Indirect adaptive neural tracking control of USVs under injection and deception attacks

Abstract

This paper investigates the tracking control issue of unmanned surface vessels (USVs) under internal/external uncertainties as well as injection and deception attacks, and proposes a novel adaptive neural network tracking control solution. In this work, the backstepping design framework is used to solve the control design issue. To resolve the time-varying gain caused by the injection and deception attacks, indirect adaptive neural and single-parameter learning approaches are employed. In the kinematic channel, internal and external uncertainties as well as the additional terms caused by the injection and deception attacks are regarded as heterogeneous uncertainties. And then, the heterogeneous uncertain term is transformed into a linearly parameterized form with only one parameter with the aid of the neural network technique and single-parameter learning approach. Furthermore, an adaptive law is constructed to update the unknown parameter, and an indirect adaptive neural learning approach is developed. Finally, a novel indirect adaptive neural tracking control solution is proposed to implement the suppression of heterogeneous uncertainties including internal and external uncertainties as well as attacks. The theoretical analysis indicates that all signals in the closed-loop tracking control system of USVs are bounded under the presented scheme. Simulation results demonstrate the effectiveness of the proposed scheme.