Design of Tracking System and Disturbance Rejection using Neural Networks for Autonomous Underwater Vehicle (AUV)

Abstract

Autonomous Underwater Vehicle (AUV) is under-actuated system with highly nonlinear dynamic and Multi Input Multi Output (MIMO). AUV has stability problem due to external disturbance. These characteristics cause the AUV is difficult to be controlled and to track the reference signal automatically. Based on these problems, this paper present tracking control design and disturbance rejection using Neural Networks for AUV with 6 Degree of Freedom (DOF). The proposed control system is used to generate control signals to overcome the nonlinear dynamics of AUV. The reference signal is processed using Line of Sight (LOS) method to obtain the desired yaw of AUV. Simulation results show that the AUV is able to track the reference signal even in the presence of ocean current disturbance.