Design and Control of a Portable Soft Exosuit by Musculoskeletal Model-Based Optimization.

Nianfeng Wang,Zicong Wang,Zitian Li,Shuhao Xia,Xianmin Zhang
DOI: https://doi.org/10.1007/978-981-99-6486-4_34
2023-01-01
Abstract:Universally applicable ankle assistance profiles can be generated through musculoskeletal model-based optimization, which can be applied to lightweight assistance conditions like unpowered exoskeletons and home rehabilitation. However, the effect of model-calculated optimal profiles on the energetics and kinematics of human walking still needs to be verified. Here, a portable soft exosuit was developed to apply model-based optimized moments and proportional biological moments to the ankle joint for comparison experiments. A low-level control strategy based on Gaussian process optimization was applied to each participant to track the target experimental moment accurately. Six participants walked on a treadmill at 1.25 ms $$^{-1}$$ under one unpowered and two powered conditions, where metabolic costs and joint kinematics were measured. Compared with the Exo-off condition, the optimal assistance exosuit significantly reduced metabolic cost by an average of 6.23%, while the proportional condition was 2.09%. This study presents a simulation and optimization design paradigm using local musculoskeletal modeling to develop new assistance patterns more efficiently.
What problem does this paper attempt to address?