Adaptive Fixed-Time Constraint Control for Human-Robot Interaction with Uncertainties using Neural Networks

Abstract

In this paper, a new control scheme using exponential-type barrier Lyapunov function (EBLF) is proposed for human-robot interaction, which can achieve high-performance trajectory tracking without dependence on the initial value. It has shown that the tracking error driven by the proposed control scheme will converge to a small set around equilibrium within a fixed time on different initial conditions. Moreover, human motion dynamics is captured by radial basis function neural networks (RBFNN) featured by universal approximation. Simulation results have demonstrated the satisfied performance of the developed control scheme.