A Neural Network Based Control System for Underwater Vehicle

A Neural Network based Controller (NNC) is presented for the motion control of the underwater vehicles. With this kind of Neural Network Controller, an on-line learning method can be performed without the sample data in several sample periods, resulting in the high learning speed of the controller. The developed control system was applied to the 4-dimension control of a UV and succeeded in simulation platform. It is shown that the NNC has a great possibility to solve the problems on the control system design of the future underwater vehicles.