MARIN designed and built a modular autonomous underwater vehicle to be used in its model basins. In order to improve the control design, as well as its digital twin, the motions of the simulation model should closely resemble the motions of the true system. The goal of this work is to improve the estimates of the model parameters. The estimation is done during selfpropulsion tests to gather data from the true system response, while emphasising on one parameter at a time. Our AUV has more thrusters than there are degrees of freedom, and this property is used to find the relative thrust coefficients of the individual thrusters. The coefficients found are validated and it was found that they better match than the settings previously obtained with a captive test. The added mass is estimated in combination with the rigid body mass based on i) forced harmonic excitations, ii) free exponential decay and iii) closed loop step tests. The principle drag terms were estimated by sailing at a set of constant velocities and fitting the drag coefficients to the median values of the steady state values. Next to a scaling factor, the added masses and principle drag terms for a 6 DOF AUV are identified with self-propulsion tests without the use of an external captive set up. The tests for the hydrodynamic parameters can be performed while in closed-loop control, making them easy to repeat when the modular aspect of our AUV is tested.
measurements and controlautonomy and decision support