Experimental Dynamics Identification of Autonomous Underwater Vehicle and Modified Model Reference Adaptive Controller Design*

This paper investigates experimental diving dynamic model identification of Autonomous Underwater Vehicles (AUVs). An effective identification method based on tracking differentiator and augmented recursive least square (TD-ARLS) estimator is introduced to identify unknown model parameters. Depending on the identified model, a modified Model Reference Adaptive Control (MRAC) law is proposed to obtain optimal control performance. The identified model is adopted as the reference model to design the modified MRAC for the AUV’s diving dynamics. The lake trials of Explorer 1000 AUV and simulation results illustrate the effectiveness of the proposed method.