Gait Pattern Prediction with Adaptability of Stride Time Variation for Lower Limb Exoskeletons

Lower limb exoskeletons are the promising tools to assist in rehabilitation training for patients with walking disabilities. However, the fixed gait pattern cannot adapt to the variation of the user's gait pattern, which hinders the user's movement. In this paper, we consider the effect of stride time on gait patterns and propose a gait pattern prediction method with the adaptability of stride time variation. The method first constructs a gait pattern dataset for normal walking with the stride time variation. Then, the prediction function is learned based on the impact of the stride time at each phase of the gait patterns. Finally, the prediction function is used to predict the gait pattern for a given stride time. Experimental results show that our method significantly outperforms the dataset averaging-based method. Our gait pattern prediction method, which is adaptable to the variation of stride time, can provide a reliable reference trajectory for the lower limb exoskeleton, thereby improving the coordination of exoskeleton's movement.