Neural network based control system architecture proposal for underwater ship hull cleaning robot

Controlling of underwater robots require the estimation of unknown hydrodynamic forces and moments and disturbances acting on these vehicles. Using conventional control theory for design and synthesis of the controller seems to be almost impossible when we are dealing with such a nonlinear system. In this paper an artificial neural network based control system of a robot for underwater ship hull cleaning is proposed. The underwater robot has to move on a trajectory specified in terms of desired position according to the hydroacoustic localisation system.