Optimal Measurement Strategies for Environmental Mapping and Localization of a Biomimetic Autonomous Underwater Vehicle

Abstract

A Bayesian method, using an extended Kalman filter, combining localization and environmental mapping by a biomimetic autonomous underwater vehicle (BAUV) is implemented. An optimal-cost measurement strategy is applied to decide the way-points to be navigated. This strategy selects the best sensor measurement by choosing one of several forward-looking directions. The localization and environmental mapping problem is then transformed into a nonlinear two-point boundary value problem. The optimal policy is to maintain the accuracy of the predicted states and to approach minimal cost of observation by solving the control problem. Experiments performed using a testbed BAUV confirm the effectiveness of the proposed method.