A robust neuro-based adaptive control system design for a surface effect ship with uncertain dynamics and input saturation to cargo transfer at sea

This paper is concerned with the problem of safely cargo transfer over a ramp between a cargo vessel and a lighter surface effect ship (SES). For this purpose, an adaptive neural network (NN) controller is proposed to control of ramp motions in the presence of entirely unknown dynamics and disturbances wherein the controller is subject to input saturation constraint. In this regard, we develop a novel non-affine nonlinear SES model by considering the nonlinear relationship among the air cushion pressure, the air flow into and out of the air cushion and the air cushion volume. To deal with the saturation constraint, an auxiliary system is proposed. To ensure the system stability, Lyapunov's direct method is employed to investigate the uniformly ultimately boundedness (UUB) of all closed-loop system states. The simulation results demonstrate the effectiveness of the proposed controller in critical sea conditions, including regular and irregular waves with high frequencies and large amplitudes. A comparative analysis with model-based approaches including Proportional Integral Derivative (PID) and Linear-Quadratic Regulator (LQR) control systems are given to highlight the performance of the controller.