Design and implementation of a neuromuscular controller on the e-walk hip orthosis for partial assistance during the gait cycle

Abstract

Among the large spectrum of wearable robotic devices, research have been highly prolific in investigating the control of lower limbs orthoses and most specifically the assistance of the gait cycle as walking is considered one of the most important activity of daily living (ADL). The core of this internship inserts within the described quest as it bears the promise of a new mid-level controller dedicated to a hip orthosis in order to provide partial assistance during the gait cycle. This new assistance method belongs to the category of so-called bio-inspired controllers. The theoretical work as well as the software development carried out intervene at 3 levels; The upper level of the mid-level proposes a synchronization layer based on finite state machines (FSM) which aims to synchronize the state of the controller with the evolution of the walker according to the different phases of the gait cycle. The lower level offers firstly, an im- proved skeletal muscle model for state estimation that aims to capture the physiology of human muscles, and secondly, a state feedback controller that mimics commands of the nervous system during the gait cycle in terms of muscle stimulation. The final objective is to provide adequate torque assistance, close to the biological torque pro- file and that reduces the effort when walking. The controller will be implemented for the very first time on a partial assistance hip orthosis; The E-Walk V1 orthosis of the REHAssist research group. The obtained results support well the effectiveness of this control approach, paving the way for entirely new applications and future research questions and adding notable contributions to the literature.