Skill learning approach based on impedance control for spine surgical training simulators with haptic playback

Abstract

In this article, a robust adaptive impedance control based on the modified function approximation technique, and augmented with a novel integral nonsingular terminal sliding mode control, is proposed for a surgical haptic system. The haptic device is maneuvered by surgeons, whereby the mode of action, motion path, and haptic dynamics are rendered as unknowns. The inclusion of integral nonsingular terminal sliding mode control leverages the capability of ensuring consistent tracking of system’s trajectories performance, fast transient response, finite-time convergence, improved robustness, and reduced chattering. However, the modified function approximation technique strategy allows to estimate model’s dynamics regardless of the availability of dynamic uncertainties’ lower and upper bounds. In addition, an estimate of the intended motion mode and path is integrated in the evolved adaptive impedance control, via radial basis function neural network, permitting the haptic arm to track the target impedance model. Finally, using the haptic arm, controlled experimental cases and comparative study were done, after which the proposed surgical simulator adopted the results for real-time validation.

Adaptive-Robust Controller for Smart Exoskeleton Robot

“…This surface ensures finite-time convergence, a swift transient response, diminished chattering, and the avoidance of singularities. These enhancements address challenges observed in prior SMC strategies [ 23 , 24 ], providing a more-robust and -efficient control foundation. The development of an adaptive controller based on the modified function approximation technique (MFAT) for unknown robot dynamics models.…”

“…However, FAT comes with its drawbacks, such as the requirement for basis functions and sharing limitations with SMC, including poor disturbance response, chattering, and the need to predetermine uncertainty limits. In previous works like [ 23 , 24 ], a modified function approximation technique was explored. Nevertheless, in [ 23 ], the main emphasis was on adaptive impedance control within the haptic system, specifically targeting the force control loop.…”

“…In previous works like [ 23 , 24 ], a modified function approximation technique was explored. Nevertheless, in [ 23 ], the main emphasis was on adaptive impedance control within the haptic system, specifically targeting the force control loop. This approach significantly contributed to estimating the desired motion intention of the surgeon, intending to enhance the overall impedance of the system.…”

“…Terminal sliding mode control (TSMC) achieves this [ 25 ], but has slow convergence and singularities. Faster techniques have been proposed, like faster TSMC [ 26 ], nonsingular TSMC [ 25 ], and integral nonsingular TSMC [ 23 , 24 ], addressing fast convergence. However, singularities persist.…”

“…The contributions of this research, as outlined in the paper, can be summarized as follows: The introduction of a groundbreaking double-integral nonsingular terminal sliding mode surface, overcoming limitations in conventional SMC approaches [ 23 , 24 ]. This surface ensures finite-time convergence, a swift transient response, diminished chattering, and the avoidance of singularities.…”