Trajectory Tracking of the Cormoran AUV Based on a PI-MIMO Approach

Abstract

This paper describes the control strategy developed for the Cormoran, a low cost ocean observing platform, hybrid between AUVs and ASVs. This robot moves slightly underneath the sea surface following a previously planned route and regularly dives to make vertical profiles of the water column. Obtained data is transmitted in real time to the laboratory and assimilated into a numerical coastal model. The control structure proposed in this article comprises a Navigation Block and a MIMO Control Block. The former is based on a following Lyapunov-based algorithm and computes on-line the reference signals required by the latter to reach and robustly track the desired trajectory. The MIMO Control Block architecture is based on a PI-MIMO control strategy that commands the rudder and the propeller. Computer simulations utilizing a dynamic model of the Cormoran and realistic disturbances are presented.