Litcius/Paper detail

An Autonomous Positioning Method of Tube-to-Tubesheet Welding Robot Based on Coordinate Transformation and Template Matching

Junqiong Cai, Ting Lei

2021IEEE Robotics and Automation Letters27 citationsDOI

Abstract

In order to change the positioning method “teach-playback” for tube-to-tubesheet welding by robots, this letter proposes an autonomous positioning method based on coordinate transformation and template matching. The coordinate trans-formation between the robot coordinate system (RCS) and the drawing coordinate system (DCS) has been established. The CCD camera will move to the position of target tube by the coordinate transformation of robot. This is coarse positioning. An improved template matching algorithm is used in machine vision. It can automatically create a template for each image instead of using a template to match the images of all the tubes. The welding torch will accurately locate the target tube by the result of template match. This is fine positioning. Finally, the comparative test showed that the proposed method is superior to the method of Ref [27]. The detection errors in x and y directions are within ±0.18 mm.

Topics & Concepts

WeldingTransformation (genetics)Template matchingMatching (statistics)Tube (container)RobotComputer scienceCoordinate systemRobot weldingComputer visionArtificial intelligenceMechanical engineeringEngineering drawingEngineeringMathematicsImage (mathematics)StatisticsGeneChemistryBiochemistryRobotics and Sensor-Based LocalizationRobot Manipulation and LearningRobotic Mechanisms and Dynamics