Estimation of Vehicle Pose Using Artificial Landmarks for Navigation of an Underwater Vehicle

Abstract

This paper describes a localization method to localize a mobile vehicle in underwater environment. Particle filter based localization method isimplemented, which is based on Bayesian filter to deal with non-linear system.This method comprises prediction step and correction step to estimate the pose of the vehicle. The prediction step is achieved by a motion model of the vehicle with inertial sensor data acquired from Doppler Velocity Log, inertia sensors,and electronic compass. In the correction step, the pose of the vehicle is updated using range and bearing information of externally fixed landmarks in the vehicle work space. The performance of the proposed localization method is verified by experiment in a tank environment using four artificial landmarks. In the experiment, the motion information of the underwater vehicle is used as surge and yaw velocities obtained from DVL and AHRS sensors. The landmark information is acquired from the artificial landmarks using an image sensor. The experimental result shows that the proposed method successfully estimated the pose of the vehicle.