An Approach to combine Methods for Cooperative Navigation and Optimal Sensor Placement in Marine Robotics

In this paper we present an approach to perform position estimations of a single Autonomous Underwater Vehicle (AUV) on an unknown track, employing noisy range measurements to a single Autonomous Surface Vehicle (ASV). The ASV has to perform the estimations and to use them simultaneously to compute its own trajectory in a way that enables it to continuously perform the position estimation with an acceptable accuracy. To this extend, methods for position estimation with an Extended Kalman Filter (EKF) will be merged with methods for Optimal Sensor Placement (OSP), namely with Empirical Gramians. The methods will be described, and their feasibility will be shown with numerical simulations using MATLAB.