Generating gait pattern of myoelectric active ankle prosthesis

Abstract

Aim of this paper is to generate human gait pattern in a myoelectric active ankle prosthesis. A prototype of motorized active ankle joint has been developed and it is controlled by users own EMG signal. EMG signal of six different lower limb muscles has been acquired and fused using standard fusion technique discarding spurious data. From the fused EMG data, different gait parameters like stride time, gait phase etc has been calculated. Joint trajectory during a gait cycle is obtained, discretized and combined with the gait parameters acquired from EMG and together they are fed to the robotic ankle prototype developed and programmed to behave like a servo. Servo motor follows the ankle trajectory using a PID position controller. The system has massive application in gait rehabilitation for people suffering from transtibial amputation.