In the application of laser line sensors for visual guidance of welding robots,the traditional hand-eye calibration methods often suffer from problems such as cumbersome procedures,significant human errors,difficulty in reuse,and inabil-ity to perform automatic recalibration.To address these issues,this paper proposes a rapid hand-eye calibration method for robots based on standard spheres and a general automated calibration program.The method utilizes a laser line sensor to scan a standard sphere,employs a sphere center fitting algorithm to obtain the sphere center position,and combines it with robot pose information to establish the transformation relationship between the sensor coordinate system and the robot coor-dinate system.The automated calibration program supports various brands of robots,laser line sensors,and communication protocols,and utilizes the PCL point cloud library and MathNET numerical calculation library for computation,thereby achieving automation and versatility in the calibration process.Experimental results show that this method only requires the robot to roughly teach 5 program points,and can automatically scan and obtain calibration results within 2 minutes,with a calibration accuracy better than 0.5 mm,meeting the requirements of robot arc welding applications.Moreover,the program is easy to reuse,simplifying the preliminary debugging and later operation and maintenance work of the robot equipped with a laser line 3D vision system,effectively solving the shortcomings of traditional hand-eye calibration methods,and provid-ing convenience for the application of laser line sensors in robot visual guidance.In the future,the calibration method will be further optimized from aspects such as reducing the sphere center fitting error and optimizing the scanning position and attitude,and its application will be expanded to fields such as high-energy beam precision welding.
关键词
线激光传感器/机器人手眼标定/标准球/自动化程序/弧焊
Key words
laser line sensor/robot hand-eye calibration/standard sphere/automated program/arc welding
维普
万方数据