Design and Implementation of an Intelligent Control System for a Lower-Limb Exoskeleton to Reduce Human Energy Consumption

Abstract

Power augmentation is known to be one of the important applications of Exoskeletons. This paper designs a control strategy to reduce the energy consumed by users in power augmentation mode. The strategy aims to calculate and apply the interaction force between humans and robots according to human intentions. To realize human intentions, the movement's kinematic characteristics and the user's muscular activity were used. The movement patterns were learned by the robot using a set of adaptive oscillators. The human movement pattern in each movement cycle was considered the basis for predicting human intention in the next cycle. Thereby, the robot's optimal path and interaction torque were calculated and applied to the robot by the internal control loop. Within this process, Electromyography (EMG) signals are used to coordinate the robot's interaction torque with human intention. This torque is modified continuously based on the EMG signals for each moment of the movement phase. Further, this control strategy's performance was first simulated and eventually evaluated by implementing them in the experimental testbed. The results confirmed that the control strategy adopted help to achieve predefined goals.