Data-driven system identification of hydrodynamic maneuvering coefficients from free-running tests
Guillermo Chillcce, Ould el Moctar
Abstract
A data-driven system identification approach was developed to identify the hydrodynamic coefficients of a mathematical maneuvering model. The method, developed primarily for use in the context of autonomous shipping, solved the ship motion equations using measurements from free-running model tests, whereby an efficient recently developed Euler equation-based numerical approach determined the zero-frequency added masses. The method is simple and robust and incorporates the physical properties of hydrodynamic forces to enforce a physically realistic solution. The method was verified and validated with free-running maneuver tests. The predicted ship kinematics and trajectories compared favorably with the measurements. The potential of the method was demonstrated.