Dynamical analysis of human gait at low speeds

Abstract

Within the context of exoskeleton applications for rehabilitation of patients with partial spinal cord injury, there is a degree of actuator variety used to promote and control human gait movements. To study possible actuator alternatives, it is necessary to know the torques acting in the lower limbs joints during the gait cycle. This work’s objective is to obtain such required joint torques in a low speed gait scenario, in order to address rehabilitation gait in future works. A non-linear dynamical model capable of describing a simplified human gait dynamics is developed and implemented for simulation. Kinematic data retrieved from both scientific literature and experimental tests conducted by the research team are used as input to an inverse dynamics analysis of gait at the same low speed, allowing the validation and study of the limitations of the proposed model. Simulation results show that the developed model shares the shape and scale of the reference data and remains between its standard deviation limits. Therefore, it can be concluded that even with the data limitations of reference literature, statistical deviation of the experimental tests and modeling simplifications, the developed dynamical model presents satisfactory results. Future works will focus on experimental gait tests at rehabilitation speeds, which are slower than the speed of this paper, to allow a more complete validation of the model for such cases.