Author ORCID Identifier

https://orcid.org/0009-0001-7273-1086

Semester

Fall

Date of Graduation

2023

Document Type

Thesis

Degree Type

MS

College

Statler College of Engineering and Mineral Resources

Department

Industrial and Managements Systems Engineering

Committee Chair

Imtiaz Ahmed

Committee Member

Srinjoy Das

Committee Member

Bin Liu

Abstract

In the realm of marine surveillance, track association constitutes a pivotal yet challenging task, involving the identification and tracking of unlabelled vessel trajectories. The need for accurate data association algorithms stems from the urge to spot unusual vessel movements or threat detection. These algorithms link sequential observations containing location and motion information to specific moving objects, helping to build their real-time trajectories. These threat detection algorithms will be useful when a vessel attempts to conceal its identity. The algorithm can then identify and track the specific vessel from its incoming signal. The data for this study is sourced from the Automatic Identification System, which serves as a communication medium between neighboring ships and the control center. While traditional methods have relied on sequential tracking and physics-based models, the emergence of deep learning has significantly transformed techniques typically used in trajectory prediction, clustering, and anomaly detection. This transformation is largely attributed to the deep learning algorithm’s capability to model complex nonlinear relationships while capturing both the spatial and temporal dynamics of ship movement. Capitalizing on this computational advantage, our study focuses on evaluating different deep learning architectures such as Multi Model Long Short-Term Memory (LSTM), 1D Convolutional-LSTM, and Temporal-Graph Convolutional Neural Networks— in addressing the problem of track association. The performance of these proposed models are compared against different deep learning algorithms specialized in track association tasks using several real-life AIS datasets.

Share

COinS