A Multi-AUV Collaborative Mapping System With Bathymetric Cooperative Active SLAM Algorithm

Abstract

Autonomous underwater vehicles (AUVs) play a pivotal role in the underwater Internet of Things (IoT). However, their capacity to fulfil large-scale bathymetric mapping is often constrained by limitations in navigational capabilities. This article proposes a homogeneous distributed collaborative mapping system, consisting of bathymetric mapping vehicles and its isomorphic server vehicle. The system achieves accurate positioning through bathymetric cooperative active simultaneous localization and mapping (BCA-SLAM) technology. This article mainly focuses on the server’s online path planning in BCA-SLAM using D-optimality metrics of the Fisher information matrix (FIM), to maximize the positioning accuracy of the collaborative system. A method for predicting the intervehicle loop-closure factor FIM was proposed for selecting the subsequent target point for the server, while a lemma for multiple augmented matrix determinants was devised to mitigate its computational burden. Experimental results have proved both the accuracy and efficiency of the proposed algorithm have been tested in semi-physical simulation.