Intelligent Fractional-Order Active Fault-Tolerant Sliding Mode Controller for a Knee Joint Orthosis

Over the last few decades, the appearance of robot applications has been continuously increasing in rehabilitation treatment due to the increasing numbers of stroke patients. In this paper, a novel fractional-order active fault-tolerant controller (FOAFTC) based on an adaptive nonlinear observer is proposed to detect, estimate, and compensate faults of a knee joint orthosis. The controlling term is derived based on the fractional-order sliding mode control (FOSMC) approach, while the switching term of this controller is designed by the fractional-order interval type-2 fuzzy logic control (IT2FLC) technique. In this regard, a nonlinear disturbance observer (NDO) is combined with this controller to estimate the muscular torque inserted by the patient, while increasing the accuracy and speed of the system performance. Adopting the Lagrange equations, a unique model is presented for the orthosis and the human lower-limb. The proposed strategy reduces modeling difficulties and the chattering phenomenon of the SMC technique. Additionally, the number of rules in the fuzzy controller is decreased as well. Following on, the closed-loop system stability is proved theoretically. Finally, the system performance based on the suggested controller is evaluated and compared with the results of two other different controllers including the conventional SMC and the fractional-order terminal sliding mode controller. Implying the proposed controller to the system leads to the smaller frequency and magnitude of the chattering effects, which provides more convenience for patients. Moreover, the efficiency of the suggested controller is illustrated in the presence of both actuator and sensor faults.