Accurate geo-referencing method for AUVs for oceanographic sampling

Abstract

The objective of the paper is to understand, characterize and enhance the achievable performance of the system of a state-of-the-art marine observation device, the Oceanserver IVER2 Autonomous Underwater Vehicle(AUV), in the Singapore coastal zone and with regard to accurate sampling of oceanographic properties. This paper discusses modifications made to the AUV, in order to make it useful for experiments in this region, which includes shallow water, strong currents, poor visibility, heavy traffic and a poor acoustic environment. These factors strongly influence the choice of navigation method and the system architecture which will enable the AUV to obtain accurate geo-referenced oceanographic properties, and to ensure its safe operation. Our science experiments usually involve sampling at various areas around the coast of Singapore within the same day; this calls for consistent positioning methods that allow ease in deploying and retrieving related equipment.This paper illustrates the use of ultra-short baseline (USBL) tracking system and ranges from an acoustic modem fused by a particle filter to aid the dead reckoning algorithm of the IVER2. The performance of the modified system is shown through simulations and field experiments.

Fusing acoustic ranges and inertial measurements in AUV navigation: The Typhoon AUV at CommsNet13 sea trial

“…The simultaneous equipment of a DVL and an USBL can however be too much expensive on a low-cost vehicle. In [6] the authors improve dead-reckoning navigation by integrating USBL measurements, proving even in this case the benefit of sensor fusion on experimental data. The results obtained in the previous works are relevant in terms of the estimate error, but the computational cost required by a particle filter is considerable.…”

“…Particularly interesting are the experimental results reported in some of these works. In [5] and [6] the measurement integration process is realized using a particle filter. The first combines DVL data with USBL observations to obtain the trajectory estimate of a lawn-mower path mission, showing how USBL corrections are able to reduce the longterm drift of the DVL estimate.…”

“…The results obtained in the previous works are relevant in terms of the estimate error, but the computational cost required by a particle filter is considerable. For instance, the approach presented in [6] relies in a dedicated CPU for the filtering algorihtm. In [7] an Extended Kalman Filter (EKF) is implemented to tighly couple the inertial-based estimate with USBL differential range and time measurements.…”

Sensor Modeling for Underwater Localization Using a Particle Filter

“…Acoustic positioning systems are the most used underwater external positioning approaches, from long baseline (LBL) [ 5 , 6 , 7 ] to sort baseline (SBL) [ 8 ] and ultrashort baseline (USBL) [ 9 , 10 ]. However, these sensors suffer from multipath Doppler effects and thermoclines, which induce acoustic reflection effects.…”

A single acoustic beacon-based positioning method for underwater mobile recovery of an AUV

“…Khan et al fused the USBL data with a particle filter estimate of dead reckoning, and the simulation and field experiments verify the performance of MSDF. 15…”