuracy optimization parameters, including: obtaining a plurality of target improvement values corresponding to the plurality of deformation indicators; Determining a set of first welding parameter adjustment amounts that can make the overall deformation of the mattress frame approach the multiple target improvement values based on the multiple deformation indices, the multiple target improvement values, and the influence relationship between the welding parameter adjustment and the multiple deformation indices; The welding parameter adjustment amount is applied to the initial welding parameters to determine the welding accuracy optimization parameters.

[0010] Through the above scheme, the macro deformation can be refined into multiple quantifiable indicators, and the parameters can be adjusted based on the influence relationship between these indicators and welding parameters, making the parameter adjustment more targeted and effective.

[0011] Furthermore, the present application also proposes, based on the above-mentioned mattress frame molding precision control method, determining a set of first welding parameter adjustment amounts that can make the overall deformation of the mattress frame approach the multiple target improvement values based on multiple deformation indicators, multiple target improvement values, and the influence relationship between welding parameter adjustment on the multiple deformation indicators, including: Obtaining a plurality of actual deformation indicators measured after a preset number of mattress frame welding productions are performed using the first welding parameter adjustment amount; Determining whether the deviations between the multiple actual deformation indicators and the multiple target improvement values meet the preset correction trigger conditions; If the correction trigger condition is met, the influence relationship on multiple deformation indicators is adjusted according to the deviation and welding parameters, and the influence relationship is corrected to obtain a corrected influence relationship; the influence relationship is used to characterize the correlation between the adjustment amount of each welding parameter and the change amount of each deformation indicator; Based on the predicted multiple deformation indicators, multiple target improvement values, and the corrected influence relationship in the macro deformation tendency of the mattress frame, a set of first welding parameter adjustment amounts that can make the overall deformation of the mattress frame approach the multiple target improvement values is determined.

[0012] Furthermore, the present application also proposes that, according to the above-mentioned mattress frame molding precision control method, if the correction trigger condition is met, the influence relationship of multiple deformation indicators is adjusted according to the deviation and welding parameters, and the influence relationship is corrected to obtain the corrected influence relationship, including: According to the deviation of each deformation index in the deviation, combined with the influence relationship, analyze and determine the welding parameters that affect the deviation of the deformation index; Based on the analyzed and determined comprehensive influence of each welding parameter on the combination of multiple deformation index deviations in the obtained deviation, one or more welding parameter-deformation index associations having a preset contribution level to the formation of the deviation are identified from the obtained influence relationships; The strength of the association of one or more identified welding parameter-deformation index associations in the obtained influence relationship is adjusted to obtain a revised influence relationship.

[0013] Furthermore, the present application also proposes that, based on the aforementioned mattress frame molding precision control method, based on the analyzed and determined degree of comprehensive influence of each welding parameter on the combination of multiple deformation index deviations in the obtained deviation, one or more sets of welding parameter-deformation index associations having a preset contribution level to the formation of the deviation are identified from the obtained influence relationship, including: For each welding parameter included in the influence relationship, the expected influence value of the adjustment of the welding parameter on the deviation of each deformation index in the obtained deviation is calculated; Determining, based on the expected influence value of each welding parameter and the actual deviation of each deformation index in the obtained deviation, a quantitative contribution of each welding parameter to the deviation combination of multiple deformation indicators in the currently obtained deviation; Compare the quantified contribution degree with the preset contribution level judgment condition, and select the welding parameters that meet the contribution level judgment condition; The association formed by the screened welding parameters and the corresponding deformation indicators when calculating the expected influence value is used as one or more sets of welding parameter-deformation indicator associations identified from the obtained influence relationship and having a preset contribution level to the formation of the currently obtained deviation.

[0014] Furthermore, the present application also proposes that, according to a mattress frame molding precision control method, the step of extracting preset optical characteristic parameters from the heat-affected zone image includes: Acquire heat-affected zone images; analyze the heat-affected zone images and identify interference areas in the images; Based on the recognition result of the interference area, the heat-affected zone image is processed to obtain a processed image; Extract preset optical feature parameters from the processed image.

[0015] Furthermore, the present application also proposes, according to the above-mentioned mattress frame molding precision control method, determining the quantitative contribution of each welding parameter to the combination of multiple deformation index deviations in the currently obtained deviation based on the expected impact value of each welding parameter and the actual deviation of each deformation index in the currently obtained deviation, including: Obtain the expected impact value of each welding parameter on the deviation of each deformation index, and obtain the actual deviation of each deformation index in the deviation; Performing preset numerical processing on the expected influence value of each welding parameter obtained and the actual deviation amount of each deformation index obtained to obtain a numerical processing result; Obtain the influencing factors of each deformation index on the final molding quality of the mattress frame; Based on the influencing factors of each deformation index obtained, a comprehensive calculation is performed in combination with the numerical processing results to determine the quantitative contribution of each welding parameter to the deviation combination of multiple deformation indexes in the currently obtained deviation.

[0016] Furthermore, the present application also proposes that, based on the above-mentioned mattress frame molding precision control method, a preset numerical processing is performed on the expected influence values of each welding parameter obtained and the actual deviations of each deformation index obtained, to obtain numerical processing results, including: Obtain the expected impact value of each welding parameter and the actual deviation of each deformation index in the deviation; Perform function mapping or normalization processing on the expected impact value and the actual deviation according to preset rules; The numerical processing result is determined based on the expected impact value and actual deviation after function mapping or normalization.

[0017] Furthermore, the present application also proposes a mattress frame forming precision control system for adjusting mattress frame welding parameters, the system comprising: An acquisition module is used to weld a test piece using a preset set of initial welding parameters before formal mass production, capture an image of the heat-affected zone located in a preset observation area of the test piece, and extract optical characteristic parameters of the heat-affected zone image; the test piece is a mattress frame sample undergoing a test welding test before mass production; a determination module for determining the macro deformation tendency of the mattress frame under the initial welding parameters based on a pre-established mapping relationship between the optical characteristic parameters and the macro deformation tendency of the mattress frame; A parameter adjustment module, configured to adjust the initial welding parameters according to the macro deformation tendency of the mattress frame under the initial welding parameters and the welding parameter compensation rules to obtain welding accuracy optimization parameters; Production of welding modules, using welding precision optimization parameters for batch welding production of mattress frames.

[0018] From the above, it can be seen that the mattress frame molding precision control method and system provided by the present application predicts the macro deformation tendency based on the optical characteristics of the heat-affected zone image of the test weld, and adjusts the welding parameters accordingly, thereby actively controlling and compensating for deformation before production or in the early stage. It has the advantage of being able to predict the macro deformation tendency based on the heat-affected zone characteristics during the welding process, and adjust the welding parameters accordingly, thereby actively controlling and compensating for deformation before production or in the early stage, improving molding precision, and reducing scrap rate, overcoming the shortcomings of the existing technology that relies on lagging detection and experience adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 It is a flow chart of a mattress frame molding precision control method disclosed in an embodiment of the present invention.

[0021] Figure 2It is a structural schematic diagram of a mattress frame forming precision control system disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be embodied in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.

[0023] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically separate entities. That is, these functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0024] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, while others may be combined or partially combined. Therefore, the actual execution order may vary depending on the actual situation.

[0025] It should be noted that the “plurality” mentioned in this article refers to two or more.

[0026] This application proposes a mattress frame molding precision control method, such as Figure 1 Shown, including: S101, before formal mass production, welding a test piece using a preset set of initial welding parameters, collecting an image of the heat-affected zone located in a preset observation area of the test piece, and extracting optical characteristic parameters of the heat-affected zone image; the test piece is a mattress frame sample undergoing a test welding test before mass production; S102, determining the macro deformation tendency of the mattress frame under the initial welding parameters based on the pre-established mapping relationship between the optical characteristic parameters and the macro deformation tendency of the mattress frame, and the optical characteristic parameters; S103, adjusting the initial welding parameters according to the macro deformation tendency of the mattress frame under the initial welding parameters and the welding parameter compensation rule to obtain welding accuracy optimization parameters; S104, using welding precision optimization parameters to carry out batch welding production of mattress frames.

[0027] Among them, in S101, the preset observation area of the test weld refers to a specific position or range on the test weld that is predetermined during the test welding process in order to collect the heat-affected zone image. It can be achieved by visual positioning, mechanical positioning or mark point positioning, such as the area near the welding joint and the specific width range on both sides of the weld centerline. It is mainly used to obtain representative welding heat-affected zone information.

[0028] The heat-affected zone image refers to the visual information obtained by image acquisition equipment, which reflects the area formed by the welding heat on the weld joint and its surrounding base material. It can be collected using a visible light camera, infrared camera or high-speed camera. It is mainly used to capture the thermal physical phenomena and material response during the welding process.

[0029] Optical characteristic parameters refer to numerical values or descriptors extracted from the heat-affected zone image that can quantify the visual characteristics of the heat-affected zone. They can be extracted using image processing algorithms, such as the width, area, shape characteristics, color distribution, texture characteristics or grayscale gradient of the heat-affected zone. They are mainly used to convert image information into data that can be used for analysis and prediction.

[0030] The mapping relationship between the pre-established optical characteristic parameters and the macroscopic deformation tendency of the mattress frame refers to a model or rule set established through historical data, experiments or simulations, which associates the optical characteristic parameters of the heat-affected zone with the geometric deformation of the overall structure of the mattress frame after welding. It can be implemented using a regression model, a neural network model or a rule lookup table, and is mainly used to predict the overall structural deformation based on local welding characteristics.

[0031] Macro deformation tendency refers to the trend, direction and degree of geometric deformation such as shrinkage, bending and torsion of the overall structure of the mattress frame during the post-weld cooling process under specific welding parameters and material conditions. It is mainly used to evaluate the impact of current welding parameters on the geometric accuracy of the final product.

[0032] Welding parameter compensation rules refer to strategies, algorithms or rule sets that determine how to adjust the initial welding parameters to reduce or eliminate macro deformation based on the predicted macro deformation tendency. They can be implemented using model-based control, fuzzy control or expert systems, and are mainly used to guide the optimization adjustment of welding parameters.

[0033] Welding accuracy optimization parameters refer to a set of welding parameters used for subsequent mass production, which are obtained by adjusting the initial welding parameters based on the macro deformation tendency and welding parameter compensation rules. They are mainly used to improve the molding accuracy of the mattress frame.

[0034] The core innovation of this application lies in that by combining the optical characteristic parameters extracted from the collected heat-affected zone image with the pre-established mapping relationship between the optical characteristic parameters and the macro-deformation tendency of the mattress frame, the macro-deformation tendency of the mattress frame under the initial welding parameters is determined, and the welding parameters are adjusted based on the predicted results, thereby achieving the effect of actively optimizing the welding parameters and effectively suppressing welding deformation before mass production.

[0035] The solution of this application achieves precision control of mattress frame molding through a strategy that combines feedforward control with feedback correction. First, before formal mass production, a test weld is performed using a set of initial welding parameters. During or after the test weld, images of the heat-affected zone (HAZ) of a pre-set observation area of the test weld are captured. The captured HAZ images are analyzed and processed to extract optical characteristic parameters that reflect the welding heat input and material response. These optical characteristic parameters contain key information about the local welding process. Based on a pre-established mapping relationship that correlates the HAZ optical characteristic parameters with the overall macroscopic deformation of the mattress frame, combined with the optical characteristic parameters extracted from the test weld, the potential macroscopic deformation tendency of the entire mattress frame structure, such as deviations in length, angle, or flatness, is predicted under the current initial welding parameters. The predicted macroscopic deformation tendency quantifies the potential accuracy issues introduced by the initial parameters. Then, based on the predicted macroscopic deformation tendency and pre-set welding parameter compensation rules, the required adjustments to the initial welding parameters are calculated. The welding parameter compensation rules define how parameters such as welding current, voltage, and speed should be adjusted to offset or reduce deformation under different deformation tendencies. The calculated adjustments are applied to the initial welding parameters to obtain a set of parameters optimized for welding accuracy. Finally, these optimized parameters are applied to subsequent batch welding production of mattress frames. This entire process forms a closed loop, analyzing local welding characteristics to predict overall deformation and proactively adjusting welding parameters accordingly. This prevents potential deformation before production begins, improving forming accuracy.

[0036] In some embodiments, specifically, an industrial camera can be used to capture images of the weld joint area of the test weld to obtain an image of the heat-affected zone. The image processing unit can use a deep learning-based image segmentation algorithm to identify the boundaries of the heat-affected zone and calculate the area, perimeter, maximum width, average grayscale value, and other optical characteristic parameters of the heat-affected zone. The pre-established mapping relationship can be a neural network model trained using a large amount of test welding data from different materials and parameters. The input of this model is the optical characteristic parameters, and the output is the predicted values of multiple macroscopic deformation indicators of the mattress frame. The macroscopic deformation tendency can be specifically expressed as the predicted values of the angular deviation, diagonal length difference, and overall flatness deviation of the four corners of the mattress frame. The welding parameter compensation rules can be a rule base constructed based on expert experience or a parameter adjustment matrix obtained by solving an optimization algorithm. This rule base or matrix determines the adjustment amounts to be made to the initial welding parameters, such as welding current, welding voltage, and welding speed, based on the predicted deformation indicators. The calculated adjustment amounts are added to the initial parameters to obtain the welding accuracy optimization parameters for mass production.

[0037] The above technical solution predicts the macroscopic deformation tendency of the mattress frame based on images of the heat-affected zone of the test weld, and proactively adjusts the welding parameters based on the predictions. This allows for optimization of welding parameters before mass production begins, effectively avoiding or significantly reducing macroscopic deformation caused by improper welding heat input, and improving the molding accuracy and product consistency of the mattress frame. Compared to traditional end-of-line detection feedback methods, this solution can identify and resolve potential accuracy issues earlier, shorten production commissioning time, and reduce scrap rates. It is particularly suitable for automated production environments where product specifications and materials frequently change.

[0038] In some of the aforementioned embodiments of the present application, the method may specifically capture images of the heat-affected zone during the welding process during the trial welding phase, analyze the optical features in the images, such as temperature distribution, heat-affected zone width, and color change, and input these features into a pre-trained model (e.g., a regression model or a neural network model based on historical data). The model can predict the overall deformation of the mattress frame that may occur after welding, such as deviations in length, width, angle, or flatness, based on the input optical features. Based on the predicted deformation and preset compensation rules, the model calculates the required adjustments to the initial welding parameters (e.g., welding current, voltage, and speed) to obtain optimized welding parameters. These optimized parameters are then applied to subsequent batch production. In this way, macroscopic deformation caused by welding can be predicted and compensated based on real-time or near-real-time information during the welding process, thereby improving the molding accuracy of the mattress frame. However, in its implementation, mattress frames have a variety of joint types, and different joint types have different sensitivities to welding parameters. Using a unified set of optical feature parameters for prediction may result in reduced prediction accuracy and an inability to accurately reflect the macroscopic deformation tendencies of different joint types.

[0039] In this regard, in step S202 of the present application, the step of determining the macro deformation tendency of the mattress frame under the initial welding parameters includes: Obtaining the joint types of the mattress frame and the correspondence between the pre-established joint types and a set of specific optical characteristic parameters used to predict macro deformation tendencies; the joint types include corner joints, butt joints, and T-joints; Determining a specific optical characteristic parameter set from the corresponding relationship according to the connector type, and selecting parameters corresponding to the specific optical characteristic parameter set from the extracted optical characteristic parameters as optical characteristic parameters used in the current prediction; Based on the optical characteristic parameters currently used in the prediction and the pre-established mapping relationship between the optical characteristic parameters and the macro deformation tendency of the mattress frame, the macro deformation tendency of the mattress frame under the initial welding parameters is determined.

[0040] Among them, the joint type refers to the geometric configuration of the mattress frame at the connection point, which is specifically the connection form between metal wires through welding, such as a butt joint formed by the butt joint of two wire ends, a T-joint formed by the vertical intersection of two wires, and a corner joint formed by the connection of two wires at a corner. Its purpose is to distinguish the effects of different connection methods on welding heat input and stress distribution; the specific optical feature parameter set refers to a subset of all extracted optical feature parameters, which is selected according to the joint type and has a high correlation and influence on the prediction of the macro deformation tendency of this type of joint. Specifically, it can be determined by prior experimental analysis or data modeling which optical features (such as heat affected zone width, temperature) under different joint types are The optical characteristic parameters (such as heat-affected zone width and maximum temperature) are most closely correlated with the final macro deformation. The purpose is to focus on the most critical welding process characteristics for different joint types and improve the pertinence of the prediction. The correspondence relationship refers to the pre-established association rules or lookup table between the joint type and the specific optical characteristic parameter set. Specifically, each joint type (such as fillet joint, butt joint, T-joint) can be bound to a predefined optical characteristic parameter set. For example, the fillet joint corresponds to set A (including heat-affected zone width and maximum temperature), and the butt joint corresponds to set B (including heat-affected zone average temperature gradient and cooling time). The purpose is to quickly and accurately determine which optical characteristic parameters should be used for prediction based on the current joint type to be predicted.

[0041] The solution of the present application obtains the joint type of the mattress frame and, based on the pre-established correspondence between the joint type and the specific optical characteristic parameter set, can specifically determine the optical characteristic parameter set used for the current prediction. It is precisely because different joint types exhibit different thermophysical behaviors and stress distribution patterns during the welding process that their macroscopic deformation tendencies are affected to varying degrees by different optical characteristic parameters. It is difficult to capture these differences by relying solely on a general optical characteristic parameter set for prediction. The present application identifies the joint type and selects the specific optical characteristic parameter set corresponding to it, accurately selecting the parameter subset that has the most significant impact on the deformation of the current joint type from all the extracted optical characteristic parameters. It is precisely because of the use of more targeted optical characteristic parameters, combined with the pre-established mapping relationship between optical characteristic parameters and macroscopic deformation tendencies, that the determination of the macroscopic deformation tendency of the mattress frame under the initial welding parameters is more accurate, thereby providing a more reliable basis for subsequent welding parameter adjustments.

[0042] In some preferred embodiments, the present application is implemented as follows: First, before welding a mattress frame, the joint type of the current mattress frame is determined by reading a production order or using a machine vision system to identify the structural features of the frame to be welded. For example, it is determined that the frame primarily comprises corner joints and butt joints. Simultaneously, the system stores a pre-established correspondence, which can be a database or lookup table, recording which optical characteristic parameter sets correspond to different joint types (e.g., corner joints, butt joints, and T-joints). For example, the correspondence may specify that a specific optical characteristic parameter set for a corner joint includes heat-affected zone width, weld length, and maximum temperature; a specific optical characteristic parameter set for a butt joint includes heat-affected zone average temperature, temperature gradient, and cooling rate; and a specific optical characteristic parameter set for a T-join includes heat-affected zone area, peak temperature, and heat-affected zone shape characteristics. Next, based on the obtained joint type (e.g., corner joint and butt joint), the specific optical characteristic parameter sets to be used are determined from the correspondence: the set corresponding to the corner joint and the set corresponding to the butt joint. Then, from all optical characteristic parameters extracted from the heat-affected zone image of the test weld, parameters belonging to these specific sets are selected as the optical characteristic parameters used in the prediction. For example, if multiple parameters such as heat-affected zone width, maximum temperature, average temperature, temperature gradient, and cooling rate are extracted, and the current joint type requires prediction using heat-affected zone width and maximum temperature, only these two parameters are selected. Finally, these selected optical characteristic parameters are input into a pre-established mapping model between optical characteristic parameters and the macroscopic deformation tendency of the mattress frame to predict the macroscopic deformation tendency of the mattress frame under the current initial welding parameters.

[0043] Through the above technical solution, it is possible to select optical characteristic parameters that have a more significant impact on the deformation of different joint types of the mattress frame for prediction, avoiding the introduction of noise by using irrelevant or less influential parameters, thereby improving the accuracy of the prediction of the macroscopic deformation tendency of the mattress frame, providing more precise input for subsequent welding parameter optimization, and helping to improve the final molding accuracy of the mattress frame.

[0044] In some of the above embodiments of the present application, in step S103, adjusting the welding parameters according to the macro deformation tendency and the welding parameter compensation rule to obtain welding accuracy optimization parameters includes: obtaining a plurality of target improvement values corresponding to the plurality of deformation indicators; Determining a set of first welding parameter adjustment amounts that can make the overall deformation of the mattress frame approach the multiple target improvement values based on the multiple deformation indices, the multiple target improvement values, and the influence relationship between the welding parameter adjustment and the multiple deformation indices; The welding parameter adjustment amount is applied to the initial welding parameters to determine the welding accuracy optimization parameters.

[0045] The macro-deformation tendency includes at least multiple deformation indicators, including length, width, angle, and / or flatness. These indicators represent the deviation between the post-formed geometric shape of the mattress frame and design requirements, such as length deviation, width deviation, angle deviation, and flatness deviation. Together, these indicators represent the macro-forming accuracy of the mattress frame. The compensation rule includes the relationship between the influence of welding parameter adjustment on the multiple deformation indicators. This relationship represents the correlation between the adjustment of specific welding parameters (e.g., current, voltage, speed) and the change in each macro-deformation indicator of the mattress frame (e.g., length, width, angle, and flatness). This relationship can be linear or nonlinear and can be established using experimental data, simulation models, or empirical knowledge. Its purpose is to quantify the impact of parameter adjustment on deformation and provide a basis for subsequent parameter optimization. Multiple target improvement values corresponding to the multiple deformation indicators are obtained. These target improvement values represent the target state or allowable range that each macro-deformation indicator of the mattress frame is expected to achieve. For example, the target improvement value can be set to approach zero for each deformation indicator or to be within a specific tolerance range. The purpose is to provide a clear optimization direction and goal for parameter adjustment. A set of first welding parameter adjustments that can bring the overall deformation of the mattress frame closer to multiple target improvement values is determined. This adjustment refers to a set of adjustment value combinations of welding parameters found through calculation or search, so that after applying this set of adjustment values, the predicted macroscopic deformation indicators of the mattress frame can be closer to the preset multiple target improvement values as a whole. This process takes into account the comprehensive impact of various welding parameters on multiple deformation indicators and aims to achieve optimal control of overall deformation.

[0046] This solution decomposes the mattress frame's macroscopic deformation tendency into multiple quantifiable deformation indicators, such as length, width, angle, and flatness, and obtains multiple target improvement values corresponding to these indicators. Based on the predicted deformation indicators, target improvement values, and the pre-established relationship between the effects of welding parameter adjustments on these deformation indicators, the system calculates a set of welding parameter adjustments.

[0047] This technical solution comprehensively considers multiple macro-deformation indicators, such as the mattress frame's length, width, angle, and flatness, avoiding the negative impact of adjusting parameters solely for a single indicator. Based on the comprehensive impact of welding parameter adjustments on these multiple deformation indicators, a set of welding parameter adjustments can be determined that will bring the mattress frame's overall macro-deformation closer to multiple target improvement values, effectively controlling and optimizing the overall molding accuracy of the mattress frame.

[0048] In some of the aforementioned embodiments of the present application, a method is proposed to determine a set of first welding parameter adjustments that can make the overall deformation of the mattress frame approach the multiple target improvement values based on the macro deformation tendency, multiple target improvement values, and the influence relationship between the welding parameter adjustment on the multiple deformation indicators, including: Obtaining a plurality of actual deformation indicators measured after a preset number of mattress frame welding productions are performed using the first welding parameter adjustment amount; Determining whether the deviations between the multiple actual deformation indicators and the multiple target improvement values meet the preset correction trigger conditions; If the correction trigger condition is met, the influence relationship on multiple deformation indicators is adjusted according to the deviation and welding parameters, and the influence relationship is corrected to obtain a corrected influence relationship; the influence relationship is used to characterize the correlation between the adjustment amount of each welding parameter and the change amount of each deformation indicator; Based on the predicted multiple deformation indicators, multiple target improvement values, and the corrected influence relationship in the macro deformation tendency of the mattress frame, a set of first welding parameter adjustment amounts that can make the overall deformation of the mattress frame approach the multiple target improvement values is determined.

[0049] Among them, the preset number refers to the number of mattress frame samples used for actual production testing to obtain feedback data, the purpose of which is to obtain sufficient data to reflect the actual production situation without affecting the production efficiency too much; the preset correction trigger condition refers to the standard for judging whether it is necessary to correct the influence relationship between the welding parameter adjustment on multiple deformation indicators, which can be that the deviation between one or more actual deformation indicators and the target improvement value exceeds a specific threshold, or the deformation trend of multiple mattress frames produced continuously shows a systematic deviation, the purpose of which is to avoid unnecessary frequent corrections and ensure that timely adjustments are made when significant deviations occur; correcting the influence relationship refers to adjusting and updating the original correlation model or data between the welding parameter adjustment and the deformation indicator change based on the deviation data fed back from actual production, which can be achieved by machine learning algorithms or rule-based algorithms. The purpose is to make the influence relationship more accurately reflect the actual situation under the current production conditions; the revised influence relationship refers to the influence relationship of the welding parameter adjustment on multiple deformation indicators after the correction process. This relationship is closer to the parameter-deformation correlation in the actual production process than the original influence relationship. Its purpose is to provide a basis for more accurate determination of the welding parameter adjustment amount in the subsequent process; the influence relationship is used to characterize the association between the adjustment amount of each welding parameter and the change amount of each deformation indicator. It means that the influence relationship model or data structure describes how the various macro deformation indicators of the mattress frame will change when the value of a certain welding parameter or a group of welding parameters is specifically adjusted. It can be a linear or nonlinear mathematical model or a multidimensional lookup table. Its purpose is to quantify the degree and direction of the influence of the welding parameter adjustment on the deformation result.

[0050] In some preferred embodiments, specifically, a predetermined number of mattress frames, for example, 10 mattress frames, can be welded using a first welding parameter adjustment calculated based on the macroscopic deformation tendency and a preset influence relationship. Then, the actual macroscopic deformation indicators of these 10 mattress frames, such as length, width, diagonal length, and flatness, are measured using a three-dimensional coordinate measuring machine or a machine vision system, and the actual deformation indicator deviations corresponding to the target improvement values are calculated. These actual deformation indicator deviations are then compared with preset correction trigger conditions, such as determining whether the average length deviation exceeds 0.5 mm or whether the average flatness deviation exceeds 0.3 mm. If any of these trigger conditions is met, the influence relationship correction process is initiated. The correction process can include inputting the obtained actual deformation indicator deviation data and the corresponding first welding parameter adjustment into a correction algorithm. This algorithm can be an iterative optimization algorithm based on historical data and current deviations, which adjusts the parameters in the original influence relationship model so that the model prediction results are closer to the actual observed deviations, thereby obtaining a corrected influence relationship. Finally, using this revised influence relationship model, combined with the predicted macro deformation tendency and target improvement value, a more accurate set of welding parameter adjustments is recalculated for subsequent mass production.

[0051] In some of the above-mentioned embodiments of the present application, it is proposed to adjust the influence relationship on the deformation index according to the deviation and welding parameters, and to correct the influence relationship. The corrected influence relationship can be specifically to distribute the overall deviation proportionally to each welding parameter-deformation index relationship in a simple way, or to adjust the corresponding welding parameter influence relationship only according to the deformation index with the largest deviation. In this way, the influence relationship can be preliminarily adjusted. However, in its implementation process, in actual production, the influence relationship between the welding parameters and the deformation index is complex. It may not be possible to accurately identify the key factors that actually cause the deviation by making corrections based on the overall deviation in a simple way, resulting in insufficient accuracy of the corrected influence relationship, which in turn affects the optimization effect of the welding parameters.

[0052] In this regard, the present application further proposes that if the correction trigger condition is met, the steps of analyzing and determining the welding parameters that affect the deviation of each deformation index in the deviation in combination with the influence relationship include: Based on the analyzed and determined comprehensive influence of each welding parameter on the combination of multiple deformation index deviations in the obtained deviation, one or more welding parameter-deformation index associations having a preset contribution level to the formation of the deviation are identified from the obtained influence relationships; The strength of the association of one or more identified welding parameter-deformation index associations in the obtained influence relationship is adjusted to obtain a revised influence relationship.

[0053] Among them, the deviation of each deformation index in the deviation refers to the difference between the actually measured deformation index and the target improvement value. The deviation value is calculated for each specific deformation index, and its purpose is to quantify the degree of inaccuracy of each deformation index; wherein, the influence relationship refers to the correlation model or function between the adjustment amount of the welding parameter and the change amount of each macro deformation index of the mattress frame, which can be implemented by a linear model, a nonlinear model, a lookup table or a model based on machine learning, and its purpose is to characterize the influence of the change of the welding parameter on the deformation index; wherein, analyzing and determining the various welding parameters that affect the deviation of the deformation index means analyzing which changes in welding parameters will cause a specific deformation index to deviate according to the current influence relationship model, which can be achieved through model sensitivity analysis, partial derivative calculation or rule-based reasoning. Its purpose is to locate the key parameters that cause the deviation of a single deformation index; wherein, the combination of multiple deformation index deviations refers to the set or vector consisting of the deviations of all the deformation indexes of interest relative to their target improvement values in a single measurement, and its purpose is to comprehensively reflect the current overall deformation state; wherein, the comprehensive influence degree refers to the influence or contribution of the adjustment of a specific welding parameter or a group of welding parameters on the overall combination of the multiple deformation index deviations, and its purpose is to evaluate the contribution of each welding parameter to the overall deformation; wherein, the preset contribution level refers to a threshold or judgment condition used to measure whether the contribution of a welding parameter-deformation index association to the formation of the overall deviation is significant enough, which can be a numerical threshold, a ranking standard or a statistical significance judgment, and its purpose is to screen out the key associations that have a significant impact on the deviation; Among them, the welding parameter-deformation index association refers to the association term or coefficient between a specific welding parameter and a specific deformation index in the influence relationship model, and its purpose is to quantify the influence of a specific parameter on a specific index; wherein, the association strength refers to the numerical size or weight of the welding parameter-deformation index association in the influence relationship model, and its purpose is to quantify the degree of influence of the welding parameter on the deformation index.

[0054] The solution of the present application can more accurately trace the source of the problem by decomposing the overall deviation into the deviation of each deformation indicator and analyzing the influencing parameters corresponding to each deviation in combination with the influence relationship. Furthermore, by evaluating the comprehensive impact of each parameter on the deviation combination of multiple deformation indicators and identifying the key associations with preset contribution levels to the deviation, the blindness of making corrections based on a single indicator or overall deviation in a simple manner is avoided. This contribution-based identification of key associations and targeted adjustments makes the correction of the influence relationship more accurate and can more truly reflect the complex interactions in the actual welding process. This refined correction mechanism overcomes the problem of insufficient correction accuracy caused by simple correction methods in the prior art, provides a more reliable basis for subsequent parameter optimization, and significantly improves the molding precision control effect of the mattress frame. As a refined influence relationship correction mechanism, this solution can improve the accuracy of parameter adjustment based on influence relationships, thereby optimizing the overall molding precision control process.

[0055] In some preferred embodiments, the macroscopic deformation indicators of the mattress frame may include length deviation, width deviation, and angle deviation. Welding parameters may include welding current, welding voltage, and welding speed. The influence relationship can be a linear model that characterizes the linear relationship between the adjustments in welding current, welding voltage, and welding speed and the changes in length deviation, width deviation, and angle deviation. Specifically, if a correction trigger condition is met, such as a significant difference between the actual measured length deviation, width deviation, and angle deviation and the target improvement value, the deviation of each deformation indicator within these deviations is first obtained. For example, the actual length deviation is +2 mm, the actual width deviation is -1 mm, and the actual angle deviation is +0.5 degrees. Based on the current influence relationship model, the welding parameters that influence these deviations are analyzed and determined. For example, model analysis reveals that the current settings of welding current, welding voltage, and welding speed all affect the actual deviations of length, width, and angle. Next, the combined impact of the determined welding current, welding voltage, and welding speed on the combination of multiple deformation indicator deviations (+2 mm, -1 mm, +0.5 degrees) is evaluated. For example, the expected impact of each parameter adjustment on the deviation combination is calculated, and a quantitative contribution is calculated based on the importance of each deformation indicator. Then, from the current influence relationship, one or more welding parameter-deformation indicator associations with a preset contribution level to the deviation are identified. For example, if the calculated contribution of welding current to the overall deviation is the highest and exceeds a preset threshold, the associations between welding current and length deviation and welding current and width deviation are identified as key associations. If the contribution of welding voltage also exceeds the threshold, the associations between welding voltage and width deviation and welding voltage and angle deviation are also identified as key associations. Finally, the strength of the associations of the identified one or more welding parameter-deformation indicator associations within the current influence relationship is adjusted. For example, if the positive impact of welding current on length is underestimated, the coefficient corresponding to welding current-length deviation in the influence relationship model is increased. If the negative impact of welding voltage on width is overestimated, the coefficient corresponding to welding voltage-width deviation in the influence relationship model is decreased. Through this targeted adjustment, a revised influence relationship is obtained.

[0056] In some of the above-mentioned embodiments of the present application, it is proposed to adjust the influence relationship of multiple deformation indicators according to the deviation and welding parameters, correct the influence relationship, and obtain a corrected influence relationship. The corrected influence relationship can be specifically determined by analyzing the deviation of each deformation indicator in the deviation, determining the various welding parameters that affect the deviation, and based on the comprehensive influence of each welding parameter on the deviation combination of multiple deformation indicators in the deviation, identifying one or more groups of welding parameter-deformation indicator associations with a preset contribution level to the deviation, and then adjusting the correlation strength of these identified associations in the influence relationship. In this way, according to the deviations occurring in actual production, the influence model between welding parameters and deformation indicators can be corrected in a targeted manner, thereby improving the model's fit to the actual situation. However, in its implementation process, how to identify the welding parameter-deformation indicator association that has an important influence on the formation of deviations from the complex influence relationship and make targeted adjustments to improve the correction efficiency and accuracy is a problem that needs to be solved.

[0057] In this regard, the present application further proposes the steps of identifying one or more welding parameter-deformation index associations having a preset contribution level to the formation of the deviation from the obtained influence relationship based on the comprehensive influence of each welding parameter determined by analysis on the combination of multiple deformation index deviations in the obtained deviation, including: For each welding parameter included in the influence relationship, the expected influence value of the adjustment of the welding parameter on the deviation of each deformation index in the obtained deviation is calculated; Determining, based on the expected influence value of each welding parameter and the actual deviation of each deformation index in the obtained deviation, a quantitative contribution of each welding parameter to the deviation combination of multiple deformation indicators in the currently obtained deviation; Compare the quantified contribution degree with the preset contribution level judgment condition, and select the welding parameters that meet the contribution level judgment condition; The association formed by the screened welding parameters and the corresponding deformation indicators when calculating the expected influence value is used as one or more sets of welding parameter-deformation indicator associations identified from the obtained influence relationship and having a preset contribution level to the formation of the currently obtained deviation.

[0058] The influence relationship is a model or data structure that characterizes the association between the adjustment of each welding parameter and the change in each deformation index. It can be implemented using a matrix, function, lookup table, or machine learning model. The expected impact value is the predicted degree of change in the deviation of a specific deformation index when a specific welding parameter is adjusted, as predicted by the influence relationship model. This can be obtained using linear calculation, nonlinear function mapping, or model inference. The quantitative contribution is a numerical measure of the combined impact of a welding parameter on the overall deviation of multiple currently observed deformation indices. This can be determined using weighted summation, correlation analysis, or model-based contribution analysis. The contribution level judgment condition is the criterion used to screen out welding parameters with a significant impact on the deviation. This can be set using a threshold, ranking selection, or statistical significance test. The welding parameter-deformation index association refers to the portion of the influence relationship that connects a specific welding parameter to a specific deformation index, characterizing the strength or pattern of the parameter's influence on the index. This can be represented by elements in a matrix, coefficients in a function, or specific connection weights in a model.

[0059] The solution of this application evaluates the potential effect of each parameter on each deformation by calculating the expected impact of adjusting each welding parameter on the deviation of each deformation index. Based on this, combined with the actual deviation of each deformation index measured, the actual contribution of each welding parameter to the current overall deviation is quantified. By comparing the quantified contribution with preset judgment criteria, the key welding parameters that play a major role in the formation of the current deviation can be effectively screened. Furthermore, the associations formed by these screened key welding parameters and their corresponding deformation indicators are identified as key associations with a preset contribution level to the current deviation extracted from the complex influence relationship. This approach eliminates the need for blind global adjustments to the influence relationship and instead focuses on the local associations that have the greatest impact on the actual deviation, thereby improving the efficiency and accuracy of the correction. This targeted identification and correction method, combined with the overall correction of the influence relationship in the basic solution, can converge to the optimal welding parameters more quickly and accurately, effectively controlling the molding accuracy of the mattress frame.

[0060] In some preferred embodiments, specifically, a currently used influence relationship model can be first obtained, such as a matrix R, where R(i, j) represents the influence coefficient of the adjustment of the i-th welding parameter on the change in the j-th deformation index. For each welding parameter i included in the influence relationship, the expected influence value E(i, j) of its adjustment on each deformation index j in the obtained deviation is calculated. This influence value can be calculated using the model R. Then, based on these expected influence values E(i, j) and the actual deviation D(j) of each deformation index j in the obtained deviation, the quantitative contribution C(i) of each welding parameter i to the combination of multiple deformation index deviations in the current deviation is determined. For example, the contribution C(i) can be calculated based on the degree of match or correlation between the expected influence value E(i, j) and the actual deviation D(j), taking into account the importance of each deformation index. Next, the calculated quantitative contribution C(i) is compared with a preset contribution level judgment condition, such as setting a threshold T, and screening out all welding parameters i whose contribution C(i) is greater than or equal to T. Finally, the association (i, j) formed by these screened welding parameters i and all the corresponding deformation indicators j when calculating the expected influence value is used as one or more sets of welding parameter-deformation indicator associations identified from the influence relationship R and having a preset contribution level to the current deviation.

[0061] Furthermore, this embodiment proposes that the steps of extracting preset optical characteristic parameters from the heat-affected zone image include: obtaining the heat-affected zone image; analyzing the heat-affected zone image to identify the interference area in the image; processing the heat-affected zone image based on the identification result of the interference area to obtain a processed image; and extracting the preset optical characteristic parameters from the processed image.

[0062] Among them, the heat-affected zone image refers to an image containing the weld joint and its nearby base material area, which is collected by imaging equipment during or after welding; the preset optical feature parameters refer to quantitative indicators extracted from the heat-affected zone image that can reflect the welding heat input, material state or deformation trend, which can be realized by grayscale distribution features, texture features, geometric features or spectral features; the interference area refers to the optical information of the heat-affected zone image that is not the heat-affected zone itself, but is introduced by external factors or the acquisition process and may affect the accuracy of feature extraction. It can be realized by welding spatter, reflective points, shadows, sensor bad points or environmental debris; processing the heat-affected zone image refers to preprocessing or enhancing the heat-affected zone image using image processing technology to reduce or eliminate the influence of the interference area and improve the image quality. It can be realized by filtering, denoising, contrast adjustment, morphological operation, region segmentation or repair and other methods.

[0063] The solution of the present application optimizes the extraction process of optical characteristic parameters by adding the steps of identifying and processing the interference area in the image before extracting the preset optical characteristic parameters from the heat-affected zone image. First, the heat-affected zone image is obtained as data input. Next, the image is analyzed to identify the interference area, which makes the subsequent processing targeted and avoids blindly processing the entire image and possibly damaging the effective information. Then, based on the identified interference area, the image is selectively or locally processed to remove or reduce the damage to the image pixel value or structure caused by the interference factors, thereby obtaining a clearer processed image. Finally, the optical characteristic parameters are extracted from the processed image. Since the influence of the interference factors is reduced, the extracted parameters can more accurately reflect the true state of the heat-affected zone. This optimized feature extraction process provides more reliable input data for subsequent deformation prediction and parameter adjustment, so that the entire mattress frame molding precision control method can more effectively cope with the challenges brought by material and specification switching, reduce macro deformation, and improve product consistency.

[0064] In some preferred embodiments, the heat-affected zone image can be obtained by photographing the welding area with an industrial camera. The interference area can be identified by using a segmentation method based on a pixel value threshold to identify highlight spatter points, or a texture analysis method can be used to identify areas with textures different from the heat-affected zone as interference. When processing the heat-affected zone image, the identified spatter point area can be smoothed by using a local median filter; the noise in the image can be denoised by using a Gaussian filter; and the uneven illumination can be processed by using a homomorphic filter or a local contrast enhancement method. When extracting the preset optical feature parameters from the processed image, the average grayscale value, grayscale variance, the number or density of edge pixels after edge detection, the texture descriptor of a specific area, etc. of the heat-affected zone can be calculated as the preset optical feature parameters.

[0065] By adding the step of identifying and processing interference areas in the image before extracting optical characteristic parameters, the above technical solution effectively reduces the impact of factors such as ambient lighting changes, weld spatter, and sensor noise on image quality. The processed image more accurately reflects the true optical properties of the heat-affected zone, resulting in higher accuracy and reliability of the optical characteristic parameters extracted from the processed image. This improves the ability of the optical characteristic parameters to characterize welding heat input and deformation, providing a more reliable data foundation for subsequent macroscopic deformation tendency prediction, enhancing the effectiveness of welding parameter adjustments, and ultimately helping to improve the molding precision and production consistency of mattress frames.

[0066] In some embodiments, based on the expected impact value of each welding parameter and the actual deviation of each deformation indicator in the obtained deviation, determining the quantitative contribution of each welding parameter to the combination of multiple deformation indicator deviations in the currently obtained deviation includes: Obtain the expected impact value of each welding parameter on the deviation of each deformation index, and obtain the actual deviation of each deformation index in the deviation; Performing preset numerical processing on the expected influence value of each welding parameter obtained and the actual deviation amount of each deformation index obtained to obtain a numerical processing result; Obtain the influencing factors of each deformation index on the final molding quality of the mattress frame; Based on the influencing factors of each deformation index obtained, a comprehensive calculation is performed in combination with the numerical processing results to determine the quantitative contribution of each welding parameter to the deviation combination of multiple deformation indexes in the currently obtained deviation.

[0067] The expected impact value refers to the theoretical and expected change in the deviation of a specific deformation index caused by the adjustment of welding parameters; the actual deviation refers to the actual deviation of the deformation index of the mattress frame from the target value or design value in actual production; the preset numerical processing refers to a predetermined mathematical transformation of the expected impact value and the actual deviation, the purpose of which is to eliminate the dimensional differences or numerical range differences between different data to make them comparable, which can be achieved by function mapping or normalization processing; the influence factor refers to a pre-set weight or coefficient that reflects the degree of influence of different deformation indicators on the final molding quality of the mattress frame, which can be determined based on expert experience, historical data analysis, or quality requirements; the comprehensive calculation refers to the process of combining the numerical processing results with the influence factor, the purpose of which is to comprehensively consider the expected impact, actual deviation, and importance to obtain a more comprehensive quantitative contribution, which can be achieved by weighted summation, multi-factor product model, or other mathematical models; the quantitative contribution refers to the numerical value obtained through comprehensive calculation that measures the quantitative influence of each welding parameter on the combination of multiple deformation index deviations in the currently obtained deviation.

[0068] In some preferred embodiments, the specific implementation is as follows. Assume that it is necessary to determine the quantitative contribution of welding parameters P1 and P2 to the deformation deviation of the current mattress frame, which is primarily reflected in the length deviation D1, width deviation D2, and angle deviation D3. First, the expected impact value of parameter P1 on D1, D2, and D3, as well as the expected impact value of parameter P2 on D1, D2, and D3, are obtained. Simultaneously, actual measurement results of the current mattress frame are obtained to obtain the actual deviations of D1, D2, and D3. Next, these expected impact values and actual deviations are numerically processed. For example, all values are normalized to the range of 0 to 1 to obtain the normalized expected impact values and normalized actual deviations. Simultaneously, based on preset quality requirements, the impact factors of D1, D2, and D3 on the final molding quality of the mattress frame are obtained. For example, the importance factor of D1 is 0.4, D2 is 0.3, and D3 is 0.3. Finally, a comprehensive calculation is performed based on these importance factors and the normalized values. For example, the quantitative contribution of parameter P1 to the overall deviation can be calculated using a weighted summation method: Contribution (P1) = (Normalized Expected Impact (P1, D1) * Normalized Actual Deviation (D1) * Importance Factor (D1)) + (Normalized Expected Impact (P1, D2) * Normalized Actual Deviation (D2) * Importance Factor (D2)) + (Normalized Expected Impact (P1, D3) * Normalized Actual Deviation (D3) * Importance Factor (D3)). The quantitative contribution of parameter P2 is calculated similarly. This method allows for more accurate contribution values for each parameter, taking into account the actual degree of deviation and the importance of each indicator.

[0069] This technical solution eliminates data discrepancies by numerically processing the expected impact values and actual deviations. By introducing importance factors, the contribution assessment is more closely aligned with actual quality requirements. A comprehensive calculation based on the numerical processing results and importance factors more accurately quantifies the contribution of each welding parameter to the current overall deformation deviation. This more accurate quantification of contribution provides a more reliable basis for subsequent welding parameter adjustments, helping to more effectively identify the key parameters that cause deviations, enabling more targeted parameter optimization and ultimately improving the molding accuracy of the mattress frame.

[0070] In some of the above-mentioned embodiments of the present application, it is proposed to determine the quantitative contribution of each welding parameter to the deviation combination of multiple deformation indicators in the currently obtained deviation based on the expected influence value of each welding parameter and the actual deviation of each deformation indicator in the obtained deviation. The determination of the quantitative contribution can be specifically through a simple linear model or empirical formula, and the expected influence value and the actual deviation are directly substituted into the calculation, so that the influence of the parameters can be preliminarily evaluated. However, in its implementation process, directly processing the expected influence value and the actual deviation may lead to distorted processing results due to differences in their units, dimensions or numerical ranges, and cannot accurately reflect the true influence of the welding parameters on the deformation indicators, thereby affecting the accuracy and efficiency of the welding parameter adjustment.

[0071] In this regard, the present application further proposes performing a preset numerical processing on the expected influence values of the obtained welding parameters and the actual deviations of the obtained deformation indicators. The steps of obtaining the numerical processing results include: Obtain the expected impact value of each welding parameter and the actual deviation of each deformation index in the deviation; Perform function mapping or normalization processing on the expected impact value and the actual deviation according to preset rules; The numerical processing result is determined based on the expected impact value and actual deviation after function mapping or normalization.

[0072] The expected impact value refers to the theoretical numerical effect that a welding parameter adjustment would have on the deviation of a deformation index, assuming an influence relationship holds. This can be achieved through multiplication using an influence relationship matrix or through predictive calculations based on a regression model. The actual deviation refers to the actual numerical difference between the mattress frame deformation index measured during actual welding production and the target value. This can be achieved through direct measurement using measuring equipment or indirect acquisition through image processing techniques. Preset numerical processing refers to the preprocessing performed on the original expected impact value and actual deviation before calculating the quantitative contribution. This can be achieved through methods such as function mapping or normalization. The numerical processing result refers to the intermediate data obtained after the preset numerical processing and used for subsequent quantitative contribution calculation. This can be represented by a processed numerical vector or matrix. The preset rule refers to the set standard or algorithm used to guide the function mapping or normalization of the expected impact value and actual deviation. This can be achieved through rules based on data distribution characteristics, expert experience, or machine learning models. Function mapping transforms data values by applying a mathematical function. This can be achieved using logarithmic, exponential, sigmoid, or custom nonlinear functions. Normalization scales data to a specific range or distribution. This can be achieved using minimum-maximum normalization, Z-score normalization, or decimal scaling.

[0073] In some preferred embodiments, specifically, after obtaining the expected impact value of each welding parameter and the actual deviation of each deformation index in the deviation, these values can be processed according to preset rules. For example, the preset rules can be set as follows: if the distribution of the data shows obvious skewness, a logarithmic function is used for function mapping; if the data distribution is relatively uniform but the numerical range varies greatly, a minimum-maximum normalization process is used to scale the values to the range of [0, 1]. Assume that for a certain welding parameter, its expected impact value on length deformation is a larger value, while its expected impact value on angular deformation is a smaller value; at the same time, there are also numerical differences in the actually measured length deviation and angular deviation. When performing normalization processing, all expected impact values and all actual deviations can be normalized separately so that they are all within the range of [0, 1]. For example, the expected contribution of length might be normalized from the original 10 mm / A to 0.8, and the expected contribution of angle might be normalized from the original 0.5 degrees / A to 0.3. The actual length deviation might be normalized from the original 5 mm to 0.7, and the actual angle deviation might be normalized from the original 2 degrees to 0.9. Based on these normalized values, combined with the relative importance factors of each deformation indicator, a comprehensive calculation is performed to determine the quantitative contribution of each welding parameter to the overall deviation. This approach ensures fair comparability when calculating contributions from data of different dimensions or numerical ranges.

[0074] In addition, this application further proposes a mattress frame forming precision control system for adjusting mattress frame welding parameters, such as Figure 2 As shown, the system includes: Acquisition module 201 is used to weld a test piece using a preset set of initial welding parameters before formal mass production, collect an image of the heat-affected zone located in a preset observation area of the test piece, and extract optical characteristic parameters of the heat-affected zone image; the test piece is a mattress frame sample undergoing trial welding testing before mass production; a determination module 202 for determining the macro deformation tendency of the mattress frame under the initial welding parameters based on a pre-established mapping relationship between the optical characteristic parameters and the macro deformation tendency of the mattress frame and the optical characteristic parameters; The parameter adjustment module 203 is used to adjust the initial welding parameters according to the macro deformation tendency of the mattress frame under the initial welding parameters and the welding parameter compensation rules to obtain the welding accuracy optimization parameters; The production welding module 204 uses welding precision optimization parameters to perform batch welding production of mattress frames.

[0075] Among them, the acquisition module refers to a functional unit used to obtain image data related to the welding process and extract specific information from it, which can be implemented by a combination of an image acquisition device and an image processing unit; the determination module refers to a functional unit used to infer the deformation trend of the mattress frame under specific conditions based on input data and a preset model, which can be implemented by a combination of a computing unit and a storage unit; the parameter adjustment module refers to a functional unit used to calculate and output parameters for optimizing the welding process based on prediction results and preset rules, which can be implemented by a combination of a computing unit and a control interface; the production welding module refers to a functional unit used to receive optimized welding parameters and perform actual welding operations, which can be implemented by welding equipment.

[0076] In some preferred embodiments, the acquisition module may specifically include an industrial camera for capturing images of the heat-affected zone of the test weld, and an image processing unit connected to the camera for analyzing the images and extracting optical characteristic parameters. The determination module and the parameter adjustment module may be integrated into an industrial control computer, which runs a prediction model algorithm and a parameter adjustment algorithm. The prediction model algorithm calculates the macro deformation tendency based on the optical characteristic parameters extracted by the image processing unit and the mapping relationship stored in the computer memory. The parameter adjustment algorithm calculates the adjustment amount of the welding parameters based on the predicted macro deformation tendency and preset compensation rules. The production welding module may be a welding robot system that receives the welding accuracy optimization parameters output by the industrial control computer through a communication interface and performs the welding operation of the mattress frame according to these parameters.

[0077] The above description is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, various modifications and variations of the present application are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A mattress frame molding precision control method, characterized in that: include: Before formal mass production, a test weld is performed using a preset set of initial welding parameters, an image of the heat-affected zone located in a preset observation area of the test weld is collected, and optical characteristic parameters of the heat-affected zone image are extracted; the test weld is a mattress frame sample that is tested for test welding before mass production; determining the macro deformation tendency of the mattress frame under the initial welding parameters according to a pre-established mapping relationship between the optical characteristic parameters and the macro deformation tendency of the mattress frame; According to the macro deformation tendency of the mattress frame under the initial welding parameters and the welding parameter compensation rules, the initial welding parameters are adjusted to obtain welding accuracy optimization parameters; The welding precision optimization parameters are used to carry out batch welding production of the mattress frames.

2. A mattress frame molding precision control method according to claim 1, characterized in that: The determining of the macro deformation tendency of the mattress frame under the initial welding parameters based on the pre-established mapping relationship between the optical characteristic parameters and the macro deformation tendency of the mattress frame includes: Obtaining the joint type of the mattress frame and the pre-established correspondence between the joint type and a set of specific optical characteristic parameters for predicting macro deformation tendency; the joint type includes corner joint, butt joint, and T-joint; Determining the specific optical characteristic parameter set from the corresponding relationship according to the connector type, and selecting parameters corresponding to the specific optical characteristic parameter set from the extracted optical characteristic parameters as optical characteristic parameters used in current prediction; Based on the optical characteristic parameters currently used in the prediction and the pre-established mapping relationship between the optical characteristic parameters and the macro deformation tendency of the mattress frame, the macro deformation tendency of the mattress frame under the initial welding parameters is determined.

3. A mattress frame molding precision control method according to claim 1, characterized in that: The macro deformation tendency includes at least a plurality of deformation indicators including length, width, angle and / or flatness; the compensation rule includes the influence relationship between the welding parameter adjustment and the plurality of deformation indicators; The step of adjusting the initial welding parameters according to the macro deformation tendency and the welding parameter compensation rule to obtain welding accuracy optimization parameters includes: obtaining a plurality of target improvement values corresponding to the plurality of deformation indicators; Determining, based on the multiple deformation indices, the multiple target improvement values, and the influence of the welding parameter adjustment on the multiple deformation indices, a set of first welding parameter adjustment amounts capable of causing the overall deformation of the mattress frame to approach the multiple target improvement values; The first welding parameter adjustment amount is applied to the initial welding parameter to determine a welding accuracy optimization parameter.

4. A mattress frame molding precision control method according to claim 3, characterized in that: The step of determining, based on the multiple deformation indices, the multiple target improvement values, and the influence of the welding parameter adjustment on the multiple deformation indices, a set of first welding parameter adjustment amounts capable of causing the overall deformation of the mattress frame to approach the multiple target improvement values comprises: Obtaining a plurality of actual deformation indicators measured after a preset number of mattress frame welding operations are performed using the first welding parameter adjustment amount; Determining whether deviations between the multiple actual deformation indicators and the multiple target improvement values meet preset correction trigger conditions; If the correction trigger condition is met, the influence relationship is corrected according to the deviation and the influence relationship of the welding parameter adjustment on the multiple deformation indicators to obtain a corrected influence relationship; the influence relationship is used to characterize the correlation between the adjustment amount of each welding parameter and the change amount of each deformation indicator; Based on the multiple deformation indicators in the predicted macro deformation tendency of the mattress frame, the multiple target improvement values, and the corrected influence relationship, a set of first welding parameter adjustment amounts that can make the overall deformation of the mattress frame approach the multiple target improvement values is determined.

5. A mattress frame molding precision control method according to claim 4, characterized in that: If the correction trigger condition is met, adjusting the influence relationship on the multiple deformation indicators according to the deviation and the welding parameters, correcting the influence relationship, and obtaining a corrected influence relationship, includes: According to the deviation amount of each deformation index in the deviation, combined with the influence relationship, analyzing and determining each welding parameter that affects the deviation amount of the deformation index; Based on the analyzed and determined comprehensive influence of each welding parameter on the obtained combination of multiple deformation index deviations, one or more welding parameter-deformation index associations having a preset contribution level to the formation of the deviation are identified from the obtained influence relationships; The strength of the association between the identified one or more welding parameter-deformation index associations in the acquired influence relationship is adjusted to obtain a revised influence relationship.

6. A mattress frame molding precision control method according to claim 5, characterized in that: The method further comprises: determining, based on the analysis of the comprehensive influence of each welding parameter on the obtained deviation combination of multiple deformation index deviations, identifying, from the obtained influence relationship, one or more welding parameter-deformation index associations having a preset contribution level to the formation of the deviation, comprising: For each welding parameter included in the influence relationship, calculating an expected influence value of adjustment of the welding parameter on each deformation index deviation in the obtained deviation; Determining, based on the expected impact value of each welding parameter and the actual deviation of each deformation indicator in the obtained deviation, a quantitative contribution of each welding parameter to a combination of deviations of multiple deformation indicators in the currently obtained deviation; Comparing the quantified contribution degree with a preset contribution level judgment condition, and screening out welding parameters that meet the contribution level judgment condition; The association formed by the screened welding parameters and the deformation indicators corresponding to the expected influence values when calculating the expected influence values is used as one or more sets of welding parameter-deformation indicator associations identified from the obtained influence relationship and having a preset contribution level to the formation of the currently obtained deviation.

7. A mattress frame molding precision control method according to claim 1, characterized in that: The collecting of the heat-affected zone image located in the preset observation area of the test weld and extracting the optical characteristic parameters of the heat-affected zone image includes: Acquire the heat-affected zone image; analyze the heat-affected zone image to identify interference areas in the image; Based on the recognition result of the interference area, the image of the heat-affected zone is processed to obtain a processed image; The optical characteristic parameters of the heat-affected zone image are extracted from the processed image.

8. A mattress frame molding precision control method according to claim 6, characterized in that: The step of determining, based on the expected impact value of each welding parameter and the actual deviation of each deformation indicator in the obtained deviation, a quantitative contribution of each welding parameter to a combination of deviations of multiple deformation indicators in the currently obtained deviation comprises: Obtaining the expected impact value of each welding parameter on the deviation of each deformation index, and obtaining the actual deviation of each deformation index in the deviation; Performing preset numerical processing on the expected influence value of each welding parameter obtained and the actual deviation amount of each deformation index obtained to obtain a numerical processing result; Obtain the influencing factors of each deformation index on the final molding quality of the mattress frame; Based on the influencing factors and in combination with the numerical processing results, a comprehensive calculation is performed to determine the quantitative contribution of each welding parameter to the deviation combination of multiple deformation indicators in the currently obtained deviation.

9. A mattress frame molding precision control method according to claim 8, characterized in that: The performing of a preset numerical processing on the expected influence value of each welding parameter obtained and the actual deviation amount of each deformation index obtained to obtain a numerical processing result includes: Obtaining the expected impact value of each welding parameter and the actual deviation of each deformation index in the deviation; Performing function mapping or normalization processing on the expected impact value and the actual deviation respectively according to preset rules; The numerical processing result is determined based on the expected impact value and the actual deviation after the function mapping or normalization processing.

10. A mattress frame forming precision control system, used for adjusting mattress frame welding parameters, characterized in that: The system includes: An acquisition module is used to weld a test piece using a preset set of initial welding parameters before formal mass production, capture an image of the heat-affected zone located in a preset observation area of the test piece, and extract optical characteristic parameters of the heat-affected zone image; the test piece is a mattress frame sample undergoing a test welding test before mass welding production; a determination module, configured to determine the macro deformation tendency of the mattress frame under the initial welding parameters based on a pre-established mapping relationship between the optical characteristic parameters and the macro deformation tendency of the mattress frame; A parameter adjustment module, configured to adjust the initial welding parameters according to the macro deformation tendency of the mattress frame under the initial welding parameters and welding parameter compensation rules to obtain welding accuracy optimization parameters; A welding module is produced, and the welding precision optimization parameters are used to carry out batch welding production of the mattress frame.

Images