From Nonlinear Dynamics to Neurorehabilitation: Translational Modeling of a Vibro-Impact Actuator for Ankle Exoskeletons

Background: Robotic ankle exoskeletons typically provide continuous torque assistance, yet impulse-based mechanical stimulation may enhance proprioceptive feedback and push-off training. Vibro-impact actuation offers a compact mechanism capable of generating controlled force pulses, but its nonlinear dynamics and sensitivity to interface conditions remain insufficiently understood for clinical deployment. A single-degree-of-freedom model of a vibro-impact actuator integrated into an ankle rehabilitation device was developed using Lagrange’s formulation under ideal excitation. Coulomb, viscous, and Coulomb–Stribeck friction laws were implemented to represent different orthotic interfaces. Numerical simulations evaluated amplitude–frequency responses, time histories, phase portraits, and basins of attraction across gait-relevant frequencies. Three dynamic regimes—non-impact, impact, and multistability—were identified. Friction characteristics significantly shifted regime boundaries, while stable impact operation produced repeatable impulses within therapeutic ranges. Multistable regions indicated sensitivity to initial conditions, highlighting the need for controlled startup. Vibro-impact actuation is a promising strategy for robotic ankle rehabilitation, providing tunable impulsive assistance with clear implications for actuator design, interface selection, and control robustness.