A hybrid admittance control strategy for fluid-driven origami actuators in lower limb exoskeleton

A hybrid admittance control strategy is proposed for a lower limb exoskeleton actuated by a fluid-driven origami-inspired artificial muscle (FOAM). This approach combines the force-responsive characteristics of conventional admittance control with the adaptive learning capability of neural networks to assist human subjects during knee rehabilitation. To determine the appropriate level of assistive force, electromyography (EMG) signals are incorporated via fuzzy logic control, thereby promoting optimal muscle engagement throughout the recovery process. The experimental results validate the controller's ability to maintain performance despite external load variations and unpredictable user exertion. This robustness is critical for managing the nonlinear dynamics inherent in exoskeleton actuation, ensuring consistent assistance across diverse operating conditions. Performance evaluation reports a root mean square error (RMSE) of 1.944 ± 0.447 Nm under hybrid admittance control, indicating high stability and adaptability of assistive torque output in response to varying muscular effort. To affirm the feasibility and adaptability of the proposed strategy, the experimental validation is conducted under controlled laboratory conditions. These findings highlight the potential of the proposed system to enhance human–robot interaction and broaden the applicability of exoskeletons in diverse rehabilitation scenarios.