Research On Robot Welding Operation Planning Technology Based On Digital Twin

Robots are widely used in the field of automated welding,which improves the stability and efficiency of welding and becomes an important equipment for modern industrial production.With the rapid development of information technology,welding robots are connected and fused with various equipment and sensors,and a series of automatic control technologies have been derived.With the introduction of intelligent manufacturing and Industry 4.0,the concept of digital twins was born and gradually attracted attention by scholars.At present,digital twin technology is mainly used in the fields of aircraft manufacturing,workshop control and numerical control processing,and there is little research in the scope of robot welding.This paper takes the 9-axis welding unit as the research object.First,it analyzes the concept and definition,technical characteristics and application status of digital twins.Combining the needs of welding automation and the existing intelligent control technology,a robot welding digital twin system architecture is proposed,designed and implemented the virtual model layer,core processing layer and physical entity layer respectively.The main research contents of this paper are as follows:Analyzed the mainstream model architecture of the current digital twin technology research,combined with the characteristics of the unit-level twin system,determined to divide the robot welding digital twin system into three layers,namely the virtual model layer,the core processing layer and the physical entity layer,combined with virtual simulation.With offline programming,robot positioner coordinated movement,genetic algorithm optimization,welding process library,welding seam real-time tracking and other technologies,it constitutes a complete digital twin system architecture for robot welding.First study the virtual model layer.Aiming at common welding scenes such as robot ground assembly and ceiling assembly,a simulation environment is built,and the virtual scene of welding is generated and imported based on XML technology,and the geometric model of the physical entity is established.Based on the VTK three-dimensional visualization technology,the high-speed rendering of the welding motion simulation process is realized,and the collision detection of the simulation model is completed,the production data of the welding unit is classified,and the simulation motion data and the real-time motion data of the welding process are based on the QWT tool library.The curve is drawn,and the verification model of the welding movement is established.Aiming at common weld shapes,the coordinated motion strategy of robot and positioner is analyzed,and the advantages and scope of application of the boat-shaped welding algorithm are analyzed.On this basis,an auxiliary motion strategy is proposed to improve the proximity of the robot.The inverse kinematics solution of the positioner under the strategy,the planning model of the welding unit is established.Finally,the triangular closed-loop relationship between the geometric model,the planning model and the verification model is established in the virtual model layer to realize the reproduction of the geometric,behavior and state characteristics of the physical entity.Then study the core processing layer.Aiming at the problems of welding unit energy consumption and motion stability,the placement position of the workpiece positioner is used as the input condition,and the genetic algorithm is used for iterative analysis and optimization based on the triangular closed-loop relationship of the virtual model layer.The welding knowledge base is designed based on SQLite3,case-based reasoning based on kd-tree and fuzzy processing,combined with fuzzy rule reasoning to form a complete reasoning method,based on TFN-AHP to design a welding surface quality evaluation method,and a process library update mechanism is designed to form a common structure The process knowledge model of the twin system.Combining the technical characteristics of real-time welding seam tracking and digital twins,a fusion scheme is proposed.Due to the overall workload arrangement of the subject,real-time welding seam tracking is only a scheme design.Finally,the physical entity layer is studied.The combined feature point method and the three-point correction method are selected as the calibration scheme of the welding unit.Combined with project requirements,Modbus/TCP and TCP/IP are selected as the network communication methods between upper and lower computers.In this paper,through virtual simulation test and actual welding experiment,the functional test of the robot welding digital twin system is carried out,which provides a new idea for the application of digital twin technology in the field of robot welding.