Preliminary experimental evaluation of a Doppler-aided attitude estimator for improved Doppler navigation of underwater vehicles

Abstract

Doppler navigation is one of the most widely practiced navigation methods for underwater robotic vehicles, yet its accuracy is limited both by sensor calibration and by the accuracy of the vehicle's attitude measurement. This paper reports a method to improve the dynamic attitude estimation for underwater vehicles employing Doppler sonar and low-cost micro-electro-mechanical systems (MEMS) attitude and heading reference systems (AHRSs). Our approach employs three-axis velocity measurements from a Doppler velocity log (DVL) to estimate acceleration to provide corrections to the AHRS accelerometer signals that improve the dynamic estimates of vehicle roll and pitch. We report an implementation of this approach with a nonlinear complementary filter to estimate the vehicle attitude dynamically. We report the results of an in-water comparative experimental evaluation of the proposed and previously reported attitude estimation methods using low-cost MEMS AHRSs. We also report an experimental evaluation of the effect of the attitude estimation accuracy on the spatial accuracy of three-dimensional Doppler navigation of underwater vehicles.