On The Dynamic Calibration and Trajectory Control of ARASH:ASiST

This article investigates the dynamic parameter calibration of ARAS Haptic System for EYE Surgery Training (ARASH:ASiST). ARASH:ASiST is a 3-DOF haptic device developed for intraocular surgery training. In this paper, the linear regression form of the dynamic formulation of the system with respect to its dynamic parameters is derived. Then the dynamic parameters of ARASH:ASiST are calibrated using the least square (LS) identification scheme. The cross-validation results for different trajectories indicate that the identified model has a suitable approximate fitness percentage in both translational and rotational motions of the surgical instrument. Finally, a robust model-based controller is implemented on the real prototype by the use of the calibration outcome, and it is verified that by using the estimated dynamic model, the trajectory tracking performance is significantly improved and the tracking error is reduced 50% compared to that of using the nominal dynamic model of the robot.