Digital Twin of a Maneuvering Ship: Real-Time Estimation of Derivatives and Resistance Coefficient Based on Motion Sensor

Abstract

Retrieving certain hydrodynamics coefficients from a marine craft during a maneuver can be useful for various reasons, such as the validation of project specifications or the rapid verification of structural changes that could impact the vessel movement. Intended to estimate some of these parameters, the present work proposes a method purely based on traditional Extended Kalman Filter (EKF) focused for limited drift angles. Albeit not posing as a replacement to conventional estimations, such as from Computer Fluid Dynamics (CFD) — which solve equations in order of millions — and experimental tests — with its time-consuming preparation setups and post-analyses — the method can possibly present itself as a convenient and quicky technique to estimate the hydrodynamics coefficients in real time, as each iteration resorts only into a few dozen of equations. Preliminary results in the simulated environment called pyDyna — a python version of the Numerical Offshore Tank ship maneuvering simulator — indicate this procedure is faster along with an acceptable margin of accuracy, possibly pointing as a feature for future digital twin applications.