Multi-product-line-oriented robot welding integration method, device and equipment and medium

Abstract

The invention relates to a multi-product-line-oriented robot welding integration method, device and equipment and a medium, and the method comprises the steps that three-dimensional geometrical characteristics and material attribute data of multiple types of workpieces are collected, and geometric parameters and material mechanical properties of weld joints are extracted; an order demand template is intelligently matched, and initial welding parameters and a clamp configuration scheme are generated; the health state of production equipment is evaluated in real time, and the collision-free movement track of the mechanical arm is planned; dynamically analyzing the joint angle sequence to generate a high-precision servo control signal; the welding process is monitored in real time, and current and molten pool temperature parameters are corrected; and iteratively optimizing the algorithm weight of the process knowledge base to improve the adaptive capacity. Through cooperation of multi-source data fusion and an intelligent algorithm, rapid switching of multiple kinds of workpieces, self-adaptive compensation of the equipment operation state and closed-loop optimization of welding process parameters are achieved, and the equipment utilization rate and the welding quality stability in a flexible production scene are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of robotic welding, and particularly to a robotic welding integration method, device, equipment and medium for multi-production lines. Background Art

[0002] In the field of industrial manufacturing, with the rapid transformation of the production mode towards multi-variety and small-batch customization, flexible production technology has become a key pillar for improving manufacturing efficiency and market competitiveness. This technology enables a production line to adapt to the requirements of multiple product specifications, playing an irreplaceable and important role in promoting the transformation and upgrading of the manufacturing industry. However, in the current process of realizing flexible production, especially in the scenario of multi-variety mixed-line welding, there are still many deficiencies in existing methods, making it difficult to fully meet the requirements of rapid switching and efficient production.

[0003] Existing solutions are often restricted by equipment and systems when dealing with multi-variety production, showing problems such as long production switching cycles and low efficiency. Especially when frequent equipment or program adjustments are required, a large amount of time and resources are wasted. These limitations directly affect the response speed and overall effectiveness of the production line, and there is an urgent need for new technological breakthroughs to make up for the shortcomings.

[0004] Focusing on specific challenges, firstly, there is the adaptability problem of production equipment when facing products of different specifications. The traditional fixed tool design is difficult to adjust quickly, resulting in a large amount of time-consuming manual adjustments every time the product specification is switched. This inefficient switching method further leads to the problem of program adaptation. Because there is no intelligent system that can automatically match product features and processing requirements, parameters need to be reset and paths verified every time, increasing the additional workload. These problems are interrelated, ultimately disrupting the production rhythm and making it difficult to achieve the goal of true flexible production. Summary of the Invention

[0005] Based on this, the purpose of the present invention is to provide a robotic welding integration method, device, equipment and medium for multi-production lines that can achieve rapid switching of multi-variety workpieces, adaptively and dynamically adjust welding parameters and motion paths, and optimize production efficiency through intelligent learning.

[0006] The purpose of the present invention is achieved by the following solutions: In the first aspect, the present invention provides a robotic welding integration method for multi-production lines, including the following steps: S1: Collect and extract data on the three-dimensional geometric features and material properties of multi-variety workpieces, and call the dynamic parameter adaptation algorithm in the process knowledge base to generate product specification data including weld geometric parameters and material mechanical properties; S2: Perform a similarity matching process on the geometric features of the workpiece in the product specification data with the preset templates in the order requirement database to generate initial configuration data including the fixture configuration plan and the initial welding current and voltage parameters; S3: Perform a health value evaluation process on the real-time operating status data of the acquired production equipment to generate an equipment evaluation result, and based on the equipment evaluation result and the initial configuration data, call the path planning algorithm in the process knowledge base to calculate the robotic arm movement trajectory, and generate equipment adjustment data including the robotic arm joint angle adjustment sequence and the fixture displacement trajectory; S4: Perform a motion control instruction conversion process on the equipment adjustment data, parse the joint angle sequence and the displacement trajectory to generate pulse control signals, and send the pulse control signals to the actuator. The pulse control signals are used to control the actuator to drive the robotic arm and the fixture to move along the planned path; S5: Based on the actual position data feedback by the actuator, perform real-time monitoring and dynamic correction processing on the current fluctuation and the molten pool temperature distribution during the welding process to generate dynamic welding parameters adapted to the working condition changes, and send the dynamic welding parameters to the welding power supply. The dynamic welding parameters are used to adjust the arc energy output; S6: Optimize the weight parameters of the dynamic parameter adaptation algorithm and the path planning algorithm in the process knowledge base according to the dynamic welding parameters and the historical production data to generate updated configuration mapping relationship data.

[0007] In one embodiment, S1 of a robot welding integration method provided by the present invention for multiple product lines specifically includes the following steps: S11: Perform three-dimensional line laser scanning on the surface three-dimensional geometric features of the workpiece to be processed by a three-dimensional line laser scanner to generate original point cloud data including millions of spatial coordinate points; S12: Perform outlier removal processing on the original point cloud data based on the normal vector consistency detection algorithm, calculate the angle between the normal vector of each point and the neighborhood average normal vector, and generate noise-reduced geometric point cloud data; S13: Perform weld seam trajectory extraction processing on the geometric point cloud data, calculate the inclination angle of the weld seam cross-section normal vector and the cumulative value of the Euclidean distance between adjacent points through the principal component analysis method, and generate a geometric parameter set including the weld seam length, inclination angle and gap size; S14: Perform material property association processing on the geometric parameter set, detect the yield strength and thermal conductivity of the workpiece material through laser-induced breakdown spectroscopy technology, and fuse the geometric features and physical property indicators to generate product specification data including weld seam geometric parameters and material mechanical properties.

[0008] In one embodiment, S2 of a robot welding integration method provided by the present invention for multiple product lines specifically includes the following steps: S21: Perform multi-dimensional feature vector construction processing on product specification data, extract the weld length, inclination angle, and material yield strength as feature dimensions, and generate a standardized feature vector. S22: Perform similarity matching processing on the standardized feature vector and the preset template in the order demand database, calculate the cosine value of the angle between vectors based on the cosine similarity algorithm, and generate a similarity score matrix. S23: Perform threshold screening processing on the similarity score matrix, select the template items with score values greater than the preset threshold, and generate initial configuration data including the fixture electromagnetic locking scheme and the initial welding current and voltage parameters.

[0009] In one embodiment, S3 of a robot welding integration method for multi-product lines provided by the present invention specifically includes the following steps: S31: Perform multi-source sensing data synchronous acquisition processing on the servo motor torque, cylinder pressure, and guide rail wear amount of the production equipment, and generate a device real-time status data set including time stamps. S32: Perform health quantification processing on the device real-time status data set based on the weighted summation formula, calculate the device health index, and generate a device evaluation result for evaluating the device health status. S33: Perform three-dimensional path planning processing on the device health index, call the path planning algorithm in the process knowledge base to search for a collision-free motion trajectory in the voxel grid space, and generate an initial path node sequence. S34: Perform health compensation processing on the initial path node sequence, calculate the robotic arm joint angle adjustment amount, and generate device adjustment data including the robotic arm joint angle adjustment sequence and the fixture displacement trajectory. The calculation formula for the robotic arm joint angle adjustment amount is: ; Where [[]]is the robotic arm joint angle adjustment amount, representing the target angle of the i [[]]th joint, [[]]is the inverse kinematics function, [[]]is the end effector target pose, where [[]]represents the position of the end effector in three-dimensional space, [[]]is the rotation angle of the end effector, [[]]is the tolerance compensation coefficient, [[]]is the maximum value 1 of the health index, [[]], [[]]is the device health index.

[0010] In one embodiment, S4 of a robot welding integration method for multi-product lines provided by the present invention specifically includes the following steps: S41: Perform inverse kinematic solution on the joint angle sequence in the device adjustment data, and inversely solve the joint angles corresponding to the end effector pose through the D-H parameter model of the robotic arm to generate the six-axis joint angle adjustment amount; S42: Perform pulse signal conversion processing on the joint angle adjustment amount, calculate the pulse frequencies of each joint drive, and generate the pulse control signal for the servo motor; S43: Perform bus transmission protocol encapsulation processing on the pulse control signal, and send the pulse control signal to the servo driver through the EtherCAT real-time industrial Ethernet.

[0011] In one embodiment, S5 of a robotic welding integration method provided by the present invention for multiple product lines specifically includes the following steps: S51: Based on the actual position data fed back by the actuator, perform multi-modal sensing synchronous acquisition processing on the current and molten pool temperature during the welding process. Collect the current waveform data through a Hall sensor, obtain the molten pool temperature field distribution matrix through an infrared thermal imager, and generate a process monitoring data set; S52: Perform abnormal fluctuation analysis on the process monitoring data set, calculate the percentage value of the current standard deviation to the mean, and generate a quality warning signal; S53: Perform dynamic parameter correction processing on the quality warning signal based on the molten pool balance equation, adjust the wire feeding speed, and generate dynamic welding parameters including current, voltage, and speed. The adjustment formula for the wire feeding speed is: ; where is the adjusted wire feeding speed, is the initial wire feeding speed, is the material thermal deformation coefficient, is the difference between the measured temperature of the molten pool and the target temperature, is the reference temperature value.

[0012] In one embodiment, S6 of a robotic welding integration method provided by the present invention for multiple product lines specifically includes the following steps: S61: Perform correction amount feature extraction processing on the dynamic welding parameters, calculate the current adjustment amount and voltage adjustment amount, and generate a parameter correction feature set; S62: Perform weight optimization processing on the dynamic parameter adaptation algorithm in the process knowledge base based on the parameter correction feature set, and update the feature matching weight coefficient of the dynamic parameter adaptation algorithm according to the welding quality qualification rate in the historical production data; S63: Perform weight optimization processing on the path planning algorithm in the process knowledge base based on the path tracking deviation amount in the dynamic welding parameters, and update the trajectory smoothness weight coefficient of the path planning algorithm according to the path execution accuracy index in the historical production data.

[0013] In a second aspect, the present invention provides a robot welding integration device for multiple product lines, which is configured with the following modules: A data acquisition and preprocessing module, which is used to collect and extract the three-dimensional geometric features and material properties of workpieces of multiple varieties, call the dynamic parameter adaptation algorithm in the process knowledge base, and generate product specification data including weld geometric parameters and material mechanical properties; A configuration parameter matching module, which is used to perform a similarity matching process on the geometric features of the workpieces in the product specification data with the preset templates in the order demand database, and generate initial configuration data including fixture configuration schemes and initial welding current and voltage parameters; A device parameter adjustment module, which is used to perform a health value evaluation process on the real-time operation status data of the production equipment obtained, generate a device evaluation result, and based on the device evaluation result and the initial configuration data, call the path planning algorithm in the process knowledge base to calculate the robot arm movement trajectory, and generate device adjustment data including the robot arm joint angle adjustment sequence and the fixture displacement trajectory; A device pulse control module, which is used to perform a motion control instruction conversion process on the device adjustment data, parse the joint angle sequence and the displacement trajectory to generate a pulse control signal, and send the pulse control signal to the actuator. The pulse control signal is used to control the actuator to drive the robot arm and the fixture to move along the planned path; A welding dynamic correction module, which is used to perform real-time monitoring and dynamic correction processing on the current fluctuation and the molten pool temperature distribution during the welding process based on the actual position data fed back by the actuator, generate dynamic welding parameters adapted to the working condition changes, and send the dynamic welding parameters to the welding power source. The dynamic welding parameters are used to adjust the arc energy output; A process knowledge base update module, which is used to optimize the weight parameters of the dynamic parameter adaptation algorithm and the path planning algorithm in the process knowledge base according to the dynamic welding parameters and the historical production data, and generate updated configuration mapping relationship data.

[0014] In a third aspect, the present application provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements any one of the above-mentioned robot welding integration methods for multiple product lines.

[0015] In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements any one of the above-mentioned robot welding integration methods for multiple product lines.

[0016] In summary, the robot welding integration method for multi-product lines provided by the present invention can achieve rapid process adaptation and dynamic precise control in the scenario of mixed-line production of multiple varieties through the collaboration of multi-source data fusion and intelligent algorithms, and can effectively solve the core problems of insufficient flexibility, response lag and low intelligent level of traditional welding systems. The robot welding integration method for multi-product lines provided by the present invention constructs a closed-loop control link from data perception, intelligent decision-making to precise execution, and can achieve rapid switching and adaptation of multiple varieties of workpieces, adaptive compensation of equipment performance and continuous optimization of process parameters without manual intervention, providing a systematic solution for flexible welding production.

[0017] For better understanding and implementation, the present invention will be described in detail below with reference to the accompanying drawings. Brief Description of the Drawings

[0018] Figure 1 It is a schematic flow chart of a robot welding integration method for multi-product lines provided by an embodiment of the present application; Figure 2 It is a schematic flow chart of generating equipment adjustment data provided by an embodiment of the present application; Figure 3 It is a schematic flow chart of generating dynamic welding parameters including current, voltage and speed provided by an embodiment of the present application; Figure 4 It is a schematic structural diagram of a robot welding integration device for multi-product lines provided by another embodiment of the present application. Detailed Embodiment

[0019] To facilitate the understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. Preferred embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, these embodiments are provided to make the understanding of the disclosure of the invention more thorough and comprehensive.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this invention belongs. The terms used in the description of the present invention herein are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The term “and / or” used herein includes any and all combinations of one or more of the related listed items.

[0021] In one embodiment, as Figure 1As shown, a robot welding integration method for multiple product lines is provided. In this embodiment, taking the application of this method to a terminal as an example, it can be understood that this method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps: S1: Collect and extract data on the three-dimensional geometric features and material properties of multi-variety workpieces, and call the dynamic parameter adaptation algorithm in the process knowledge base to generate product specification data including weld geometric parameters and material mechanical properties.

[0022] Specifically, the system can use a coordinate measuring machine to perform contact measurement on the key points on the surface of the workpiece to obtain accurate three-dimensional coordinate data, including point cloud data and surface contour information. At the same time, in combination with a laser scanner that emits a laser beam and receives the reflected light, three-dimensional point cloud data of the workpiece surface is obtained through the principle of triangulation. The laser scanner is suitable for the rapid measurement of complex curved surfaces and flexible workpieces, and supplements the measurement of the overall shape and local detail features. In addition, the system can also use structured light three-dimensional scanning technology. By projecting a specific grating pattern onto the surface of the workpiece through a structured light projector, an industrial camera captures the deformed grating image, and based on the principle of parallax, the three-dimensional coordinate information of the workpiece surface is calculated, further improving the measurement accuracy and speed. At the same time, through material composition analysis instruments, mechanical property testing equipment, etc., the material properties such as the strength, toughness, hardness, and thermal expansion coefficient of the workpiece material are detected and data is collected, providing a basic basis for determining subsequent welding parameters. After the data is collected, the system uses professional CAD software and data processing algorithms for digital processing and feature extraction to obtain key geometric feature parameters of the workpiece, such as length, width, height, curvature, angle, etc., and material property feature parameters, such as strength grade, hardness range, composition ratio, etc., forming a standardized feature data set, laying a foundation for subsequent welding process planning and parameter matching.

[0023] After the data collection is completed, the system deeply analyzes and processes the collected massive three-dimensional geometric feature data, identifies and extracts the shape features of the workpiece, such as determining whether the workpiece has complex curved surfaces and special-shaped structures, and determining representative geometric feature parameters such as the edge contours of key parts, so as to simplify the original large amount of complex data into key information that can directly represent the geometric characteristics of the workpiece, laying a solid foundation for the adaptation of subsequent process parameters.

[0024] Subsequently, the system automatically invokes the dynamic parameter adaptation algorithm stored in the process knowledge base, and intelligently performs complex calculations and logical judgments based on the input workpiece geometric feature and material property data. By matching and analyzing a large number of pre-stored welding process data models, the algorithm can generate weld geometric parameters that match the current workpiece, including specific values such as the width, depth, and groove angle of the weld. At the same time, combining with the material mechanical property data, it further determines the mechanical property performance parameters of the material during the welding process, and finally integrates to form complete and accurate product specification data, providing comprehensive and targeted basic data support for the subsequent welding process planning to ensure that the welding process can accurately adapt to the characteristics and requirements of different workpieces.

[0025] S2: Perform a similarity matching process on the geometric features of the workpiece in the product specification data and the preset templates in the order demand database to generate initial configuration data including the fixture configuration plan and the initial welding current and voltage parameters.

[0026] Specifically, the preset templates cover typical patterns of various workpiece geometric features and reasonable ranges of corresponding process parameters. The system can use a similarity matching algorithm based on vector similarity calculation, spatial geometric transformation, and machine learning technology to perform precise calculations from multiple key dimensions such as workpiece shape, size, and contour, and find the preset template that is most similar to the current workpiece geometric feature one by one, and calculate the specific similarity value in a quantitative way, which is used as an important basis for determining the optimal matching result.

[0027] After determining the best matching template, the system automatically generates initial configuration data including a detailed fixture configuration plan and the initial welding current and voltage parameters according to the template and its corresponding process requirement information. In the process of generating the fixture configuration plan, the system comprehensively analyzes various factors such as the workpiece fixing method, positioning accuracy requirements, and production efficiency goals associated with the template, selects a suitable fixture type from the fixture library, and determines its specific specification parameters and layout method to ensure that the workpiece can be stably and accurately fixed and positioned during the welding process, thus ensuring the smooth progress of the welding operation.

[0028] At the same time, the system combines the welding process experience data related to the matching template, and through calculation and analysis, determines the initial welding current and voltage parameters that match the current workpiece. The determined parameters fully consider the influence of various factors such as the material properties, size specifications, and welding position of the workpiece, and can provide preliminary and reasonable process parameter guidance for the subsequent welding production, enabling the welding process to start under relatively appropriate process conditions and creating favorable conditions for further optimizing the welding quality.

[0029] S3: Evaluate the health value of the real-time operating status data of the production equipment obtained, generate equipment evaluation results, and based on the equipment evaluation results and initial configuration data, call the path planning algorithm in the process knowledge base to calculate the robotic arm motion trajectory, and generate equipment adjustment data including the robotic arm joint angle adjustment sequence and the fixture displacement trajectory.

[0030] Specifically, the system collects various status information of the production equipment during operation in real time, including the angle, speed, and acceleration data of each joint of the robotic arm, the wear detection value and operation accuracy index of the fixture, and the working status signals of various sensors. These rich data obtained in real time are input into a pre-constructed equipment health evaluation model. Based on advanced data fusion technology, fault diagnosis algorithms, and machine learning methods, the model comprehensively analyzes and evaluates the performance indicators of each component of the equipment. Through learning and training on a large amount of historical fault data and normal operation data, the model can accurately judge the health status of each component of the equipment, and then calculate a health value that can quantitatively reflect the overall performance status and reliability level of the equipment.

[0031] Based on the evaluated equipment health value and initial configuration data, the system calls the path planning algorithm suitable for the current production scenario from the process knowledge base. This algorithm comprehensively considers various factors such as the constraint conditions of the equipment health status, the geometric characteristics of the workpiece, the layout position of the fixture, and the initial welding parameters, and uses complex mathematical operations and optimization techniques to accurately calculate the optimal motion trajectory of the robotic arm during the welding operation. This trajectory includes the detailed adjustment sequence of the angles of each joint of the robotic arm over time, as well as key information such as the corresponding position displacement trajectory of the fixture at different welding stages, so as to generate complete and accurate equipment adjustment data. These data can guide the equipment to move in the most efficient and stable manner during the welding process, ensure the quality and efficiency of the welding operation, and at the same time fully consider the actual operating conditions of the equipment, avoid accelerating the wear of the equipment due to overuse of certain parts, and extend the overall service life of the equipment.

[0032] S4: Perform motion control instruction conversion processing on the equipment adjustment data, parse the joint angle sequence and displacement trajectory to generate pulse control signals, and send the pulse control signals to the actuator. The pulse control signals are used to control the actuator to drive the robotic arm and the fixture to move along the planned path.

[0033] Specifically, the system parses in detail the robotic arm joint angle adjustment sequence and the fixture displacement trajectory in the equipment adjustment data. These data contain the target angles of each joint of the robotic arm during the welding process and the displacement position information of the fixture at different time points. The system will convert these target positions and angles into corresponding pulse control signals according to the kinematic models of the robotic arm and the fixture.

[0034] For a joint of the robotic arm, the system calculates the number of pulses to be sent and the timing of the pulse sequence based on its pulse equivalent and the target angle change. At the same time, the system also takes into account the kinematic constraints of the robotic arm and the fixture, such as the speed limit and acceleration limit of the joint, to ensure that the generated pulse control signal can make the device operate smoothly within the allowable motion range. The generated pulse control signal is sent to the corresponding actuators, such as the servo motor driver of the robotic arm and the displacement actuator of the fixture. After receiving the pulse signal, these actuators drive the robotic arm and the fixture to move along the planned motion trajectory precisely. For example, the servo motor driver of the robotic arm accurately controls the rotation speed and rotation angle of the motor according to the frequency and number of the pulse signal, so that the joint of the robotic arm moves according to the predetermined angle sequence, ensuring that the welding torch can accurately reach each position of the weld seam.

[0035] During the execution process, the system also monitors the feedback signals of the actuators in real time, such as the actual position data fed back by the encoder of the motor. These feedback data are used for closed-loop control. The system compares the actual position with the target position, calculates the deviation value, and adjusts the subsequent pulse control signal according to the deviation value to achieve high-precision motion control. For example, if a small deviation is detected between the actual position and the target position of the robotic arm joint, the system immediately sends a correction pulse signal to quickly adjust the joint to the correct position. Through this real-time feedback and correction mechanism, the system can effectively overcome external interference and the errors of the device itself, ensuring that the welding equipment moves strictly according to the planned path, thus guaranteeing the stability and reliability of the welding quality.

[0036] S5: Based on the actual position data fed back by the actuator, monitor and dynamically correct the current fluctuation and the molten pool temperature distribution during the welding process in real time, generate dynamic welding parameters adapted to the change of working conditions, and send the dynamic welding parameters to the welding power source. The dynamic welding parameters are used to adjust the arc energy output.

[0037] Specifically, the system uses sensor technology to collect the current fluctuation data and the molten pool temperature distribution information during the welding process in real time. The current fluctuation data is obtained in real time through a high-precision current sensor, reflecting the stability change of the welding current; the molten pool temperature distribution is measured non-contact by an infrared thermal imager or a fiber optic temperature sensor to obtain the real-time temperature field information of the molten pool area.

[0038] The system compares and analyzes these real-time monitoring data with the preset welding process requirements, and uses a dynamic correction algorithm to timely adjust abnormal situations during the welding process. When it detects that the current fluctuation exceeds the allowable range or the molten pool temperature distribution does not meet the expectation, the dynamic correction algorithm quickly calculates and generates dynamic welding parameters adapted to the changes in the working conditions according to the current working conditions, such as adjusting the magnitude of the welding current and voltage, changing the welding speed or wire feeding speed, etc. At the same time, the system sends these dynamic welding parameters to the welding power source, and the welding power source adjusts the arc energy output in real time according to the received parameter instructions to ensure the stability of the welding process and the reliability of the welding quality. In addition, the system stores and analyzes the real-time monitoring data to provide data support for subsequent process optimization and quality control.

[0039] S6: Optimize the weight parameters of the dynamic parameter adaptation algorithm and the path planning algorithm in the process knowledge base according to the dynamic welding parameters and historical production data, and generate updated configuration mapping relationship data.

[0040] Specifically, the system collects the dynamic welding parameters generated during this welding process, including the actually adjusted parameter values such as welding current, voltage, welding speed, wire feeding speed, etc., and the quality inspection data during the welding process, such as the forming quality of the weld, penetration depth, undercut situation, etc. At the same time, the system retrieves a large amount of past welding task data stored in the historical production database, and these data cover the welding parameters and quality records under different workpiece types, welding positions, welding materials and other conditions.

[0041] The system integrates and preprocesses these data, and extracts the feature information related to the algorithm weight optimization. For example, for the dynamic parameter adaptation algorithm, the system analyzes the change trends of the weights of various welding parameters under different workpiece specifications and material properties, and the correlation between these weights and the welding quality indicators. For the path planning algorithm, the system studies the influence of the weight adjustment of various influencing factors such as weld length, welding angle, obstacle avoidance requirements, etc. in the path planning on the welding efficiency and quality under different weld geometries and welding positions.

[0042] Based on data analysis, the system uses optimization methods in data mining techniques and machine learning algorithms, such as gradient descent method, genetic algorithm, etc., to iteratively optimize the weight parameters in the algorithm. For example, during the optimization process of the dynamic parameter adaptation algorithm, the system constructs an optimization model with the welding quality index as the objective function and the weights of various welding parameters as variables. By continuously adjusting the weight parameters, calculating the objective function value, and updating the weights according to the rules of the optimization algorithm, the weight combination that optimizes the welding quality index is finally found. The new configuration mapping relationship data can more accurately reflect the association and mapping rules between different workpiece specifications, equipment states, and welding working conditions. For example, in subsequent welding tasks, when encountering workpieces with similar specifications and materials, the system can generate initial welding process parameters and equipment adjustment plans faster and more accurately according to the updated configuration mapping relationship. At the same time, the optimized algorithm weight parameters can also enable the system to have stronger adaptability and optimization ability in the face of complex and changeable welding environments, thereby continuously improving the intelligent level of the welding system and the stability of welding quality, realizing the continuous improvement and optimization of the welding process, and providing strong technical support for the improvement of the enterprise's production efficiency and cost reduction.

[0043] In summary, the robot welding integration method for multi-product lines provided by the present invention can achieve rapid process adaptation and dynamic precise control in the multi-variety mixed-line production scenario through the collaboration of multi-source data fusion and intelligent algorithms, and can effectively solve the core problems of insufficient flexibility, lagging response, and low intelligent level of traditional welding systems. Specifically, through the automatic acquisition and extraction of three-dimensional geometric features and material properties, combined with the dynamic parameter adaptation algorithm of the process knowledge base, the system can construct a product specification data system that accurately represents the workpiece characteristics to achieve the precise mapping of process requirements and equipment capabilities; based on the intelligent matching mechanism of the order demand template, it can quickly generate fixture configuration plans and initial welding parameters suitable for different workpieces, significantly shortening the process preparation cycle when switching product specifications; through the dynamic invocation of real-time equipment health assessment and path planning algorithms, it can generate a robotic arm motion trajectory that incorporates equipment performance degradation compensation under complex working conditions to ensure the stability of the processing process and positioning accuracy; relying on the parsing conversion of pulse control signals and the collaboration of actuators, it can achieve servo drive control with high dynamic response to ensure the precise reproduction of complex trajectories; based on the real-time monitoring of multi-physical field data during the welding process and the dynamic parameter correction, a closed-loop feedback mechanism of the molten pool state and energy output can be established to suppress process fluctuations and improve the forming quality; through the iterative optimization of algorithm weights driven by historical production data, the intelligent decision-making ability of the process knowledge base can be continuously improved, forming a self-evolving production control system.

[0044] The robot welding integration method for multiple product lines provided by the present invention constructs a closed-loop control link from data perception, intelligent decision-making to precise execution, enabling rapid switching and adaptation of multi-variety workpieces, adaptive compensation of equipment performance, and continuous optimization of process parameters without manual intervention, providing a systematic solution for flexible welding production.

[0045] In one embodiment, step S1 of the robot welding integration method for multiple product lines provided by the present invention specifically includes the following steps: S11: Perform three-dimensional line laser scanning on the surface three-dimensional geometric features of the workpiece to be measured and processed through a three-dimensional line laser scanner to generate raw point cloud data containing millions of spatial coordinate points.

[0046] Specifically, the system performs three-dimensional line laser scanning on the surface three-dimensional geometric features of the workpiece to be measured and processed through a three-dimensional line laser scanner. The scanner emits a high-frequency laser line, and the laser line forms stripes on the workpiece surface. These stripes are deformed due to the undulation of the workpiece surface. The system uses a high-speed camera synchronized with the laser scanner to capture the deformed laser stripe images, and applies the phase shift algorithm. Based on the Moiré fringe principle, the height information corresponding to each pixel point is calculated through the phase information of multiple images, and then raw point cloud data containing millions of spatial coordinate points is generated. These data accurately reflect the macroscopic three-dimensional geometric features of the workpiece surface, providing basic data support for subsequent geometric feature extraction and analysis. During this process, the system ensures the high resolution and high precision of the point cloud data through a high-precision optical system and advanced image processing algorithms, providing reliable data guarantee for subsequent processing steps.

[0047] S12: Perform outlier removal processing on the raw point cloud data based on the normal vector consistency detection algorithm, calculate the angle between the normal vector of each point and the average normal vector of the neighborhood, and generate denoised geometric point cloud data.

[0048] Specifically, in three-dimensional geometric data, outliers are usually incorrect data points caused by noise or interference during the scanning process, and these outliers will affect subsequent feature extraction and analysis. The normal vector consistency detection algorithm identifies outliers by calculating the angle between the normal vector of each point and the average normal vector of other points in the neighborhood. Specifically, the system calculates the normal vector of each point and then compares it with the average normal vector of the points in the neighborhood. If the angle between the normal vector of a certain point and the average normal vector of the neighborhood exceeds the set threshold, then this point is determined as an outlier and removed. This process can effectively remove noise data and generate denoised geometric point cloud data, providing a clearer and more accurate data basis for subsequent weld seam trajectory extraction. Through this outlier removal processing, the quality of the data can be significantly improved, ensuring the reliability of subsequent analysis.

[0049] S13: Perform weld seam trajectory extraction and processing on the geometric point cloud data. Calculate the inclination angle of the normal vector of the weld seam cross-section and the cumulative value of the Euclidean distances between adjacent points through the principal component analysis method, and generate a set of geometric parameters including the weld seam length, inclination angle, and gap size.

[0050] Specifically, the system uses the principal component analysis (PCA) method to calculate the inclination angle of the normal vector of the weld seam cross-section and the cumulative value of the Euclidean distances between adjacent points. The PCA method identifies the main directions of data variation through statistical analysis of the point cloud data, thereby determining the orientation and cross-sectional characteristics of the weld seam. Specifically, the system first identifies the weld seam area in the point cloud data, which is usually achieved by analyzing the density, curvature, and directionality of the point cloud data. Then, the system extracts a series of cross-sections along the weld seam direction within the weld seam area and performs PCA analysis on the point cloud data within each cross-section. By calculating the covariance matrix of the point cloud data within each cross-section and solving its eigenvectors, the normal vector of the cross-section is obtained. The inclination angle of the normal vector is determined by comparison with the reference coordinate system, which provides key parameters for subsequent welding posture adjustment. At the same time, the system calculates the Euclidean distances between adjacent points and accumulates them to obtain the weld seam length and gap size. This process involves complex mathematical operations and geometric modeling. Through efficient numerical calculation algorithms and geometric reconstruction techniques, the system can accurately extract the weld seam trajectory in a short time. The finally generated set of geometric parameters details the geometric characteristics of the weld seam, including the weld seam length, inclination angle, gap size, and curvature changes, etc., providing comprehensive data support for subsequent welding process planning and equipment configuration.

[0051] S14: Perform material property association processing on the set of geometric parameters. Detect the yield strength and thermal conductivity of the workpiece material through laser-induced breakdown spectroscopy technology, fuse geometric features and physical property indicators, and generate product specification data including weld seam geometric parameters and material mechanical properties.

[0052] Specifically, the system uses Laser-Induced Breakdown Spectroscopy (LIBS) technology to detect mechanical properties such as the yield strength and thermal conductivity of the workpiece material. The LIBS technology excites a plasma on the material surface through a high-energy laser pulse, and then analyzes the spectrum emitted by the plasma to determine the elemental composition and performance indicators of the material. The system first controls the laser emitter to irradiate the workpiece surface with precise pulse energy and frequency, exciting the plasma. Then, a spectrometer is used to collect the spectral signals emitted by the plasma, and advanced spectral analysis algorithms are used to process and analyze the signals. The system can identify the characteristic spectral lines of the main elements in the material and calculate mechanical property indicators such as the yield strength and thermal conductivity of the material based on a pre-established material property database. These material attribute data are fused with the geometric parameter set to generate complete product specification data.

[0053] The fusion process involves data format conversion, coordinate system unification, and feature parameter matching. The system ensures the accuracy and integrity of the data through efficient data processing algorithms and database management technologies. The finally generated product specification data not only includes the geometric features of the weld, but also includes the mechanical properties of the material, such as yield strength, thermal conductivity, melting point, etc. These data provide comprehensive information support for subsequent welding process planning and equipment configuration, enabling the welding system to formulate the optimal welding strategy according to the specific geometric and material characteristics of the workpiece, thereby improving the welding quality and production efficiency.

[0054] In one embodiment, S2 of a robot welding integration method for multi-product lines provided by the present invention specifically includes the following steps: S21: Perform multi-dimensional feature vector construction processing on the product specification data, extract the weld length, inclination angle, and material yield strength as feature dimensions, and generate a standardized feature vector.

[0055] Specifically, the system extracts the weld length, inclination angle, and material yield strength from the product specification data as feature dimensions. The weld length data is in meters, and the system converts it to the [0, 1] interval through a standardization algorithm. The specific method is to calculate the relative difference between this value and the minimum and maximum values in the dataset, thereby ensuring that welds of different lengths can be compared on a unified scale. For the inclination angle data, the system extracts it in radians, with a range between 0 and π, and then realizes standardization through trigonometric function transformation. For example, using the sine or cosine function to map it to the [-1, 1] interval to ensure the linear separability of the angle information in subsequent calculations. The material yield strength data is in Pascals, and the system uses a logarithmic function to compress it to reduce the numerical range and highlight the relative differences. After standardization processing, these data are integrated into a multi-dimensional feature vector.

[0056] The multi-dimensional feature vector is presented in matrix form, where each row represents a workpiece instance and each column corresponds to a feature dimension. The system ensures that the numerical values of each feature dimension are on the same scale through matrix operations, thus avoiding certain features dominating the similarity calculation results due to their large numerical ranges. For example, the unnormalized weld length may vary within a range of several meters, while the radian value of the inclination angle is usually between 0 and π. Without normalization, the variation in weld length may dominate the similarity calculation and obscure the differences in inclination angle and material properties. Through normalization, the system ensures that each feature dimension has equal importance in the similarity matching process, providing a solid foundation for subsequent similarity matching. The generated normalized feature vector can not only accurately reflect the key geometric and material characteristics of the workpiece but also provide a unified data format for subsequent matching algorithms, ensuring the accuracy and efficiency of the calculation.

[0057] S22: Perform a similarity matching process on the normalized feature vector and the preset templates in the order demand database, calculate the cosine value of the angle between vectors based on the cosine similarity algorithm, and generate a similarity score matrix.

[0058] Specifically, the system obtains the feature vectors of all preset templates from the order demand database. These templates also contain three feature dimensions: weld length, inclination angle, and material yield strength, and have been normalized. The system calculates the dot product of the input feature vector and each preset template vector through dot product operations, that is, the sum of the products of the corresponding dimension values, which reflects the direction consistency of the two vectors in the feature space. At the same time, the system calculates the norm of each preset template vector, that is, the square root of the sum of the squares of the vector dimension values, which represents the magnitude of the template vector in the feature space.

[0059] The cosine similarity value is obtained by dividing the dot product by the product of the norms of the two vectors. The value ranges from -1 to 1. The closer the value is to 1, the more similar the two vectors are; the closer the value is to -1, the less similar they are; and 0 indicates orthogonality or irrelevance. The system organizes all the calculated cosine similarity values into a similarity score matrix, where each row corresponds to the feature vector of an input product specification data, each column corresponds to a preset template, and each element in the matrix represents the similarity score between the corresponding row and column. In this way, the system can comprehensively quantify the similarity relationship between the input data and all preset templates, providing an objective and quantitative basis for subsequent template selection. The generation process of the similarity score matrix involves a large number of matrix operations and numerical calculations. The system ensures high efficiency and high precision even when processing large-scale data through an efficient linear algebra library and parallel computing technology, providing reliable data support for the subsequent threshold screening step.

[0060] S23: Perform threshold screening on the similarity scoring matrix, select the template items with scoring values greater than the preset threshold, and generate initial configuration data including the fixture electromagnetic locking scheme and the initial welding current and voltage parameters.

[0061] Specifically, the preset threshold of the template item is determined based on historical matching data and expert experience, and is used to distinguish valid matches from invalid matches. The system automatically selects those template items with scoring values greater than the preset threshold, and considers that these templates have sufficient similarity with the current product specification data and can be used as reliable references. For the selected template items, the system extracts the fixture electromagnetic locking scheme and the initial welding current and voltage parameters from them, and these information are integrated into the initial configuration data. The initial configuration data not only includes the specific locking method of the fixture during the welding process, such as the specific position of electromagnetic locking, the magnitude of the locking force and other detailed parameters, but also provides the basic current and voltage settings when the welding equipment starts, such as the magnitude of the initial welding current, the set value of the voltage and other specific parameters. Through this threshold screening process, the system ensures that only the template items that highly match the current product specification data are selected, thereby improving the accuracy and applicability of the initial configuration data. This process is achieved through mathematical operations and logical judgments, without manual intervention, ensuring the objectivity and efficiency of the processing, and providing accurate parameter guidance for the subsequent welding process execution.

[0062] In one of the embodiments, as Figure 2 shown, S3 of a robot welding integration method for multi-product lines provided by the present invention specifically includes the following steps: S31: Synchronously collect and process multi-source sensing data of the servo motor torque, cylinder pressure and guide rail wear amount of the production equipment, and generate a device real-time status data set including time stamps.

[0063] Specifically, during the collection process, the system obtains the output torque data of the servo motor in real time through high-precision torque sensors. These sensors are based on the strain gauge principle and can accurately measure the torsional stress of the motor shaft and convert it into an electrical signal. At the same time, the system uses pressure sensors to collect the real-time pressure values of the cylinders. These sensors use the piezoresistive effect to convert the pressure change into a resistance change and then into a voltage signal. In addition, the system uses wear detection sensors to accurately measure the wear amount of the guide rail. These sensors detect the microscopic topography changes on the surface of the guide rail through laser displacement measurement technology, so as to quantify the wear degree. The collected analog signals are preprocessed through a signal conditioning circuit for operations such as amplification and filtering to improve the quality and stability of the signals.

[0064] The system adds timestamps to each collected data point to ensure data synchronization and timeliness. The preprocessed data is integrated into a real-time device status dataset containing timestamps, which is stored in a structured manner, ensuring data integrity and availability, and providing accurate raw data support for subsequent device health assessment. Through this multi-source data fusion technology, the system ensures comprehensive and accurate monitoring of the production equipment status, providing a reliable data foundation for subsequent health assessment and path planning.

[0065] S32: Quantify the health of the real-time device status dataset based on the weighted summation formula, calculate the device health index, and generate a device evaluation result for evaluating the device health status.

[0066] Specifically, the system first determines the influence weights of three parameters, namely servo motor torque, cylinder pressure, and guide rail wear amount, on the device health status according to the device's historical operation data and expert experience. The determination of the weights is optimized by analyzing historical fault data and device performance decay curves, combined with expert knowledge, to ensure the rationality of weight allocation. The system normalizes each parameter using the min-max normalization method to uniformly adjust the numerical range of each parameter to the interval [0, 1] to eliminate the influence of dimensions and dimension ranges. Subsequently, the system can use the weighted summation formula to multiply the normalized parameter values by their corresponding weights and sum them to calculate the device health index. This index intuitively reflects the health status of the device in a quantified numerical form, where 0 indicates that the device is completely faulty and 1 indicates that the device is in the best state. The system integrates the calculated device health index into the device evaluation result, which not only includes the specific value of the health index but also a detailed analysis of the contribution of each parameter, providing a quantitative basis for subsequent path planning and device adjustment. Through this quantification process, the system can monitor the health changes of the device in real time, providing scientific support for production scheduling and maintenance decisions.

[0067] S33: Perform three-dimensional path planning on the device health index, call the path planning algorithm in the process knowledge base to search for collision-free motion trajectories in the voxel grid space, and generate an initial path node sequence.

[0068] Specifically, the system calls the path planning algorithm in the process knowledge base and performs voxelization on the working space of the production line according to the equipment health index. The voxelization process divides the complex three-dimensional space into multiple small voxel units to form a voxel grid space, and each voxel unit has uniform size and coordinate information. In the voxel grid space, the system uses the equipment health index as a constraint condition, combines the geometric features of the workpiece and the weld position information, and searches for a motion trajectory that meets the collision-free requirement. The path planning algorithm comprehensively considers the kinematic model of the robotic arm, the size and position limitations of the fixture, and the impact of the equipment health index on the motion speed and accuracy.

[0069] Preferably, the system can adopt an improved ant colony algorithm to iteratively search for the optimal path in the voxel grid space. This algorithm simulates the foraging behavior of ants, uses the pheromone update mechanism and path heuristic rules, and gradually converges to the optimal solution. Finally, the system generates an initial path node sequence, which details a series of key node positions and pose information of the robotic arm in the three-dimensional space. Each node contains the angles, positions of the joints of the robotic arm, and the state parameters of the fixture. Through this path planning process, the system ensures the efficiency and safety of the welding operation, while considering the impact of the equipment health status on path planning, and improves the reliability of the production process and the welding quality.

[0070] S34: Perform health compensation processing on the initial path node sequence, calculate the robotic arm joint angle adjustment amount, and generate equipment adjustment data including the robotic arm joint angle adjustment sequence and the fixture displacement trajectory.

[0071] Specifically, the system reads the node information of the initial path node sequence and extracts the target position and pose data of each joint of the robotic arm. Combining the equipment health assessment results, analyze the health status of each joint to determine whether there are health problems such as wear. For the problem joints, calculate their actual motion range and accuracy changes through the inverse kinematics algorithm, and determine the joint angle compensation amount. The compensation amount calculation depends on the joint health index and the kinematic model. A lower health index results in a larger compensation amount to ensure that the actual trajectory of the robotic arm is close to the initial planned path. The system applies the compensation amount to the corresponding joint positions of the initial path nodes to generate equipment adjustment data including the robotic arm joint angle adjustment sequence. At the same time, considering the fixture displacement trajectory, adjust the position and pose of the fixture according to the adjusted motion trajectory of the robotic arm and the workpiece fixing constraint conditions to ensure the stable positioning of the workpiece. The generated equipment adjustment data contains the target position and pose of the robotic arm joints after adjustment and the fixture displacement trajectory information, providing an accurate instruction set for the subsequent control of the welding equipment, ensuring the smooth progress of the welding operation, and improving the welding quality and the service life of the equipment. Among them, the calculation formula for the robotic arm joint angle adjustment amount is: ; Where, is the robotic arm joint angle adjustment amount, indicating thei The target angle of the joint is the inverse kinematics function is the target pose of the end effector, where represents the position of the end effector in three-dimensional space is the rotation angle of the end effector is the tolerance compensation coefficient is the maximum value of the health index 1 , is the equipment health index

[0072] The above-provided robot welding integration method for multi-product lines can achieve real-time and accurate perception of the operating state of production equipment through synchronous acquisition of multi-source sensing data and quantitative evaluation of equipment health, and can effectively solve the problems of reduced motion accuracy and inaccurate path planning caused by equipment performance attenuation in traditional welding systems; based on the collision-free path search algorithm in voxel grid space and the path planning strategy call mechanism combined with the process knowledge base, it can generate an initial motion trajectory that takes into account both safety and efficiency to meet the rapid path adaptation requirements under complex working conditions; through the joint angle compensation algorithm driven by the equipment health index, it can dynamically fuse equipment performance parameters during the path planning stage, and can achieve precision self-healing control of the manipulator motion trajectory, significantly reducing pose deviations caused by factors such as mechanical wear and load fluctuations; this technical system constructs a closed-loop adjustment link from state perception, path planning to compensation control, and can achieve dynamic matching of equipment health status and motion control parameters without manual intervention, providing core guarantee for the adaptive operation of equipment in multi-variety welding scenarios.

[0073] In one of the embodiments, S4 of the robot welding integration method for multi-product lines provided by the present invention specifically includes the following steps: S41: Perform inverse kinematic solution processing on the joint angle sequence in the equipment adjustment data, and inversely solve the joint angles corresponding to the end effector pose through the D-H parameter model of the manipulator to generate the six-axis joint angle adjustment amount.

[0074] Specifically, the system performs inverse kinematic solution processing on the joint angle sequence in the equipment adjustment data, and inversely solves the joint angles corresponding to the end effector pose through the D-H parameter model of the manipulator to generate the six-axis joint angle adjustment amount. Specifically, the system first establishes a mathematical relationship between the joint angles and the end effector pose according to the D-H parameter model of the manipulator. This model accurately describes the kinematic characteristics of the manipulator through four basic parameters: link length, link twist angle, link rotation angle, and link displacement. The system converts the target pose of the end effector into the form of a homogeneous transformation matrix, and then uses the inverse kinematic algorithm to solve the corresponding joint angles.

[0075] During the solution process, considering the kinematic constraints and singular configurations of the robotic arm, the system adopts a numerical iteration method to gradually approach the exact solution. The system analyzes the kinematic performance of the robotic arm through the Jacobian matrix to ensure the stability and accuracy of the solution. Finally, the system generates the adjustment amounts of the six-axis joint angles, which accurately reflect the angle changes required for each joint of the robotic arm to reach the target pose. This process is achieved through matrix operations and numerical analysis, ensuring the efficiency and accuracy of the calculation and providing accurate joint angle data for the subsequent pulse signal conversion.

[0076] S42: Perform pulse signal conversion processing on the joint angle adjustment amounts, calculate the driving pulse frequencies of each joint, and generate the pulse control signals for the servo motors.

[0077] Specifically, for the adjustment amount of each joint, the system precisely calculates the number of pulses based on the pulse equivalent of the servo motor. The pulse equivalent is the proportional coefficient between the rotation angle of the motor and the number of pulses. By calculating the number of pulses of the servo motor, the system ensures an accurate correspondence between the number of pulses and the joint angle adjustment amount. Among them, the number of pulses is equal to the joint angle adjustment amount divided by the pulse equivalent. At the same time, according to the preset welding speed and acceleration / deceleration characteristics, the system uses a speed planning algorithm to determine the frequency change law of the pulse sequences of each joint.

[0078] At the start of welding, the system outputs a low-frequency pulse sequence to smoothly start the robotic arm; as the welding process progresses, the pulse frequency gradually increases to reach the set welding speed; at the end of welding or when deceleration is required, the pulse frequency gradually decreases to ensure a smooth transition during the welding process. To ensure the motion synchronization of multiple joints, the system optimizes the timing of the pulse sequences of each joint. The system comprehensively considers the differences in kinematic characteristics of each joint and the requirements of the welding process. By analyzing the motion trajectories and time parameters of each joint, it adjusts the output timing of the pulse sequences. For example, for a joint with a longer motion trajectory, the system will appropriately advance the output time of its pulse sequence to compensate for the motion delay between the components of the robotic arm. Through this precise timing optimization, the coordinated actions of each joint of the robotic arm are ensured, avoiding welding posture deviations or weld quality problems caused by non-synchronization. The finally generated pulse control signals perfectly match the driving requirements of the servo motors, providing a key guarantee for accurately controlling the motion of the robotic arm and ensuring the precise presentation of the welding trajectory.

[0079] S43: Perform bus transmission protocol encapsulation processing on the pulse control signals and send the pulse control signals to the servo driver through the EtherCAT real-time industrial Ethernet.

[0080] Specifically, EtherCAT (Ethernet for Control Automation Technology) is a high-performance real-time industrial Ethernet with the characteristics of high real-time performance and high transmission efficiency, and is particularly suitable for application scenarios with extremely high requirements for time synchronization such as robot welding systems. The system encapsulates the pulse control signal into a data frame format compliant with the EtherCAT protocol. The data frame contains key fields such as source address, destination address, data length, control information, and check bits. During the transmission process, the system utilizes the distributed clock mechanism of the EtherCAT bus to achieve nanosecond-level time synchronization among multiple servo drives, ensuring the simultaneous arrival and execution of the joint pulse control signals. The system also performs integrity verification and error retransmission processing on the transmitted data to ensure the accuracy of the control signal. For example, when a data packet has a check error due to electromagnetic interference or other reasons, the system will immediately retransmit the data packet and monitor and compensate the motion state of the corresponding joint in subsequent control cycles to eliminate the impact of transmission errors on the welding process.

[0081] In addition, the system will also dynamically adjust the size and transmission frequency of the data frame according to network load and real-time requirements. In high-load situations, the system will appropriately reduce the size of the data frame to reduce transmission latency; while in stages with higher real-time requirements, the system will increase the transmission frequency to ensure the timely delivery of control signals. Through the efficient transmission of the EtherCAT bus, the system realizes the real-time and reliable transmission of pulse control signals, ensuring the precise motion control of the robotic arm, and is an important technical support for the efficient and stable operation of the entire robot welding system.

[0082] In one embodiment, as Figure 3 shown, S5 of a robot welding integration method for multiple product lines provided by the present invention specifically includes the following steps: S51: Based on the actual position data fed back by the actuator, perform multimodal sensing and synchronous acquisition processing on the current and molten pool temperature during the welding process. Collect current waveform data through a Hall sensor, obtain the molten pool temperature field distribution matrix through an infrared thermal imager, and generate a process monitoring data set.

[0083] Specifically, the system collects current waveform data through Hall sensors. Based on the Hall effect principle, the Hall sensors convert the changes in welding current into voltage signals. These voltage signals are processed by a signal conditioning circuit for filtering and amplification to remove noise and improve signal quality. Meanwhile, the system uses an infrared thermal imager to obtain the molten pool temperature field distribution matrix. The infrared thermal imager detects the infrared radiation emitted by the object, converts it into an electrical signal, and generates the temperature field distribution matrix after digital processing. The system performs time synchronization processing on the collected current waveform data and temperature field distribution matrix with the actual position data, adding a timestamp to each data point to ensure the synchronization and relevance of the data for precise time alignment in subsequent processing. Finally, the system integrates these data into a structured dataset, namely the process monitoring dataset. This dataset not only contains detailed information on current and temperature but also includes actual position data, providing comprehensive data support for subsequent abnormal fluctuation analysis and quality warning. Through this multi-modal data acquisition and fusion technology, the system achieves comprehensive monitoring of the welding process and ensures real-time control of welding quality.

[0084] S52: Conduct abnormal fluctuation analysis on the process monitoring dataset, calculate the percentage value of the current standard deviation to the mean, and generate a quality warning signal.

[0085] Specifically, the system extracts the current waveform data from the process monitoring dataset and preprocesses it, including removing high-frequency noise using digital filtering technology and eliminating baseline drift through a baseline correction algorithm to ensure the accuracy and stability of the data. To quantify the degree of current fluctuation, the system calculates the standard deviation and mean of the current data. Preferably, the standard deviation and mean calculation formulas are: ; where represents the standard deviation of the current data, represents the i th current data point, represents the total number of data points, represents the mean of the current data. Through the above calculations, the system obtains the standard deviation and mean of the current data and calculates the relative intensity of current fluctuation through the ratio of the standard deviation to the mean. The relative intensity of current fluctuation reflects the proportion of the current fluctuation amplitude relative to the average value and can effectively identify abnormal fluctuations. The system compares this ratio with a preset reference limit, which is determined based on historical data and expert experience and is used to distinguish normal fluctuations from abnormal fluctuations. When the ratio exceeds the reference limit, the system determines that an abnormal fluctuation has occurred during the welding process and generates a quality warning signal. The quality warning signal contains detailed information such as the time, location, and degree of abnormality to prompt the operator to take timely measures.

[0086] Through this statistical analysis method, the system can quickly identify abnormal situations during the welding process and ensure the stability of welding quality. At the same time, the system associates and stores the quality warning signals with the process monitoring data set, providing data support for subsequent quality traceability and process optimization. Through this real-time monitoring and warning mechanism, the system effectively improves the reliability of the welding process and the level of quality control.

[0087] S53: Dynamically correct the quality warning signals based on the molten pool balance equation, adjust the wire feeding speed, and generate dynamic welding parameters including current, voltage, and speed.

[0088] Specifically, considering factors such as the heat input of the molten pool, the material melting rate, and heat conduction, the system establishes a dynamic relationship model between welding parameters according to the molten pool balance equation. The molten pool balance equation is calibrated through physical models and experimental data to ensure its accuracy and applicability. When receiving a quality warning signal, the system adjusts the wire feeding speed in real time through this model to compensate for the impact of current fluctuations on the stability of the molten pool. The adjustment of the wire feeding speed is achieved through a feedback control algorithm to ensure that the change in the wire feeding speed can respond to current fluctuations in a timely manner. Preferably, the adjustment formula for the wire feeding speed is: ; where, is the adjusted wire feeding speed, is the initial wire feeding speed, is the material thermal deformation coefficient, is the difference between the measured temperature of the molten pool and the target temperature, is the reference temperature value.

[0089] At the same time, the system fine-tunes the welding current and voltage. Through the PID control algorithm, it ensures the stable combustion of the arc and the uniform formation of the molten pool. The adjusted welding parameters are updated in real time through a feedback control mechanism to ensure the dynamic adaptability of the welding process. The system integrates the adjusted current, voltage, and wire feeding speed into dynamic welding parameters and generates new control instructions. These instructions are sent to the welding equipment through a real-time communication interface to guide it to perform corresponding parameter adjustments. Through this dynamic correction method based on physical models, the system can effectively cope with abnormal fluctuations during the welding process and ensure the stability of welding quality and production efficiency. At the same time, the system records and analyzes the dynamic welding parameters in real time, providing data support for subsequent process optimization and quality improvement.

[0090] A robot welding integration method for multi-product lines provided in this embodiment can achieve real-time and accurate monitoring of the welding current and molten pool temperature through multi-modal sensing synchronous acquisition and abnormal fluctuation analysis, and can effectively solve the problem of unstable welding quality caused by process parameter fluctuations in traditional welding systems. Based on the data fusion of Hall sensors and infrared thermal imagers, a process monitoring data set characterizing the welding state is constructed to capture potential quality risks; through the calculation of the percentage value of the current standard deviation and the mean, accurate quality warning signals can be generated to identify and respond to process anomalies in a timely manner; relying on the dynamic parameter correction mechanism of the molten pool balance equation, the wire feeding speed can be adaptively adjusted to dynamically optimize the arc energy output to ensure the stability of the molten pool temperature and the quality of the weld formation; this technical system constructs a closed-loop control link from state perception, anomaly detection to parameter correction, and can realize the real-time optimization of welding process parameters under complex working conditions, significantly improving the stability of the welding process and the consistency of product quality.

[0091] In one of the embodiments, step S6 of a robot welding integration method for multi-product lines provided by the present invention specifically includes the following steps: S61: Extract the correction amount characteristics of the dynamic welding parameters, calculate the current adjustment amount and the voltage adjustment amount, and generate a parameter correction feature set.

[0092] Specifically, the system calculates the current adjustment amount in the dynamic welding parameters, and real-time collects the current data during the welding process through a high-precision current sensor. These data are sampled at a high frequency to ensure that the tiny fluctuations of the current can be captured. The system precisely compares the collected real-time current data with the preset welding process parameters to calculate the current deviation value at each moment. Similarly, the system calculates the voltage adjustment amount, obtains the real-time voltage value of the welding arc by using a voltage sensor, and compares it with the target voltage value to obtain the voltage deviation value.

[0093] To improve the accuracy of the data, the system can use advanced digital signal processing technologies to filter and denoise these deviation data. Specifically, the system can apply an adaptive filtering algorithm, such as the Kalman filter, which can dynamically adjust the filtering parameters to effectively remove the high-frequency noise and random interference generated during the welding process. On this basis, the system can perform feature extraction on the processed data through complex mathematical models and algorithms, such as the least squares method and wavelet analysis. Calculate the statistical characteristic parameters such as the mean, variance, peak value, and kurtosis of the current and voltage adjustment amounts. These parameters can comprehensively reflect the stability of the current and voltage during the welding process, as well as the frequency and amplitude changes of the adjustment. The finally generated parameter correction feature set contains these key feature parameters, providing detailed data support for the subsequent weight optimization processing of the dynamic parameter adaptation algorithm, ensuring that the optimization of the welding process parameters can be based on accurate feature data.

[0094] S62: Optimize the weight of the dynamic parameter adaptation algorithm in the process knowledge base based on the parameter-corrected feature set, and update the feature matching weight coefficient of the dynamic parameter adaptation algorithm according to the welding quality qualification rate in the historical production data.

[0095] Specifically, according to the welding quality qualification rate in the historical production data, the system updates the feature matching weight coefficient of the dynamic parameter adaptation algorithm. The historical production data contains detailed information on a large number of past welding tasks, such as welding parameters, material properties, workpiece geometric features, and the final welding quality inspection results. Preferably, the system can use data mining techniques and machine learning algorithms, such as random forest and support vector machine, to deeply analyze the historical data, find out the key feature parameters related to the welding quality qualification rate and their combined relationships. By calculating the contribution degrees of these feature parameters under different welding conditions, the system adjusts the weight coefficient to make the algorithm pay more attention to the features that have a significant impact on the welding quality. For example, if it is found that the average value of the current adjustment amount is highly correlated with the welding quality qualification rate, the weight of this feature in the algorithm will be increased. The optimized dynamic parameter adaptation algorithm can more accurately match the welding process parameters with the workpiece features, so as to quickly generate the optimal welding parameter combination in the face of different working conditions, improving the welding quality and production efficiency.

[0096] S63: Optimize the weight of the path planning algorithm in the process knowledge base based on the path tracking deviation amount in the dynamic welding parameters, and update the trajectory smoothness weight coefficient of the path planning algorithm according to the path execution accuracy index in the historical production data.

[0097] Specifically, the path tracking deviation amount is calculated by comparing the actual motion trajectory of the robotic arm with the planned trajectory. The system uses high-precision position sensors to monitor the motion trajectory of the robotic arm in real time, and compares it with the theoretical trajectory generated by the path planning algorithm to calculate the position deviation and attitude deviation at each moment. These deviation data are processed through coordinate transformation and data fusion to be converted into a unified deviation vector. At the same time, the system comprehensively evaluates the historical production data, which records information such as the motion trajectory of the robotic arm, welding speed, and weld quality in past welding tasks, as well as the specific values of the path execution accuracy. Through data mining techniques, such as association rule mining and clustering analysis, the system finds out the key factors affecting the path execution accuracy, such as the fluctuation of the welding speed, the installation error of the workpiece, and the dynamic characteristics of the robotic arm. According to the correlation degree between these factors and the trajectory smoothness, the system adjusts the weight coefficient in the path planning algorithm. For example, if the historical data shows that the trajectory smoothness has a significant impact on the weld formation quality, the corresponding weight will be increased to reduce the path tracking deviation.

[0098] Preferably, the system can adopt optimization algorithms, such as genetic algorithms and particle swarm optimization algorithms, etc., to optimize and adjust the parameters of the path planning algorithm, ensuring that the algorithm can automatically generate the optimal welding path according to the actual production requirements and equipment conditions. The optimized path planning algorithm can comprehensively consider various factors such as welding quality, path smoothness, and production efficiency, generate a smoother, continuous, and accurate motion trajectory, improve welding efficiency and quality, reduce production costs, and enhance the competitiveness of the enterprise in the market.

[0099] Preferably, as Figure 4 shown, the present invention provides a robot welding integration device 600 for multi-product lines, and this device is configured with the following modules: A data acquisition and preprocessing module 610, which is used to collect and extract data on the three-dimensional geometric features and material properties of multi-variety workpieces, call the dynamic parameter adaptation algorithm in the process knowledge base, and generate product specification data including weld geometric parameters and material mechanical properties; A configuration parameter matching module 620, which is used to perform similarity matching processing on the geometric features of the workpieces in the product specification data with the preset templates in the order demand database, and generate initial configuration data including fixture configuration schemes and initial welding current and voltage parameters; An equipment parameter adjustment module 630, which is used to perform a health value evaluation process on the real-time operation status data of the acquired production equipment, generate equipment evaluation results, and based on the equipment evaluation results and the initial configuration data, call the path planning algorithm in the process knowledge base to calculate the robotic arm motion trajectory, and generate equipment adjustment data including the robotic arm joint angle adjustment sequence and fixture displacement trajectory; An equipment pulse control module 640, which is used to perform motion control instruction conversion processing on the equipment adjustment data, parse the joint angle sequence and displacement trajectory to generate pulse control signals, and send the pulse control signals to the actuator. The pulse control signals are used to control the actuator to drive the robotic arm and fixture to move along the planned path; A welding dynamic correction module 650, which is used to perform real-time monitoring and dynamic correction processing on the current fluctuations and molten pool temperature distribution during the welding process based on the actual position data fed back by the actuator, generate dynamic welding parameters adapted to the working condition changes, and send the dynamic welding parameters to the welding power supply. The dynamic welding parameters are used to adjust the arc energy output; A process knowledge base update module 660, which is used to optimize the weight parameters of the dynamic parameter adaptation algorithm and the path planning algorithm in the process knowledge base according to the dynamic welding parameters and historical production data, and generate updated configuration mapping relationship data.

[0100] In summary, the robot welding integration device for multi-product lines provided by the present invention can achieve rapid process adaptation and dynamic precise control in the scenario of mixed-line production of multiple varieties through the collaboration of multi-source data fusion and intelligent algorithms, and can effectively solve the core problems of insufficient flexibility, response lag, and low intelligent level of traditional welding systems. Through the automatic acquisition and extraction of three-dimensional geometric features and material properties, combined with the dynamic parameter adaptation algorithm of the process knowledge base, a product specification data system that accurately characterizes the workpiece characteristics can be constructed to achieve the accurate mapping of process requirements and equipment capabilities; based on the intelligent matching mechanism of the order demand template, a fixture configuration plan and initial welding parameters suitable for different workpieces can be quickly generated, significantly shortening the process preparation cycle during product specification switching; through the dynamic invocation of real-time equipment health assessment and path planning algorithms, a robotic arm motion trajectory that incorporates equipment performance decay compensation can be generated under complex working conditions to ensure the stability and positioning accuracy of the processing process; relying on the parsing, conversion, and coordination of pulse control signals, high-dynamic-response servo drive control can be achieved to ensure the accurate reproduction of complex trajectories; based on the real-time monitoring of multi-physical field data during the welding process and the dynamic parameter correction, a closed-loop feedback mechanism for the molten pool state and energy output can be established to suppress process fluctuations and improve the forming quality; through the iterative optimization of algorithm weights driven by historical production data, the intelligent decision-making ability of the process knowledge base can be continuously improved, forming a self-evolving production control system.

[0101] The robot welding integration device for multi-product lines provided by the present invention constructs a closed-loop control link from data perception, intelligent decision-making to precise execution, and can achieve rapid switching and adaptation of multiple varieties of workpieces, adaptive compensation of equipment performance, and continuous optimization of process parameters without manual intervention, providing a systematic solution for flexible welding production.

[0102] Preferably, the data acquisition and preprocessing module 610 provided in this embodiment is configured with the following units: An original point cloud generation unit for performing three-dimensional line laser scanning on the surface three-dimensional geometric features of the workpiece to be processed by a three-dimensional line laser scanner to generate original point cloud data containing millions of spatial coordinate points; A noise reduction data generation unit for performing outlier removal processing on the original point cloud data based on the normal vector consistency detection algorithm, calculating the angle between the normal vector of each point and the neighborhood average normal vector, and generating noise-reduced geometric point cloud data; A geometric parameter set generation unit for performing weld seam trajectory extraction processing on the geometric point cloud data, calculating the inclination angle of the weld seam cross-section normal vector and the cumulative value of the Euclidean distance between adjacent points through the principal component analysis method, and generating a geometric parameter set including the weld seam length, inclination angle, and gap size; A product specification generation unit is used to perform material property association processing on a set of geometric parameters, detect the yield strength and thermal conductivity of the workpiece material through laser-induced breakdown spectroscopy technology, fuse geometric features and physical property indexes, and generate product specification data including weld geometric parameters and material mechanical properties.

[0103] Preferably, the configuration parameter matching module 620 provided in this embodiment is configured with the following units: A feature vector generation unit is used to perform multi-dimensional feature vector construction processing on the product specification data, extract the weld length, inclination angle, and material yield strength as feature dimensions, and generate a standardized feature vector; A similarity score unit is used to perform similarity matching processing on the standardized feature vector and a preset template in the order demand database, calculate the cosine value of the angle between vectors based on the cosine similarity algorithm, and generate a similarity score matrix; An initial configuration generation unit is used to perform threshold screening processing on the similarity score matrix, select template items with a score value greater than the preset threshold, and generate initial configuration data including the electromagnetic locking scheme of the fixture and the initial welding current and voltage parameters.

[0104] Preferably, the equipment parameter adjustment module 630 provided in this embodiment is configured with the following units: A real-time status data generation unit is used to perform multi-source sensing data synchronous acquisition processing on the servo motor torque, cylinder pressure, and guide rail wear amount of the production equipment, and generate a device real-time status data set including time stamps; A device evaluation result generation unit is used to perform health quantification processing on the device real-time status data set based on the weighted summation formula, calculate the device health index, and generate a device evaluation result for evaluating the device health status; A path node sequence generation unit is used to perform three-dimensional path planning processing on the device health index, call the path planning algorithm in the process knowledge base to search for a collision-free motion trajectory in the voxel grid space, and generate an initial path node sequence; A device adjustment data generation unit is used to perform health compensation processing on the initial path node sequence, calculate the joint angle adjustment amount of the robotic arm, and generate device adjustment data including the robotic arm joint angle adjustment sequence and the fixture displacement trajectory.

[0105] Preferably, the device pulse control module 640 provided in this embodiment is configured with the following units: A mechanical axis joint adjustment unit is used to perform inverse kinematics processing on the joint angle sequence in the device adjustment data, and inversely solve the joint angles corresponding to the pose of the end effector through the robotic arm D-H parameter model to generate a six-axis joint angle adjustment amount; A pulse control signal generation unit, which is used to perform pulse signal conversion processing on the joint angle adjustment amount, calculate the pulse frequencies of each joint drive, and generate the pulse control signal of the servo motor; A pulse control signal transmission unit, which is used to perform bus transmission protocol encapsulation processing on the pulse control signal, and send the pulse control signal to the servo driver through the EtherCAT real-time industrial Ethernet.

[0106] Preferably, the welding dynamic correction module 650 provided in this embodiment is configured with the following units: A process monitoring data generation unit, which is used to perform multi-modal sensing synchronous acquisition processing on the current and molten pool temperature during the welding process based on the actual position data fed back by the actuator, collect current waveform data through a Hall sensor, obtain the molten pool temperature field distribution matrix through an infrared thermal imager, and generate a process monitoring data set; A quality warning unit, which is used to perform abnormal fluctuation analysis on the process monitoring data set, calculate the percentage value of the current standard deviation to the mean value, and generate a quality warning signal; A welding dynamic detection unit, which is used to perform dynamic parameter correction processing on the quality warning signal based on the molten pool balance equation, adjust the wire feeding speed, and generate dynamic welding parameters including current, voltage and speed.

[0107] Preferably, the process knowledge base update module 660 provided in this embodiment is configured with the following units: A parameter correction feature generation unit, which is used to perform correction amount feature extraction processing on the dynamic welding parameters, calculate the current adjustment amount and voltage adjustment amount, and generate a parameter correction feature set; A dynamic parameter optimization unit, which is used to perform weight optimization processing on the dynamic parameter adaptation algorithm in the process knowledge base based on the parameter correction feature set, and update the feature matching weight coefficient of the dynamic parameter adaptation algorithm according to the welding quality qualification rate in the historical production data; A path planning optimization unit, which is used to perform weight optimization processing on the path planning algorithm in the process knowledge base based on the path tracking deviation amount in the dynamic welding parameters, and update the trajectory smoothness weight coefficient of the path planning algorithm according to the path execution accuracy index in the historical production data.

[0108] In one embodiment, the present application further provides a computer device, including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the above-mentioned robot welding integration method for multiple product lines is implemented.

[0109] In one embodiment, the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the above-mentioned robot welding integration method for multiple product lines is implemented.

[0110] In the description of this specification, the description referring to terms such as "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials, or characteristics described may be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0111] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the present disclosure solution. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0112] As mentioned above, the above are only the specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of various changes or substitutions, and these should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A robot welding integration method for multiple product lines, characterized in that It includes the following steps: S1: Collect and extract data on the three-dimensional geometric features and material properties of multi-variety workpieces, and call the dynamic parameter adaptation algorithm in the process knowledge base to generate product specification data including weld geometric parameters and material mechanical properties; S2: Perform a similarity matching process on the geometric features of the workpieces in the product specification data and the preset templates in the order demand database to generate initial configuration data including the fixture configuration scheme and initial welding current and voltage parameters; S3: Perform a health value evaluation process on the real-time operating status data of the acquired production equipment to generate an equipment evaluation result, and based on the equipment evaluation result and the initial configuration data, call the path planning algorithm in the process knowledge base to calculate the robotic arm movement trajectory, generating equipment adjustment data including the robotic arm joint angle adjustment sequence and fixture displacement trajectory; S4: Perform a motion control instruction conversion process on the equipment adjustment data, parse the joint angle sequence and displacement trajectory to generate pulse control signals, and send the pulse control signals to the actuator. The pulse control signals are used to control the actuator to drive the robotic arm and fixture to move along the planned path; S5: Based on the actual position data feedback by the actuator, perform real-time monitoring and dynamic correction on the current fluctuation and molten pool temperature distribution during the welding process to generate dynamic welding parameters adapted to the working condition changes, and send the dynamic welding parameters to the welding power source. The dynamic welding parameters are used to adjust the arc energy output; S6: Optimize the weight parameters of the dynamic parameter adaptation algorithm and path planning algorithm in the process knowledge base according to the dynamic welding parameters and historical production data to generate updated configuration mapping relationship data.

2. The method according to claim 1, characterized in that The S1 includes: S11: Perform three-dimensional line laser scanning on the surface three-dimensional geometric features of the workpiece to be measured and processed by a three-dimensional line laser scanner to generate raw point cloud data containing millions of spatial coordinate points; S12: Perform outlier removal on the raw point cloud data based on the normal vector consistency detection algorithm, calculate the angle between the normal vector of each point and the neighborhood average normal vector, and generate noise-reduced geometric point cloud data; S13: Perform weld trajectory extraction on the geometric point cloud data, calculate the weld cross-section normal vector inclination angle and the cumulative Euclidean distance of adjacent points through the principal component analysis method, and generate a geometric parameter set including weld length, inclination angle and gap size; S14: Perform material property association on the geometric parameter set, detect the yield strength and thermal conductivity of the workpiece material by laser-induced breakdown spectroscopy technology, and fuse geometric features and physical property indexes to generate product specification data including weld geometric parameters and material mechanical properties.

3. The method according to claim 1, characterized in that The S2 includes: S21: Perform multi-dimensional feature vector construction on the product specification data, extract weld length, inclination angle and material yield strength as feature dimensions to generate a standardized feature vector; S22: Perform a similarity matching process on the standardized feature vector and the preset templates in the order demand database, calculate the cosine value of the angle between vectors based on the cosine similarity algorithm, and generate a similarity score matrix; S23: Perform threshold screening on the similarity scoring matrix, select the template items with scoring values greater than the preset threshold, and generate initial configuration data including the fixture electromagnetic locking solution and the initial welding current and voltage parameters.

4. The method according to claim 1, wherein The S3 includes: S31: Synchronously collect and process multi-source sensing data of the servo motor torque, cylinder pressure, and guide rail wear of the production equipment, and generate a device real-time state data set including timestamps; S32: Quantify the health of the device real-time state data set based on the weighted summation formula, calculate the device health index, and generate a device evaluation result for evaluating the device health state; S33: Perform three-dimensional path planning on the device health index, call the path planning algorithm in the process knowledge base to search for a collision-free motion trajectory in the voxel grid space, and generate an initial path node sequence; S34: Perform health compensation on the initial path node sequence, calculate the robotic arm joint angle adjustment amount, and generate device adjustment data including the robotic arm joint angle adjustment sequence and the fixture displacement trajectory. The calculation formula for the robotic arm joint angle adjustment amount is: ； Among them, is the adjustment amount of the robotic arm joint angle, indicating the i target angle of the joint, is the inverse kinematics function, is the target pose of the end effector, where represents the position of the end effector in three-dimensional space, is the tolerance compensation coefficient, is the maximum value of the health index 1, , is the device health index.

5. The method according to claim 1, wherein The S4 includes: S41: Perform inverse kinematics on the joint angle sequence in the device adjustment data, and inversely solve the joint angles corresponding to the end effector pose through the robotic arm D-H parameter model to generate a six-axis joint angle adjustment amount; S42: Perform pulse signal conversion on the joint angle adjustment amount, calculate the driving pulse frequency of each joint, and generate a pulse control signal for the servo motor; S43: Perform bus transmission protocol encapsulation on the pulse control signal, and send the pulse control signal to the servo driver through EtherCAT real-time industrial Ethernet.

6. The method according to claim 1, wherein The S5 includes: S51: Synchronously collect and process the current and molten pool temperature during the welding process based on the actual position data feedback by the actuator. Collect the current waveform data through a Hall sensor, obtain the molten pool temperature field distribution matrix through an infrared thermal imager, and generate a process monitoring data set; S52: Analyze the abnormal fluctuations in the process monitoring data set, calculate the percentage value of the current standard deviation to the mean, and generate a quality warning signal; S53: Perform dynamic parameter correction on the quality warning signal based on the molten pool balance equation, adjust the wire feeding speed, and generate dynamic welding parameters including current, voltage, and speed. The adjustment formula for the wire feeding speed is: ； Among them, is the adjusted wire feeding speed, is the initial wire feeding speed, is the coefficient of thermal deformation of the material, is the difference between the measured temperature of the molten pool and the target temperature, is the reference temperature value.

7. The method according to any one of claims 1-6, characterized in that, The S6 includes: S61: Extract the correction amount features of the dynamic welding parameters, calculate the current adjustment amount and voltage adjustment amount, and generate a parameter correction feature set; S62: Optimize the weights of the dynamic parameter adaptation algorithm in the process knowledge base based on the parameter correction feature set, and update the feature matching weight coefficients of the dynamic parameter adaptation algorithm according to the welding quality qualification rate in the historical production data; S63: Optimize the weights of the path planning algorithm in the process knowledge base based on the path tracking deviation amount in the dynamic welding parameters, and update the trajectory smoothness weight coefficients of the path planning algorithm according to the path execution accuracy index in the historical production data.

8. A robot welding integration device for multiple product lines, characterized in that The device includes: The data acquisition and preprocessing module is used to collect and extract data on the three-dimensional geometric features and material properties of multi-variety workpieces, call the dynamic parameter adaptation algorithm in the process knowledge base, and generate product specification data including weld geometric parameters and material mechanical properties; The configuration parameter matching module is used to perform similarity matching processing on the geometric features of the workpiece in the product specification data with the preset templates in the order demand database, and generate initial configuration data including the fixture configuration scheme and the initial welding current and voltage parameters; The equipment parameter adjustment module is used to perform health value evaluation processing on the real-time operation status data of the acquired production equipment, generate equipment evaluation results, and based on the equipment evaluation results and the initial configuration data, call the path planning algorithm in the process knowledge base to calculate the robotic arm movement trajectory, and generate equipment adjustment data including the robotic arm joint angle adjustment sequence and the fixture displacement trajectory; The equipment pulse control module is used to perform motion control instruction conversion processing on the equipment adjustment data, parse the joint angle sequence and the displacement trajectory to generate pulse control signals, and send the pulse control signals to the actuator, and the pulse control signals are used to control the actuator to drive the robotic arm and the fixture to move along the planned path; The welding dynamic correction module is used to perform real-time monitoring and dynamic correction processing on the current fluctuation and molten pool temperature distribution during the welding process based on the actual position data fed back by the actuator, generate dynamic welding parameters adapted to the working condition changes, and send the dynamic welding parameters to the welding power source, and the dynamic welding parameters are used to adjust the arc energy output; The process knowledge base update module is used to optimize the weight parameters of the dynamic parameter adaptation algorithm and the path planning algorithm in the process knowledge base according to the dynamic welding parameters and the historical production data, and generate updated configuration mapping relationship data.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, the method according to any one of claims 1 to 7 is implemented.

Images