Research On Laser Recognition Technology Of Right Angle Weld Seam Of Welding Robot

With the development and maturity of robot technology,automated welding robots are gradually replacing manual welding to solve the problems of low production efficiency and dirty environment in traditional welding.Welding sensing technology is the precondition to realize automatic welding seam tracking.Aiming at the shortcomings of traditional ultrasonic,arc and other sensing technologies such as low precision and poor anti-interference ability,a welding seam identification method based on line structured light is proposed to complete right-angle welding.The welding seam is welded under special working conditions to realize the intelligent production.The main work of this paper is as follows:First of all,this paper determines the overall design scheme of laser weld seam recognition according to actual needs,determines the optical path structure of the sensor according to the optical path analysis of the line structured light vision sensor,and completes the design of each component in the visual weld seam recognition system according to the technical indicators.A hardware platform for visual tracking recognition.Secondly,this paper establishes the model of the structured light vision-guided weld seam recognition system,and studies the mapping relationship between the two-dimensional image coordinates and the three-dimensional coordinate information in the image captured by the sensor.The camera calibration principle,the laser plane calibration principle and the hand-eye calibration principle are introduced in turn,and the transformation relationship between the image coordinate system and the robot coordinate system is finally solved,and the transformation from the image recognition result to the robot execution result is completed.Then,according to the right-angle weld image collected by the vision sensor,an image processing algorithm is designed in this paper.Aiming at the problem that traditional ROI is easily interfered by subjective factors,a dynamic ROI extraction method based on weld target recognition is proposed.According to the industrial environment where the workpiece is located,median filtering is used for image smoothing;Due to the poor interference ability,a method for extracting laser center fringes based on the improved barycenter gray method is proposed.In view of the long time-consuming problem of traditional detection of straight lines,an improved hough transform is proposed to improve the detection speed.Finally,the general idea of right-angle weld seam welding is designed in this paper,and the designed visual weld seam recognition system is verified by offline and online methods.The experimental results show that the designed image processing algorithm has the advantages of high precision and good real-time performance.;The designed calibration model is accurate,and the tracking welding can be completed stably.Therefore,the right-angle weld identification system designed in this paper achieves the expected effect,and can be applied to real-time welding in right-angle weld conditions to improve welding efficiency.