Investigation Into Key Technologies Of A 6-DOF Hybrid Robot For Optical Polishing

Driven by the practical needs for innovation and development of ultra-precision optical polishing equipment,this dissertation investigates key technologies of a novel 6-DOF hybrid robot for optical polishing.The main contents cover trajectory planning for implementation precision of dwell time,inverse displacement and workspace analysis,kinematic calibration and customized CNC system software development of the polishing robot.The contributions made in this dissertation are summarized as follows.An approach for optimizing the theoretical dwell time is proposed by considering its fluctuation limits.Then,an effective algorithm is presented to determine the full set of control points along the piecewised cubic B-spline based polishing path in an efficient and accurate manner.This is followed by the motion profile planning that finally enables the implementation error of the dwell time to be minimized by modifying the feed-rate of polishing head.The simulation results show that discrepancy between the predicted PV(RMS)values and theoretical calculations can be reduced from 10.1%(7.8%)to 0.8%(0.6%)in comparison with conventional linear interpolation method.The layout of a robotized polishing machine especially designed for ultraprecision polishing aspheric optical lens is proposed,which is essentially composed of a 6-DOF hybrid robot and a polishing end-effector.The inverse displacement and reachable workspace analyses of the robot are carried out,resulting in a cylindrical task workspace that can meet the requirement for polishing aspherical optical elements of500 mm?500 mm in dimension.Drawing mainly on screw theory,a complete and linearize error model of the polishing robot is formulated by considering all possible source errors at the joint/link level.Then,a hierarchical two-step procedure for error parameter estimation and pose error compensation is proposed:(1)the coarse estimation and compensation of the encoder offsets using ordinary least squares(OLS);and(2)the fine estimation of the whole set of identifiable error parameters using a Liu estimation(LE)and subsequent modification of the NC trajectory dataset of the polishing head.The simulation results show that the overall standard deviation of the whole set of error parameters estimated by the LE is 63.9%,smaller than that estimated by the OLS.Consequently,the maximum and average volumetric positioning(orienting)error can be reduced by 23.1%(37.5%)and 20.0%(11.1%)using the LE compared with those obtained by the OLS.With the aid of the task decomposition technique,the software architecture of CNC system of the polishing robot is proposed.Then,the component-oriented technology is employed to generate the functional units with high encapsulation and reusability,leading to significant improvements in efficiency and reliability of the software development.Equipped with the knowledge and information aforementioned,a general approach is proposed for the development of the customized CNC software by taking into accout demanding analysis,architecture design,component library construction and human-computer interface generation.The developed software has the capabilities of technological parameters setting,controller parameters tunning,error compensation and trajectory interpolation,thereby fulfilling the requirement for optical polishing.Comprehensive experiments are carried out on a prototype polishing robot.The results show that the positioning(orienting)volumetric accuracy of 0.060 mm(0.05 deg)can be achieved in the task workspace after fine kinematic calibration.The results also show that PV/RMS values of the optical specimen,with five iterations,converge from 5.36 ?m/0.75 ?m to 1.84 ?m/0.20 ?m,confirming the effectiveness of the proposed trajectory planning method.The outcomes have been successfully employed in the development of a 6-DOF hybrid robot prototype for high-quality and high-efficiency optical polishing.