Optimisation of reference gait trajectory of a lower limb exoskeleton

Lower limb exoskeletons have gained considerable interest in recent years as a research topic for creating aids for people with walking disabilities and strength augmenters for pilot walkers. A crucial practical problem, however, is generating the reference trajectory of the joints. In this paper, we solve the reference trajectory problem by a novel approach which obtains the angle trajectories of knee joints from the hip joints. The relationship between the angle trajectories of the knee and hip joints is acquired through kinematic models of the lower limb exoskeleton. In these models, the parameters of the joint position trajectories are optimised by a swarm fish algorithm with variable population. The proposed approach is validated in virtual simulations and a physical prototype of an exoskeleton system. The experimental results confirm that the reference trajectory generation approach accurately reproduces human walking.