Gaussian sum filter based SLAM for autonomous navigation of underwater vehicles using acoustic sources

Abstract

Underwater localization is crucial for autonomous navigation of underwater vehicles. This paper presents a method of acoustic based underwater localization. It utilizes artificial acoustic sources as underwater landmarks. A vehicle is equipped with two hydrophones so that it acquires acoustic signals emitted by the acoustic sources. Directional information of the acoustic source is estimated by time difference of arrivals. Then, Gaussian sum filter based SLAM is performed by fusing inertial sensors with the directional angles of acoustic sources to estimate locations of both the vehicle and acoustic sources. The proposed method can provide an accurate acoustic based underwater localization without any prior information of source location. The performance of the proposed method is demonstrated by field test conducted in an inland water environment. ？ 2019 IEEE.