Finite-time tracking control of underactuated surface vehicle with tracking error constraints

Abstract

This paper studies the finite-time tracking control of an underactuated surface vehicle (USV) subject to the lumped uncertainties including actuator faults, external disturbances, and plant uncertainties. A tracking controller with error constraints is given to derive the prescribed tracking performance for the vehicle. Subsequently, the adaptive neural network (NN) is adopted to estimate the lumped uncertainties. Meanwhile, an adaptive switching mechanism is devised to avoid chattering of state output and improve adaptation rate to compensate approximation error. Furthermore, a new performance function is introduced to obtain a clear indication of actual convergence time, and the nonlogarithmic transformation function (NTF) is constructed to solve the potential singularity problem in the logarithmic error mapping function. Some simulations have been presented to prove the excellent tracking performance of the developed method.