High performance path following for marine vehicles using azimuthing podded propulsion

Abstract

Podded propulsion systems offer greater maneuvering possibilities for marine vehicles than conventional shaft and rudder systems. As the propulsion unit rotates about its vertical axis to a specified azimuth angle, the entire thrust of the propeller contributes to the steering moment without relying on lift generation by a control surface such as a rudder. However, the larger sideforce and moment cause the ship to enter the nonlinear realm sooner than a ruddered vessel. Furthermore if the rudder or azimuthing propulsor is aft of the vessel's center of gravity then the system is non-minimum phase; during a turn the ship center initially moves in the direction opposite the turn. For these reasons it is necessary to design a robust maneuvering control system to set the azimuth angle of the propulsor in an intelligent and stable manner. This thesis focuses on the path following performance of a vessel with podded propulsion. The enhanced maneuvering abilities of such vessels allow the time constant of cross-track error response to be greatly reduced. Additionally these vessels can follow course changes and waypoints more precisely than ruddered vessels.

(cont.) A simple path following algorithm was developed to achieve this performance; the algorithm uses simulation-based feedforward terms to anticipate the sliding motion of the vessel during a turn. The stability and performance analysis was performed in three domains: linear theory, a nonlinear simulation, and experiments with a 12-foot autonomous surface vessel. Experiments confirmed that path following performance was vastly improved using the feedforward algorithm for waypoints at which the course change angle was large.