Font Size: a A A

Research On Kinematic Calibration Of A Triaxial SCARA Based On Vision

Posted on:2018-01-15Degree:MasterType:Thesis
Country:ChinaCandidate:H L XiongFull Text:PDF
GTID:2348330542451781Subject:Measuring and Testing Technology and Instruments
Abstract/Summary:PDF Full Text Request
Selective Compliance Assembly Robot Arm(SCARA)is widely demanded,and the absolute accuracy is one of its main performance indexes.The calibration of kinematic parameters for SCARA can improve its absolute accuracy.Calibration is a technique that can recognize robot's parameters accurately by using proper measuring instrument.Based on accurate parameters,the absolute accuracy can be improved by controller tuning or error compensation.Technologically,the study of calibration can facilitate the application of robot off-line programming which can simulates a robot's work.There are four steps in the calibration:modeling,measuring,identification,and compensation.This thesis contributes to measurement and identification.Firstly,a program by combing axial measurement method and telecentric vision measuring system is proposed for the kinematic calibration of a triaxial SCARA.In this program,the vision measuring system is based on telecentric optical path in both object and image space.Then the triaxial SCARA is designed.Limited by the measurement field of view,the mark point should be mounted skillfully for measuring efficiently.The principle of the mark point installation is researched by means of the simulation of SCARA and then motion planning is proposed for optimizing measurement.Secondly,sobel and canny edge detection operators are tested by using the generated standard circle marks.Then three different sub-pixel location algorithmsare researched and practicedwhich are Gaussian fitting method,Gradient barycenter method,and Zernike moment method.Eventually,sobel-Gaussian method is utilized for sub-pixel location in this theis.In the step of identification,a dual-threshold algorithm based on least square method and radius constraint algorithm is proposed for small arc fitting precisely.Otherwise,Region-of-Interest extract is improved effectively.Finally,the calibration of the triaxial SCARA is implemented and verified both by the vision measuring system.The arm's lengths and zero positions are obtained.The results show that the absolute accuracy of the cursory triaxial SCARA is less than 1.5mm and it is suited for most industrial application.
Keywords/Search Tags:SCARA, kinematic calibration, vision measuring, sub-pixel location, small arc fitting
PDF Full Text Request
Related items