Control-Oriented Planar Motion Modeling of Unmanned Surface Vehicles

This paper describes a comparison of experimentally identified dynamic models for the planar motion of an unmanned surface vehicle (USV). The objective is to determine a model which is rich enough to enable effective motion planning and control, simple enough to allow straight forward parameter identification, and general enough to describe a variety of hullforms and actuator configurations. Starting from a three degree-of-freedom nonlinear model obtained from physical principles, we consider four simplified variants: (1) a linear model obtained by linearizing about straight, constant-speed motion, (2) a first order steering model (for turn rate) coupled with a first order speed model, (3) a second order steering model (for turn rate and sideslip angle), coupled with a first order speed model, and (4) a nonlinear model for low speed operation. The paper provides analysis of system identification data collected from field trials of three USV platforms in Summer 2010. The platforms represent three distinct control system implementations: a servo-actuated outboard engine, a servo-actuated jet-drive thruster, and differential thrusters.