Disturbance rejection based on adaptive neural network controller design for underwater robotic vehicle

This paper presents a disturbance rejection based on an adaptive neural network (DR-ANN) controller design for an underwater robot vehicle (URV). The disturbances are caused by environmental causes such as ocean currents or internal causes like dynamic system nonlinearity caused by uncertainties. A wave disturbance model is presented based on near-surface shallow water disturbances and deep water disturbances to test the URV controllers performance under these types of disturbances. A radial base function (RBF) is used to estimate both of the disturbances and the unknown uncertainty of the URV dynamics to reduce the nonlinearity effect. Two scenarios are presented to test the DR-ANN controller, where each scenario represent a path trajectory with a different disturbance model for the URV model. The DR-ANN stability was ensured using a Lyapunov function. The performance of the DR-ANN controller was evaluated by comparing the proposed controller with other existing works using simulation and numerical experiments. At the end, the results obtained show the superiority of the DR-ANN controller in the presence of disturbances and uncertainties.