Generic model control based on fuzzy neural network for underwater vehicles

In this paper, a generic model trajectory tracking control method is proposed for underwater vehicles. A set of sliding mode switching functions based on the trajectory tracking error are constructed as the controlled system, so that the condition of the relative degree of 1 is satisfied. Considering the problem that it is difficult to establish the accurate motion model of underwater vehicle, a method for online identification based on fuzzy neural network is presented. Due to the fact that the speed information of underwater vehicle is needed in construction of the sliding mode switching function, the state estimation is performed by the sliding mode observer. In order to compensate the uncertainties, the sliding mode switching item is added to the controller. Aiming at the chattering problem caused by sliding mode switching item, an adjustment method of sliding mode gain based on exponential function is proposed. The stability of the system is analyzed by the Lyapunov theory. Finally, the proposed control method is verified by simulation.