Optimal adaptive super-twisting sliding mode control base on zero momentum point stability criterion of a lower limb exoskeleton

Disturbance and bounded uncertainty are the most important factors which can be degrade efficient performance of the lower limb exoskeleton. While sliding mode control is a robust control approach against such disturbances, however, by applying the boundary layer in spite of chattering phenomenon, robust performance becomes feeble. In order to overcome this drawback, high order sliding mode algorithms like supper twisting has been proposed in which, chattering phenomenon is mitigated by eliminating the boundary layer. In this paper, an adaptive supper twisting sliding mode control is proposed for a lower limb exoskeleton robot in which the sliding variable and its derivative tend to zero continuously in presence of the disturbance and bounded uncertainty. In addition, the desired trajectory of the upper limb is determined so that in each moment the stability of the robot is guaranteed based on zero momentum point criterion. To achieve maximum stability and minimum error in tracking of the desired trajectories, the controller parameters and the upper limb desired trajectory parameters are optimized using the Harmony Search algorithm. Robot is modeled in ADAMS and then control inputs are applied to the Adams model. Finally, Performance of two controllers is compared. Simulation results reveal the performance of the proposed controller is better than the optimal sliding mode controller.