This work of thesis has been developed during my internship in "MDM Team", an Italian company specialized in underwater robotics. My task was to identify the model parameters for an Autonomous Underwater Vehicle (AUV) called Zeno. In order to get this result I formulated a sequence of optimization problems; each one gives out part of the set of parameters, starting from the result of the previous problems and from sensors data, collected during specific tests performed in a testing pool. The identified model has been than used in order to add a Computed Torque action to the already implemented PID motion controller, ensuring better performances in the trajectory execution.