Neural Network Fixed-Time Control of a Robotic System under Output Constraint

Abstract

A fixed-time robotic system control method is proposed with input deadzone under a delayed time constraint. Radial basis function neural networks (RBFNN) is employed, and an error-driven adaptive law is designed. The stability of the system is achieved by introducing a new error shifting function, and by the Lyapunov stability theory convergent tracking error to a small set near zero has been observed. Simulation results proved the method effectiveness.