A Robust Fuzzy Autonomous Underwater Vehicle (AUV) Docking Approach for Unknown Current Disturbances

Abstract

In this paper, a robust autonomous underwater vehicle (AUV) docking approach that can handle unknown water currents is presented. Several assumptions and design considerations are made in our work. First, there is no onboard velocity sensor, and thus water currents will need to be estimated and compensated online. Second, the vehicle uses its ultrashort baseline (USBL) system to home itself toward an acoustic source situated at a rigidly bottom-mounted underwater docking station. Third, the dock is oriented at a preknown heading, and its absolute position is a priori known. Our docking approach incorporates a Tagaki-Sugeno-Kang (TSK) fuzzy inference system (FIS) that assists the vehicle with high level guidance maneuvers in the form of fuzzified commanded heading and speed vector fields. A current compensator is designed and applied to the fuzzy docking guidance to allow the vehicle to maintain its course in the presence of current disturbance. An extended Kalman Filter (EKF) is formulated to estimate the current and vehicle states. Extensive Monte Carlo simulations were performed to evaluate the current estimator and vehicle docking performance under various docking conditions. The simulation results demonstrated the inherent robustness of this designed fuzzy docking approach against unknown current disturbances, without any real-time velocity measurements.