Research On Robot Calibration Method Based On Multiple Sensor Information Fusion Research

Digital manufacturing and personalized,customized production concept is gradually penetrate into the manufacturing industry.Intelligent manufacturing is considered to be the 21 st century advanced manufacturing technology development important direction.Industrial robot as intelligent manufacturing equipment,the most active elements,its scope and influence will be any other production tools are unmatched.However,absolute positioning accuracy of industrial robots is far from the design accuracy due to the influence of factors such as manufacturing error,assembly error and wear in the process of daily use.Through the calibration its absolute positioning accuracy can be effectively improved.Therefore,the robot calibration method was studied,given a multi-sensor information fusion technology based on the robot calibration method.Working as follows:1)Depth study of the robot modeling method.Based on the POE formula,the kinematics model of the KR5 arc robot is established.The model does not have the singularity problem,and the unified description of the rotating joint and the moving joint is realized.2)Multiple sensor information fusion technology is researched.In order to solve the problem that the sampling time and sampling frequency of each sensor are different,the interpolation extrapolation time registration algorithm is used to unify the measured data to the same time node.For each sensor,the measurement data space criterion is inconsistent.The position data registration experiment and the attitude data registration experiment are designed respectively.The adaptive weighting fusion algorithm is used to fuse the measured data of time and space registration,and the measurement results of the end effector are obtained.3)A parameter calibration model based on position information and joint constraint is established.This model avoids the orthogonalization process of the traditional peg model in the process of identifying parameter iterative,not only simplifies the calculation process,but also improves the precision of calibration result.No coordinate transformation between the survey coordinate system and base coordinate system,reduces the requirements of the survey coordinate system and the error of the process of coordinate transformation.Thereby the error compensation of the robot is more convenient and more accurate.The process of measuring attitude information of flange coordinate system is complex,so the use of the peg model can not only simplify the experimental procedure,but also decreases the requirement of measuring equipment.4)In order to verify the effectiveness of the calibration method,the KR5 arc robot was used to perform the calibration experiment.After the calibration,the position error and attitude error of the robot are obviously reduced.