The Study Of Kinematics Parameters And Vision Calibration Of SCARA

SCARA(Selective Compliance Assembly Robot Arm) is widely used inSMT(Surface Mounted Technology) for its flexibility in X and Y directions and goodstiffness in Z-axis. To ensure the stable and accurate work of SCARA, we mustcalibrate its kinematics parameters and vision system before putting it into productionpractice.This dissertation first presents a fast method to calibrate the kinematicsparameters of SCARA(including the two arms’ length determination and thecalibration of the zero position) by using a calibration plate with four holes based onSVD decomposition, we introduce the calibration tool, derive the calibrationequations detailedly, and propose advice to improve calibration accuracy. Second, wepropose a digital image processing algorithm which is used to detect the sub-pixelcenter of single circular marker from complex visual field based on a cubic curvefitting method, analyze the experimental results. This algorithm is successfully usedin the method of camera calibration, and meets the real-time and accuracyrequirements. Third, we introduce the camera pin-hole model and propose asemiautomatic method to calibrate the camera fixed on the assembly plane andanother one to meet the requirement of Hand-Eye calibration, and we finish the erroranalysis. Finally, we describe SCARA’s radial offset of U shaft while picking up thematerial, and propose a method to calibrate the offset, at the same time, come up withthe error compensation control scheme to the problem of the PCB position error andthe material radial offset.The calibration methods proposed in this paper as well as the developed software,has been used as one of the SCARA calibration solutions in Colibri AutomationTechnology (shenzhen) Co., Ltd.