Diving autopilot design for underwater vehicles using multi-objective control synthesis

This paper presents the mathematical modeling, guidance and robust control synthesis of a highly maneuverable submersible vehicle (or underwater vehic le) when performing st specific mission at shallow submergence conditions. First, the vertical plane motions (heave and pitch) of the vehicle are mode led by a set of maneuvering equations. After model simplification, a state space model is compactly obtained. Then a state-feedback controller is prop osed for the accurate depth-keeping and pitch motion controls of the vehicl e. The control actions to the generalized plant can be provided by the mixe d H-2/H-infinity optimal synthesis as well as closed-loop pole constraint w ith LMIs. The feasibility of the guidance and control approach is verified with direct numerical simulations. The proposed approach ensures reasonable depth-keeping and minimal pitch motions, even under a given uncertainty co ndition.