An autonomous location prediction model for maritime transport applications: a case study of Persian Gulf

Abstract

ABSTRACT In this paper, a machine-learning-based location prediction model for maritime transport applications is proposed. In the proposed model, only environmental parameters are used as the input of the model, which makes the model autonomous. Meaning that the proposed model does not require information from any source other than onboard sensors. The Persian Gulf is used as the test case in this study. The sea depth, magnetic field intensity, and gravitational field intensity are three input features of the proposed model, and the corresponding coordination are the output. In the proposed methodology, the problem of location prediction is formulated as a regression problem and the above three parameters are used to train the main regression models. In this paper, two methods are used as the baseline, artificial neural networks (ANN) and support vector regression. The simulation results illustrate the high accuracy and good performance of the ANN in location prediction tasks.