Forward dynamics simulation for analyzing interactions with passive lower-limb assistive devices via optimization of physical posture

抄録

Forward dynamics simulations can significantly reduce the development costs associated with new wearable assistive devices by minimizing the need for prototyping and experimentation. However, current model-based methods rely on experimental data to define the model posture for the simulation, making it challenging to design the wearer's posture suitable for reducing physical load with passive lower-limb assistive devices. To address this challenge, this study proposes a forward dynamics simulation method that focuses on optimizing physical posture. This simulation approach calculates human-device interactions, including reaction forces, physical posture, and physical load, based on the physical posture optimized by a cost function that evaluates both the physical load and the suitability of the posture for work, which contributes to estimating the impact of the device on the physical load of the wearer. To investigate the validation and advantage of the proposed simulation, this study compared the simulation and experimental results for existing and new passive lower-limb assistive devices. The results show that although the proposed method did not accurately simulate the individual human-device interactions for a given participant using the passive lower-limb assistive device, it accurately estimated the average reaction forces and lower limb posture across all participants. Additionally, the proposed method correctly represented the differences in reaction forces and lower limb posture resulting from changes in device structure. Consequently, the proposed simulation has advantages in computationally evaluating the device’s performance in reducing the lower limb loads and contributes to the development of effective assistive devices for preventing musculoskeletal disorders.