Intuitive Gait Pattern Generation for an Exoskeleton Robot

Exoskeleton robots are being studied and developed in various forms according to user. Among them, the most concentrated area is the lower limb exoskeleton robot for “walking,” which is the basic exercise of paralyzed patients. In this study, we utilize the gait cycle, speed, and stride, which are important dependent factors of gait, and not the control technique that uses the predefined gait pattern using the average value of the gait data of the general person used by the existing rehabilitation exoskeleton robot. By creating an end-point reference using a walking element and inverse kinematics, and by using a dynamic movement primitive technique to learn the gait data of the general public, generating the gait patterns of various walking environments without storing them in advance is possible. In this paper, we applied this method to the exoskeleton robot and show robot to generate various strides gait pattern by experiments.