First experimental results on adaptation algorithms for lower limbs active orthoses

In this work we present the first experimental results on gait-pattern adaptation for lower limb orthoses. The adaptation algorithm considers the orthosis-patient interaction forces and the Zero Moment Point (ZMP) criterion, allowing the patient to modify the gait-pattern as his/her degree of voluntary locomotion still maintaining the walking stability. A set of neural networks (NN) are used to decrease the time-consuming computation of the model and ZMP-based trajectory generation. The first neural network approximates the inverse dynamics and the ZMP optimization, while the second one works in the optimization procedure, giving an adapted desired trajectory according to orthosis-patient interaction. This trajectory adaptation is added directly to the trajectory generator, also reproduced by a set of neural networks. Also, a robust controller based on the ℋ∞ method is designed to attenuate the effects of external disturbances and parametric uncertainties in the trajectory tracking errors. Experimental results of the proposed NN-based adaptation algorithm, with one of the ankle joints of the model was replaced by an actual active orthosis, are presented.