Identification of hydrodynamic parameters by simulation with Smoothed Particle Hydrodynamics

Nicolas Gartner (Oct. 2016 – June 2020)


Funding : SURFBOT DGA Rapid project

Supervision : Vincent HUGEL and Mathieu RICHIER


Generally research studies conducted in underwater robotics mainly focus on the development of control laws, without including the mecatronic design of the vehicle in the optimization of the control law. Actually this approach is not required as long as the vehicle operates in open water, which is the case for most of the missions. However in coastal zones that are strongly disturbed (zones named surf zones) it could be interested to optimize the mecatronic design of the vehicle to increase the robustness and/or the efficiency of the control law through the following points : estimation/observation of hydrodynamic parameters, reduction of the impact of disturbances caused by currents, and layout adaptation of the actuators to the control needs.
Hence on the one hand it comes to develop a simulation environment enabling the modeling of an underwater drone behavior as accurately as possible in the surf zone. The evaluation of the system requires the determination of innovative criteria to quantify its capability to carry out a given mission. These criteria will allow evaluating the mechanical design and the control command of an underwater drone in simulation. On the other hand, they could be directly integrated in an optimization process of the system parameters (overall shape, motor layout, holonomy or non-holonomy of the vehicle) according to the mission to execute (pipeline tracking, point stabilization, move to another place, course keeping, etc.). Finally all this research work could conduct to the development of reconfigurable robotic solutions, which would constitute one promising application for underwater vehicles.