Task-Invariant Centroidal Momentum Shaping for Lower-Limb Exoskeletons

Yu, Miao; Lv, Ge

doi:10.1109/cdc51059.2022.9993080

Cited by 2 publications

(2 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Different from control paradigms that do not consider human inputs, the proposed approach requires the information of volitional human motion to compute control laws u d and u. Therefore, we modify an existing model-based NDO [22] to estimate human joint torques using only angular information [30]. Let B λ = B − A T λ, τhum = (I − A T λ)τ hum , and define z = M −1 (τ hum + J T F perturb ) as the term to be estimated, left-multiplying M −1 at both sides of (1), we have…”

Section: Modified Nonlinear Disturbance Observermentioning

confidence: 99%

See 1 more Smart Citation

Safety Augmentation for Volitional Human Locomotion via Lower-Limb Exoskeletons: A Case Study

Yu,

Lin,

2024

Preprint

View full text Add to dashboard Cite

User safety is a crucial factor to consider when designing control paradigms for lower-limb exoskeletons. Existing control paradigms mainly focus on providing assistance for human users under stable walking conditions, ignoring situations that human users may lose their balance due to external perturbations during locomotion. In this letter, we propose a safety framework for lower-limb exoskeletons to augment safety for volitional human motion based on Control Barrier Functions. The safety indicators are defined as the human’s center of mass and swing foot position lying within self-selected ranges. Instead of enforcing reference trajectories, we incorporate human inputs and preferences in a two-layer quadratic program structure based on Control Barrier Functions to generate assistance for ensuring safety. Simulation results on a human wearing an exoskeleton demonstrate that the proposed control paradigm can generate assistance to assist human users in maintaining balance while undergoing gait perturbations and recovering from initial unsafe postures.

show abstract

Section: Modified Nonlinear Disturbance Observermentioning

confidence: 99%

“…and e can be ensured to be uniformly ultimately bounded with selected parameters [30]. When external perturbations occur, ( 16) will estimate it as part of the "human torque".…”

Section: Modified Nonlinear Disturbance Observermentioning

confidence: 99%