An ANN Based Approach for Gait Prediction of a Lower-Limb Exoskeleton with Plantar Pressure Sensors

This paper proposes an approach based on Artificial Neural Network ANN method for gait prediction of a lower-limb exoskeleton equipped with plantar pressure sensors and a pair of crutches. This approach can be implemented to predict the exact moment to change gait motion status. Further, the proposed approach can help to decide the starting movement speed of the pilot, through predictions on angular velocities of joints of both knees and hips. In this way, the exoskeleton can cope better with the pilot. Experimental results show that the new approach can capture the starting point of a new move, as well as predict the starting movement speed based on inputs from pressure sensors installed under pilot's plantar and crutches.