Ankle gait-pattern generators

Abstract

Different lower limbs robotic rehabilitation devices are under study to help patients to recover gestures altered, affected or lost after traffic accident injury, illness or disability. Especially ankle because is a complex part of the human skeleton. In this paper we present our contribution to the ankle foot robotic device. We will present our strategy to generate the setpoint trajectory for ankle devices that is based on neural networks. The latter will generate the trajectory of the ankle from the angular position sensors of the hip joint and the knee joint. This gave good results, but we noticed during the tests that the ankle trajectory contained sporadic peaks. This was the case of the right and left ankle. An optimization algorithm is developed to eliminate this phenomenon; it will be presented soon.