Ship Tracking Utilizing Propeller Noise with a Compact Hydro-phone Array
محورهای موضوعی : Signal Processing; Image Processing
Mojgan Mirzaei Hotkani
1
(
Department of Electrical Engineering, Ferdowsi University of Mashhad, Iran
)
Seyed Alireza Seyedin
2
(
Department of Electrical Engineering, Ferdowsi University of Mashhad, Iran
)
کلید واژه: Normal-mode model, MUSIC algorithm, Matched Field Processing, Tracking ship noise, Underwater acoustic,
چکیده مقاله :
This study presents an innovative approach to ship tracking utilizing the analysis of propeller noise. It employs a hybrid technique that combines both time-domain and frequency-domain processing. In the frequency-domain aspect, the method utilizes matched field processing (MFP), which can be customized using diverse propagation models such as the normal modes, the image model, and the Lloyd-Mirror. Within this framework, an approximated time-domain frequency-limited covariance matrix is integrated, incorporating attributes of the Image propagation model, and coupled with the Multiple Signal Classification (MUSIC) algorithm. This integrated methodology leverages the strengths of both time-domain and frequency-domain strategies simultaneously. To validate the effectiveness of the proposed algorithm, actual data from a field trial conducted in collaboration with JASCO Applied Sciences Ltd at Duncan’s Cove, Canada, in September 2020 is employed. The trial involved deploying a hydrophone array in the outbound shipping lane, which collected broadband noise data from various types of vessels, enabling the estimation of vessel bearings. The results from this real-world trial demonstrate the practical performance of the proposed algorithm.
This study presents an innovative approach to ship tracking utilizing the analysis of propeller noise. It employs a hybrid technique that combines both time-domain and frequency-domain processing. In the frequency-domain aspect, the method utilizes matched field processing (MFP), which can be customized using diverse propagation models such as the normal modes, the image model, and the Lloyd-Mirror. Within this framework, an approximated time-domain frequency-limited covariance matrix is integrated, incorporating attributes of the Image propagation model, and coupled with the Multiple Signal Classification (MUSIC) algorithm. This integrated methodology leverages the strengths of both time-domain and frequency-domain strategies simultaneously. To validate the effectiveness of the proposed algorithm, actual data from a field trial conducted in collaboration with JASCO Applied Sciences Ltd at Duncan’s Cove, Canada, in September 2020 is employed. The trial involved deploying a hydrophone array in the outbound shipping lane, which collected broadband noise data from various types of vessels, enabling the estimation of vessel bearings. The results from this real-world trial demonstrate the practical performance of the proposed algorithm.
