A robust adaptive-fuzzy-proportional-derivative controller for a rehabilitation lower limb exoskeleton

Achieving high performance controller for multi-joints actuators on rehabilitation lower limb exoskeleton (RLLE) is difficult due to its non-linear characteristics. The controller performance with less tracking error is a key challenge in their controller. Therefore, this paper presents a new particle swarm optimization based initialization of model reference adaptive for fuzzy logic proportional derivative controller (Adaptive-FLC-PD), used in RLLE for passive mode rehabilitation exercise. The RLLE modelling, which integrates a lower-limb exoskeleton coupled with a direct current motor as joint actuator and a patient leg model, was simulated in MATLAB. The lower-limb exoskeleton motion is realised via a trajectory tracking method that imitates a therapist-administered manual activity during passively rehabilitation exercise. An Adaptive-FLC-PD was designed to control the direct current motor and drive the hip and the knee of the lower-limb exoskeleton. The stability analysis of Adaptive-FLC-PD has been shown by the applied Lyapunov function. The performance of the Adaptive-FLC-PD was compared with the fuzzy logic controller (FLC) and FLC-proportional derivative (FLC-PD) algorithms. The numerical analysis ascertained the performance of the Adaptive-FLC-PD in designing, tuning and simulating control system of RLLE.