An LSTM-based Approach to Fuel Consumption Estimation in Digital Twin Ship

Abstract

The maritime industry plays a vital role in global trade and transportation, yet it also contributes significantly to CO2 emissions. Efforts to reduce emissions and operational costs have spurred the need for accurate fuel consumption estimation models. This paper introduces a Long Short Term Memory (LSTM)-based approach to enhance the digital twin modeling of fuel consumption of the R/V Gunnerus research vessel. We use correlation and sensitivity analyses to select input parameters and optimize the LSTM model configuration, alongside real data from R/V Gunnerus for verification of the model. Results demonstrate the efficacy of the proposed model in accurately predicting fuel consumption rates for the three diesel engines of R/V Gunnerus, enabling informed decision-making. By integrating LSTM models into the digital twin framework, operators can optimize vessel performance, reduce costs, and minimize environmental impact, thus advancing sustainability in the maritime industry.