Adaptive Sliding Mode Control in Driving a Power Augmentation Exoskeleton Based on Minimization Interaction Forces between Human and the Robot

Exoskeleton robots are a motion-assist device having anthropomorphic structure to power the user for movement via motoric actuators. The exoskeletons are developed for two main applications: rehabilitation of disabled patients, and augmentation of human power during working heavy jobs. Long walking especially with load carrying is a case causes early fatigue. Thus, exoskeletons provide a solution to enhance human power during long marches. To this end, coordination of movements between human and robot is necessary. Many strategies have been proposed to control the interaction between human and exoskeleton, which one method is estimation of the interaction force and try for zeroing this force. In this method the interaction force is estimated based on the difference of movement between corresponding joints of human and exoskeleton, so there is no need to load cells to measure the interaction force. In order to extend this strategy, in this paper we exploit the method of sliding mode control with adaptive sliding gains, and its performance is compared with the case in which the gains are constant. For both of methods the stability of controller is proved according to Lyapunov theory. The exoskeleton robot is modeled based on a 3-link articulated structure, and for performing the leg movement during swing phase at which the joint displacements and velocities are larger and faster. Finally, performance of the controllers on this model is evaluated via numerical simulations. The results show the adaptive sliding mode control is more successful in tracking references and zeroing the interaction forces.