A Kinematic Parameter Calibration Method of a 6-Axis Industrial Robot Using an Eye-in-Hand 2-D Laser Profiler

This letter presents a novel position estimation method for a 2-D laser profiler (LPF) and its application to the offline kinematic parameter calibration of an industrial robot. Unlike traditional laser tracker systems, LPFs are more affordable, easier to configure, and can capture over 3000 data points in a single scan, which provides valuable characteristics for calibration without introducing new errors owing to motion and time effects. The method relies on a single scan of a custom-designed gauge, with profile features extracted using an edge detection algorithm that combines split-and-merge with linear regression. A gauge frame establishment approach using the LPF is also introduced. The feasibility of the method was validated through offline kinematic parameter calibration experiments on the IRB2600 industrial robot. Three methods were applied to optimize nonlinear error models of the kinematic parameters, including fmincons, particle swarm optimization (PSO), and genetic algorithms. The methodology was evaluated experimentally using a commercial industrial robot, and the results showed significant improvement in positioning accuracy with more than 90$\%$ error reduction by PSO and fmincons, demonstrating the method's effectiveness and applicability in high-precision tasks.