An adaptive neural network with nonlinear FOPID design of underwater robotic vehicle in the presence of disturbances, uncertainty, and obstacles

Abstract

An adaptive neural network with nonlinear fractional-order PID (ANNFOPID) controller design is proposed for underwater robotic vehicle (URV) to solve the path tracking problem. The path tracking problem is caused by disturbances and unknown uncertainties of the underwater vehicle dynamic model. The disturbances were presented with the URV model to evaluate the performance of the ANNFOPID controller to reject the disturbances. At the same time, an obstacle avoidance model has been presented with ANNFOPID controller to evaluate the controller ability to maneuver around different obstacle locations. A radial base function (RBF) has been used to estimate both of the disturbances and the unknown uncertainties of the dynamic model. An improved slime mould algorithm (ISMA) is presented to invent a new trajectories for the underwater vehicle and enhance the ability of the URV to overcome the obstacles problems. At the end, the results obtained show that the ANNFOPID controller present an outstanding performance in comparison with other existing works to overcome the disturbances, unknown uncertainties, and obstacles problems effectively.