Intelligent Guided Programming Technology Of Welding Robot For 3D Space Curve Welds

Traditional welding robot programming methods are divided into manual teaching and off-line programming.Manual teaching method needs to occupy the robot for long-term discontinuation teaching,low efficiency,the effect depends largely on the experience and level of the operator,unable to meet the requirements of flexible production.Off-line programming method does not require the robot to stop production,and can achieve coordinate control of multiple robots and peripheral auxiliary equipment,but it requires the precise 3D model of the workpiece.When the deviation of the 3D model of the workpiece is large,the adaptability of the program to the actual work can not be solved,especially for the welding of small batch,multi-variety and non-standard workpieces.In this paper,a line-structured light sensor is used as the main sensing tool,and a set of welding robot intelligent guidance programming system for 3D space curve welds is constructed,which realizes the perception of weldment,the extraction of welds,and the optimal generation of weld trajectory.The specific research contents are as follows:Aiming at the design problem of the intelligent guiding programming system of welding robot,guided by the modular idea,the coordinated control of each module is realized through the information interaction between the modules to realize the welding operation demand.Firstly,the requirements of the programming system were carried out through extensive research and actual welding requirements,and the functional classification and overall architecture design of the system were carried out.Secondly,the module division of the system is determined,and the weldment is sensed by the front-end sensing module,and the weld seam is extracted.Based on this,the problem of optimization of the weld trajectory is solved by the welding process package module,and the universal programming platform module is used to solve the problem of optimization of model reading,unit layout,job simulation and scanning path extraction.Then the robot Eye-in-hand hand-eye system is designed to scan the weldment.Finally,the sensor selection and the calibration of the robot Eye-in-hand hand-eye system are carried out.Aiming at the extraction problem of common type 3D space curve weld seams,a weld seam extraction method based on line-structured light sensor is proposed.Firstly,according to the manual teaching or extracting the scan path from the weldment 3D model,the weldment sensing is performed to realize the acquisition of the ordered point cloud data on the surface of the workpiece.Then,through the pre-processing techniques such as passthrough filtering and Gaussian filtering,the ROI is extracted and the outliers are removed.Finally,according to the geometric characteristics of different types of weldments,the weld points are extracted from each row in the ordered point cloud by RANSAC fitting,point cloud projection,two-line intersection,intersection angle between normal vectors and so on.These weld seam points are combined to form a weld seam point sequence.Different from the conventional method on the base of image processing,the welds is extracted directly by the method of high precision and high-resolution point cloud data processing.The extraction result of this method is less affected by the welding environment,which can solve the problem that the sensing precision of complex 3D space curve weld is not high.Aiming at the problem of optimal generation of weld trajectory,a method for optimizing weld trajectory based on NURNS curve is proposed.Firstly,the weld seam point sequence is sampled to obtain the control point of the NURBS curve.Then the optimization of the weld trajectory is realized based on the NURBS curve method.Finally,the NURBS curve parameter densification based on newton iterative method is mainly studied.The method used in this paper can form weld trajectories with good motion stability and smoothness,thus improving the welding quality and precision,and ensuring the adaptability and flexibility of the welding process.Finally,based on the ROBOLP,an existing general robot off-line programming and simulation platform of the research group,the software development of intelligent guidance programming system of welding robot is completed in the paper.The software development work in this paper mainly includes the development of robot control module and sensor control module,the programming of seam extraction algorithm for different types of seam,the optimization of seam trajectory generation based on NURBS curve,the “one-key operation” of related process,the design and development of user interface to enhance the interaction of the system.In addition,a lot of experiments,using the system software as the experiment platform,were carried out on the proposed method and technology to verify its feasibility and effectiveness.