Research On Discontinuous Seam Tracking Technology For Mobile Robot

In the process of ship manufacturing,a lattice type member is arranged at the bottom of the cabin.This structure is narrow space and have irregular drain hole,so welding seam is discontinuous and the welding robot arm can not reach.At present,the welding is mainly completed by manual.Research and develop autonomous mobile robot for complex cabin,the robot is used to welding drain hole in lattice roomand free the workers from the harsh welding environment,improve the production efficiency and welding quality at the same time,has important theoretical and practical significance.Take the autonomous mobile welding robot as platform to solve welding automation of lattice room in the manufacture of cabin.Rotating arc and laser vision are used for sensing,types of welds are involved:Straight seam,bend seam,drain hole welding seam,right angle turn seam,right angle turn seam with drain hole.The main research contents are as follows: the design of laser vision sensor,the treatment and recognition of welding seam with drain hole,the planning and real-time tracking of right angle seam and the initial localization of welding seam and so on.In the light of the characteristics that a single sensor can not recognize the drain hole welding seam,the system adopts rotating arc sensing and laser vision dual sensing mode.The rotating arc is used for seam tracking under welding condition.The laser vision sensor is used to weld seam tracking in arc stop condition,the characteristic points and mold identification of drain hole welding seam.According to the drain hole welding seam and actual welding characteristics,an active laser vision sensor was designed.The design principle,process and encapsulation of the sensor are introduced in detail.The sensor can be applied not only to the fillet weld,but also be used for other types seam recognition,which provides reference for the design of vision sensor in welding field.Aiming at the problem of welding current signals are susceptible to all kinds of interference,a combination filtering method including preprocessing,amplitude limiting,mean value and soft ware are put forward.The method can effectively suppress noise and improve signal to noise ratio.The horizontal deviation is extracted by characteristic harmonic method,the deviation is transmitted to the Fuzzy-PID controller,and the system control quantity is obtained.Through the movement of the slider,the correction of the weld seam is corrected.The feasibility and accuracy of the algorithm are verified by experiments.In the welding process,arc,spatter,smoke and so on drag-in great interference to the quality of weld image,it is very important to choose the image processing methods which have strong anti-interference ability.An image processing method based on improved gradient algorithm is proposed,it can extract the original weld information from the strong interference weld image.The treated laser stripe can completely express the characteristics of drain hole welding seam and lays a foundation for accurate seam recognition.According to different welding types of drain holes,the identification methods of weld characteristic points and types are given.In the weld process of lattice room,in order to reducelaying requirement of Robot's initial position,Welding robot should have the capability of automatically finding the initial position of welding seam.The camera system was calibrated,initial positioning analysis for straight line and turn weld was completed,the relationship between the robot,the wheels and the initial position of the slider movement was given,the automatic initial position of welding robot is realized.In order to meet the requirements of automatic tracking welding of various welds in the lattice room,Kinematics and dynamics analysisfor the welding robot system is carried out,and a mathematical model is established;The adaptive optimal weighted information fusion algorithm is adopted to fuse the rotation arc deviation and the laser deviation to realize the smooth welding of the curved weld;Aim at right angled turn welding seam and Right angled turn with drain hole welding seam,path planning and real-time seam tracking method was adopted,analysis and Simulation for movement process was completed,the coordinated control method of robot body and cross slider is given.Finally,the effectiveness of the research work is verified by the actual welding test.In addition,the actual trial of the autonomous mobile welding robot system was carried out in Jiujiang Jiangzhou shipyard in the design of,the desired results is achieved.