Can Momentum-Based Control Predict Human Balance Recovery Strategies?

Abstract

Human-like balance controllers are desired for wearable exoskeletons in order to enhance human-robot interaction. Momentum-based controllers (MBC) have been successfully applied in bipeds, however, it is unknown to what degree they are able to mimic human balance responses. In this paper, we investigated the ability of an MBC to generate humanlike balance recovery strategies during stance, and compared the results to those obtained with a linear full-state feedback (FSF) law. We used experimental data consisting of balance recovery responses of nine healthy subjects to anteroposterior platform translations of three different amplitudes. The MBC was not able to mimic the combination of trunk, thigh and shank angle trajectories that humans generated to recover from a perturbation. Compared to the FSF, the MBC was better at tracking thigh angles and worse at tracking trunk angles, whereas both controllers performed similarly in tracking shank angles. Although the MBC predicted stable balance responses, the human-likeness of the simulated responses generally decreased with an increased perturbation magnitude. Specifically, the shifts from ankle to hip strategy generated by the MBC were not similar to the ones observed in the human data. Although the MBC was not superior to the FSF in predicting human-like balance, we consider the MBC to be more suitable for implementation in exoskeletons, because of its ability to handle constraints (e.g. ankle torque limits). Additionally, more research into the control of angular momentum and the implementation of constraints could eventually result in the generation of more human-like balance recovery strategies by the MBC.

Robust Walking Control of a Lower Limb Rehabilitation Exoskeleton Coupled with a Musculoskeletal Model via Deep Reinforcement Learning

“…Because of varied conditions of patients' disability, the humanexoskeleton interaction forces are unpredictable and could vary substantially from one patient to another, a very important factor to consider for controller development. Existing controllers for LLREs often focus on trajectory tracking, conventional Proportional-Integral-Derivative (PID) control [20], fuzzy control [8], model-based predictive control [21], impedance control [22,23], and momentum-based control [24]. The trajectory tracking approaches are primarily used for early-stage rehabilitation when patients have very weak muscle strength, its robustness against unexpected large perturbations or uncertain interaction forces is not great.…”

Reinforcement Learning and Control of a Lower Extremity Exoskeleton for Squat Assistance

“…Balance training in the presence of external perturbations ( Horak et al, 1997 ) is considered as one of the more important factors in evaluating patients’ rehabilitation performance. A rehabilitation exoskeleton can be employed for balance training to achieve static stability (quiet standing) or dynamic stability (squatting, sit-to-stand, and walking) ( Bayon et al, 2020 ; Mungai and Grizzle, 2020 ; Rajasekaran et al, 2015 ). Squatting exercises are very common for resistance-training programs because their multiple-joint movements are a characteristic of most sports and daily living activities.…”

“…To the best of our knowledge, investigations presenting robust controllers against large and uncertain perturbation forces (e.g., due to human interactions) have rarely been carried out as biped balance control without perturbation itself is a challenging task. Most existing balance controller designs for such lower extremity rehabilitation exoskeletons focused mostly on the trajectory tracking method, conventional control like Proportional–Integral–Derivative (PID) ( Xiong, 2014 ), model-based predictive control ( Shi et al, 2019 ), fuzzy control ( Ayas and Altas, 2017 ), impedance control ( Hu et al, 2012 ; Karunakaran et al, 2020 ), and momentum-based control for standing ( Bayon et al, 2020 ). Although the trajectory tracking approaches can be easily applied to regular motions, its robustness against unexpected large perturbations is not great.…”

“…On the other hand, model-based predictive control could be ineffective or even unstable due to inaccurate dynamics modeling, and it typically requires a laborious task-specific parameters tuning. The momentum-based control strategies have also been applied to impose standing balancing on the exoskeleton ( Bayon et al, 2020 ; Emmens et al, 2018 ), which was first applied in humanoid robotics to impose standing and walking balance ( Lee and Goswami, 2012 ; Koolen et al, 2016 ). This method aimed to simultaneously regulate both the linear and angular component of the whole body momentum for balance maintenance with desired GRF and CoP at each support foot.…”

Variable Impedance Control for Bipedal Robot Standing Balance Based on Artificial Muscle Activation Model

“…However, the ZMP control method has some limitations: the ZMP was calculated by sensor information feedback which lags behind the actual attitude change, and the delay will cause the controller ring [4]. According to the law of conservation of momentum, some scholars simultaneously adjust the angular and linear momentum to complete the bipedal robotic standing balance control [5,6], and this approach relies on the robotic dynamic model which is difficult to improve the robustness of the standing balance control. With the development of intelligent control algorithms which are increasingly being used to solve the robotic standing balance control problems [7], the intelligent control algorithms rely on a large amount of test data.…”