Adaptive second-order sliding model-based fault-tolerant control of a lower-limb exoskeleton subject to tracking the desired trajectories augmented by CPG algorithm

Reference trajectory generation and trajectory tracking in the presence of uncertainties and disturbances are among the most challenging topics in the realm of exoskeletons. Central pattern generation (CPG) is one of the trajectory design methods for walking robots, which acts as a limit cycle and produces smooth and rhythmic trajectories by rapid cancelation of the effects of exogenous disturbances. In this paper, the reference trajectories of robot joints are designed by combining modified Hopfield oscillators, which allows adjusting the frequency and amplitude of walking. For online modification of the reference joint trajectories, feedback error signal from zero-moment point (ZMP) criterion and error signal of impedance filter are used. To counteract the adverse effects of uncertainties and exogenous disturbances and to achieve high trajectory tracking performance, dynamic adaptive fast terminal sliding mode control strategy (DAFTSMC) is applied, which provides chattering-free control signals and demonstrates finite-time convergence rate, as well as high robustness against uncertainties and disturbances with unknown range. Moreover, the robot’s maximum walking stability based on the ZMP criterion is achieved through the proper movement of the waist joint. Additionally, control parameters, oscillator gains, and their connections are optimized to ensure the highest performance. Finally, the performance of the proposed method is compared with two other control schemes: conventional sliding mode control (SMC) and CPG-based hybrid SMC. The obtained results show the superiority of the proposed approach.