Asynchronous UKF-based localization of an underwater robotic vehicle for aquaculture inspection operations

摘要

This paper presents a methodology for localizing an autonomous tethered underwater robotic vehicle deployed for aquaculture inspection operations. The developed methodology is based on an Asynchronous Unscented Kalman Filer (UKF) to fuse information from onboard sensors along with a'priori knowledge about the aquaculture's geometry and the underwater robot's hydrodynamics. The performance of the proposed methodology is assessed through accuracy and consistency metrics generated from simulation benchmarks.