Development of comprehensive models for precise prognostics of ship fuel consumption DOI
Thanh Tuan Le, Prabhakar Sharma, Nguyen Dang Khoa Pham

et al.

Journal of Marine Engineering & Technology, Journal Year: 2024, Volume and Issue: 23(6), P. 451 - 465

Published: July 2, 2024

This study incorporates two unique machine learning algorithms, Huber regression and Light Gradient Boosting Machines (LGBM), for estimating ship consumption of fuel. These methods are employed to create forecasting models fuel during journeys, which is especially useful when interacting with non-linear data. The then analyzes evaluates the prediction accuracy these approaches compared a baseline model generated using linear regression. results investigation show that both establish extremely accurate predictions while handling data quickly. However, Huber-based outperforms LGBM in terms accuracy, an R-squared value 0.979 versus 0.917 LGBM. In addition, has diminished error, RMSE 2.278, model's 4.55. graphical violin plot Taylor's diagram further established superiority ML. findings imply could be suitable option in-route usage real time. As consequence, this emphasises potential benefits accurately predicting consumption, providing encouraging possibilities optimise lowering greenhouse gas emissions.

Language: Английский

Leveraging Artificial Intelligence to Enhance Port Operation Efficiency DOI Creative Commons
Gia Huy Dinh, Hoang Thai Pham, Lam Canh Nguyen

et al.

Polish Maritime Research, Journal Year: 2024, Volume and Issue: 31(2), P. 140 - 155

Published: June 1, 2024

Abstract Maritime transport forms the backbone of international logistics, as it allows for transfer bulk and long-haul products. The sophisticated planning required this form transportation frequently involves challenges such unpredictable weather, diverse types cargo kinds, changes in port conditions, all which can raise operational expenses. As a result, accurate projection ship’s total time spent port, anticipation potential delays, have become critical effective activity management. In work, we aim to develop management system based on enhanced prediction classification algorithms that are capable precisely forecasting lengths ship stays delays. On both training testing datasets, XGBoost model was found consistently outperform alternative approaches terms RMSE, MAE, R2 values turnaround waiting period models. When used model, had lowest RMSE 1.29 during 0.5019 testing, also achieved MAE 0.802 0.391 testing. It highest 0.9788 0.9933 Similarly, outperformed random forest decision tree models, with greatest phases.

Language: Английский

Citations

4
Development of comprehensive models for precise prognostics of ship fuel consumption DOI
Thanh Tuan Le, Prabhakar Sharma, Nguyen Dang Khoa Pham

et al.

Journal of Marine Engineering & Technology, Journal Year: 2024, Volume and Issue: 23(6), P. 451 - 465

Published: July 2, 2024

This study incorporates two unique machine learning algorithms, Huber regression and Light Gradient Boosting Machines (LGBM), for estimating ship consumption of fuel. These methods are employed to create forecasting models fuel during journeys, which is especially useful when interacting with non-linear data. The then analyzes evaluates the prediction accuracy these approaches compared a baseline model generated using linear regression. results investigation show that both establish extremely accurate predictions while handling data quickly. However, Huber-based outperforms LGBM in terms accuracy, an R-squared value 0.979 versus 0.917 LGBM. In addition, has diminished error, RMSE 2.278, model's 4.55. graphical violin plot Taylor's diagram further established superiority ML. findings imply could be suitable option in-route usage real time. As consequence, this emphasises potential benefits accurately predicting consumption, providing encouraging possibilities optimise lowering greenhouse gas emissions.

Language: Английский

Citations

2