Motion optimization strategy for Bearing-Only Tracking performed with a team of Autonomous Underwater Vehicles navigating in formation

摘要

Complex systems composed of multiple robotic agents are increasingly studied for their use in the marine environment. In this work, we address the problem of Target Motion Analysis (TMA) through bearing observations only, performed with a team of Autonomous Underwater Vehicles (AUVs) navigating in formation. This objective is typical in underwater passive acoustic source localization and tracking scenarios. As demonstrated in several studies, the Bearing-Only Tracking (BOT) problem is strictly related to the relative motion of the observer and the target. We propose to optimize the team formation path with respect to the target motion. Consequently, this work presents a distributed control framework to achieve coordinated motion control of the agents and cooperative target tracking. The control framework implements a novel adaptive approach building on the Model Predictive Control (MPC) scheme, in order to increase target tracking performances and autonomy of the system. The proposed implementation explicitly accounts for the major limits associated with underwater acoustic communication as latencies and bandwidth.