Human-in-the-Loop Bayesian Optimization of a Tethered Soft Exosuit for Assisting Hip Extension

Kim, Myung‐Hee; Ding, Ye; Liu, Charles; Lee, Sang-Jun; Karavas, Nikolaos; Walsh, Conor J.; Kuindersma, Scott

doi:10.1007/978-3-030-01887-0_28

Cited by 3 publications

(2 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Chen et al [ 7 ] proposed the use of a Rayleigh oscillator to fit gait information, and this method could maintain relatively stable predictions under the condition of large changes in the gait period. Kim et al [ 8 ] proposed a human-in-the-loop Bayesian optimization algorithm that combined two parametric sinusoids at the peak of assistance to apply an assist curve to aid in hip extension. Guo et al [ 9 ] proposed a new type of assist function that can generate the desired assist function only by adjusting three parameters: the assist amplitude, the assist period, and the assist peak shift.…”

Section: Introductionmentioning

confidence: 99%

Optimization of Torque-Control Model for Quasi-Direct-Drive Knee Exoskeleton Robots Based on Regression Forecasting

Xia,

Wei,

Lin

et al. 2024

Sensors

View full text Add to dashboard Cite

The choice of torque curve in lower-limb enhanced exoskeleton robots is a key problem in the control of lower-limb exoskeleton robots. As a human–machine coupled system, mapping from sensor data to joint torque is complex and non-linear, making it difficult to accurately model using mathematical tools. In this research study, the knee torque data of an exoskeleton robot climbing up stairs were obtained using an optical motion-capture system and three-dimensional force-measuring tables, and the inertial measurement unit (IMU) data of the lower limbs of the exoskeleton robot were simultaneously collected. Nonlinear approximations can be learned using machine learning methods. In this research study, a multivariate network model combining CNN and LSTM was used for nonlinear regression forecasting, and a knee joint torque-control model was obtained. Due to delays in mechanical transmission, communication, and the bottom controller, the actual torque curve will lag behind the theoretical curve. In order to compensate for these delays, different time shifts of the torque curve were carried out in the model-training stage to produce different control models. The above model was applied to a lightweight knee exoskeleton robot. The performance of the exoskeleton robot was evaluated using surface electromyography (sEMG) experiments, and the effects of different time-shifting parameters on the performance were compared. During testing, the sEMG activity of the rectus femoris (RF) decreased by 20.87%, while the sEMG activity of the vastus medialis (VM) increased by 17.45%. The experimental results verify the effectiveness of this control model in assisting knee joints in climbing up stairs.

show abstract

Section: Introductionmentioning

confidence: 99%

Optimization of Torque-Control Model for Quasi-Direct-Drive Knee Exoskeleton Robots Based on Regression Forecasting

Xia,

Wei,

Lin

et al. 2024

Sensors

View full text Add to dashboard Cite

show abstract

“…While bioinspiration, simulation models, and user feedback-based hand tuning can provide initial suggestions for control parameters, previous research has consistently demonstrated that HILO outperforms these approaches for easily quantifiable outcomes, such as metabolic cost and walking speed [8,11,26]. Selecting effective control parameters for exoskeletons is a complex task due to the difficulty in modeling or predicting human responses [27], the variability in individual responses to exoskeleton assistance [28,29], and the need for extensive training to use exoskeletons effectively [30]. When developing a new exoskeleton or addressing a new task, such as walking with frontal hip assistance, HILO becomes particularly valuable as it can identify effective control parameters that may not be known in advance.…”

Section: Introductionmentioning

confidence: 99%

Frontal hip exoskeleton assistance does not appear promising for reducing the metabolic cost of walking: A preliminary experimental study

Kim,

Raitor,

Liu

et al. 2023

Preprint

View full text Add to dashboard Cite

Background: During walking, humans exert a substantial hip abduction moment to maintain balance and prevent pelvic drop. This significant torque requirement suggests that assisting the frontal hip muscles could be a promising strategy to reduce the energy expenditure associated with walking. A previous musculoskeletal simulation study also predicted that providing hip abduction assistance through an exoskeleton could potentially result in a large reduction in whole-body metabolic rate. However, to date, no study has experimentally assessed the metabolic cost of walking with frontal hip assistance. Methods: In this case study involving a single subject (N = 1), a tethered hip exoskeleton emulator was used to assess the feasibility of reducing metabolic expenditure through frontal-plane hip assistance. Human-in-the-loop optimization was conducted separately under torque and position control to determine energetically optimal assistance parameters for each control scheme. Results: The optimized profiles in both control schemes did not reduce metabolic rate compared to walking with assistance turned off. The optimal peak torque magnitude was found to be close to zero, suggesting that any hip abduction torque would increase metabolic rate. Both bio-inspired and simulation-inspired profiles substantially increased metabolic cost. Conclusion: Frontal hip assistance does not appear to be promising in reducing the metabolic rate of walking. This could be attributed to the need for maintaining balance, as humans may refrain from relaxing certain muscles as a precaution against unexpected disturbances during walking. An investigation of different control architectures is needed to determine if frontal-plane hip assistance can yield successful results.

show abstract

Lower limb active prosthetic systems—overview

Voloshina

Collins

2020

Wearable Robotics

View full text Add to dashboard Cite

Human-in-the-Loop Bayesian Optimization of a Tethered Soft Exosuit for Assisting Hip Extension

Cited by 3 publications

References 7 publications

Optimization of Torque-Control Model for Quasi-Direct-Drive Knee Exoskeleton Robots Based on Regression Forecasting

Optimization of Torque-Control Model for Quasi-Direct-Drive Knee Exoskeleton Robots Based on Regression Forecasting

Frontal hip exoskeleton assistance does not appear promising for reducing the metabolic cost of walking: A preliminary experimental study

Lower limb active prosthetic systems—overview

Contact Info

Product

Resources

About