Font Size: a A A

Research On Seam Tracking System Based On Laser Measurement And Visual Information Extraction

Posted on:2022-05-03Degree:MasterType:Thesis
Country:ChinaCandidate:S H TuoFull Text:PDF
GTID:2480306569953889Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
Robot welding technology has the characteristics of stable quality and high efficiency,which has been widely used in advanced manufacturing industry.At present,welding robot still adopts the traditional "teaching reappearance" mode in many applications,which lacks the function of real-time perception and feedback adjustment of external information.The welding torch may deviate from the original teaching track and affect the welding accuracy and quality after welding when the welding object or condition changes,the surface defects of the weldment and the weldment is heated and deformed.Therefore,this paper studies the seam tracking system based on laser sensor and visual information fusion.With the help of the sensor,the profile data of the welding groove is obtained,and the three-dimensional coordinates of the center point of the welding torch are obtained through industrial computer processing.The curve fitting is used to plan the movement path of the center point of torch,and the information is transmitted to the robot control cabinet.The welding torch is adjusted in real time by driving the robot to move so that it is aligned with the groove center to ensure the welding accuracy and quality.The research contents include:(1)The overall design of the welding seam tracking system is carried out,and the select of each device is completed according to the actual welding application requirements.The welding simulation workstation is established through robot studio,and the layout of the components in the system is planned.According to the planning results,the welding seam tracking experimental platform is built,which is mainly composed of welding robot system,laser sensor and industrial computer.(2)The D-H method is used to establish the kinematics model of the welding robot,and the forward and inverse kinematics analysis is carried out to obtain the transformation matrix between the head and end coordinate systems of the robot,which provides the calculation basis for calibrating the welding gun and laser sensor.The calibration method of welding gun and laser sensor is studied.The principle of multi-point calibration of laser sensor is stated.On this basis,the three-point calibration method is proposed.Through the pose calibration of welding gun and laser sensor,the transformation relationship between the two coordinate systems is obtained,which lays the foundation for the collection and processing of weld data.(3)Research on seam tracking method is carried out to realize high-precision positioning of groove feature points of weldment.Aiming at I,V and Y-groove,the weld seam tracking is realized by four steps: welding profile scanning,profile data filtering,groove feature point positioning and welding path planning.For the groove feature points,the second derivative maximum method is used for preliminary positioning,and then the linear fitting is used for accurate positioning;for the welding path planning,the least square,Gaussian and cubic splin interpolation are used to fit the movement path of the welding torch center point,which gives a theoretical basis for the experimental verification of weld tracking.(4)The calibration experiment of welding gun and laser sensor is completed,and track verification of straight and curve weld is carried out.The experimental results show that the tracking error is about 0.628 mm and 0.736 mm in the initial positioning,and the error reduce to 0.387 mm and 0.429 mm after the precise positioning,and the lifting range reaches 38.4%and 41.7%,which meets the precision requirements of the industrial automatic welding error? 0.5mm.The results show that the proposed tracking method is effective and can provide powerful technical support for the welding of complex welds such as high precision weld tracking and curve surface.
Keywords/Search Tags:Welding robot, Laser sensor, Feature point extraction, Linear fitting, Seam tracking
PDF Full Text Request
Related items