Enhanced neural network control of lower limb rehabilitation exoskeleton by add-on repetitive learning

Abstract This paper addresses neural network (NN) control of a lower limb exoskeleton for rehabilitation. Two control schemes are presented for exoskeleton to conduct trajectory tracking tasks, in the presence of unmodeled dynamics of exoskeleton, interaction between human and exoskeleton, as well as additional disturbances. The first controller is purely NN-based, and a combined error factor (CEF), which consists of the weighted sum of tracking error and its derivative, is adopted to enhance the human safety by improved transient response. The second controller is developed based on a combined scheme of repetitive learning control (RLC) and NN, where the add-on RLC is used to learn periodic uncertainties that attribute to the repetitive motion of the exoskeleton leg. Stabilities of the controllers are proved rigorously in a Lyapunov way. It is worthwhile to highlight that although the pure NN controller can deal with periodic and non-periodic uncertainties simultaneously, the main feature of exoskeleton motion during rehabilitation therapy, namely, the repetitiveness, is fully ignored, thus could degrade the tracking performance. Compared simulation reveals that, the proposed control method has achieved a significant control effect with remarkable transient performance.