부산대학교 도서관

This paper proposes a 6R robot kinematic closed-loop calibration method, in order to improve the 6R robot kinematic absolute positioning accuracy, that establishes a single-point pose constraint through the monocular vision-based pose measurement method, and realizes the robot kinematic parameter identification by using the particle swarm optimization method. The camera installed at the end of the robot is used as the measurement equipment, and the fixed high-precision calibration board shall be measured in multiple positions to achieve single-point restraint in closed-loop calibration of kinematic. In order to realize the application of calibration results in the industrial environment, a minimum parameter set of 7 parameters for kinematic calibration is obtained based on inverse solution of parameter constraints in industrial application and kinematic coupling of hand-eye system, and step-by-step parameter identification is achieved with the aid of hand-eye calibration algorithm to realize fast and stable kinematic parameter identification of robots. The proposed methodology is employed to achieve successful calibration of the HSR-JR605 robot, leading to a reduction in the average distance error of robot motion from 2.536 mm to 0.3422 mm.