Adaptive Robust Control for a Lower Limbs Rehabilitation Robot Running Under Passive Training Mode

This paper focuses on the problem of the adaptive robust control of a lower limbs rehabilitation robot (LLRR) that is a nonlinear system running under passive training mode. In reality, uncertainties including modeling error, initial condition deviation, friction force and other unknown external disturbances always exist in a LLRR system. So, it is necessary to consider the uncertainties in the unilateral man-machine dynamical model of the LLRR we described. In the dynamical model, uncertainties are (possibly fast) time-varying and bounded. However, the bounds are unknown. Based on the dynamical model, we design an adaptive robust control with an adaptive law that is leakage-type based and on the framework of Udwadia-Kalaba theory to compensate for the uncertainties and to realize tracking control of the LLRR. Furthermore, the effectiveness of designed control is shown with numerical simulations.