|
A Machine Learning-Based Hybrid Model for Forecasting Electricity Demand in Offshore Cold Ironing, Focusing on Environmental Sustainability in Maritime Operations
|
Meead Mansoursamaei   , Mahmoud Moradi * , Keikhosro Yakideh |
|
|
|
Abstract: (120 Views) |
Maritime logistics, as one of the pillars of global trade, faces challenges such as port congestion and environmental impacts. This paper investigates the use of green electric energy to power vessels outside port terminals through a method known as offshore cold ironing. This study develops an advanced machine learning approach, a hybrid stacking ensemble model, to forecast electricity demand. The focus of this research is on a maritime port, and it evaluates various machine learning models within a parallel stacking approach, demonstrating their effectiveness in forecasting the energy consumption of vessels outside the port compared to individual models. The aim of this research is to enhance sustainability in maritime energy consumption by examining the efficient transfer of renewable energy in offshore scenarios. This includes eliminating the use of fossil fuels by vessels and transitioning to renewable energy sources through accurate electricity demand forecasting. The results show that the MLP-GB ensemble model outperforms other individual or ensemble models in achieving lower root mean squared error, highlighting its potential for accurate forecasting. The findings of this research point to the critical role of accurate forecasting in promoting the replacement of maritime energy sources and contributing to global efforts to combat climate change and reduce greenhouse gas emissions in the maritime sector. |
Article number: 6 |
| Keywords: Maritime logistics, Offshore cold ironing, Machine learning, Forecast, Electricity demand |
|
|
Full-Text [PDF 960 kb]
(67 Downloads)
|
Type of Study: Research |
Subject:
Special
Received: 2024/06/6 | Accepted: 2024/11/7 | Published: 2025/06/18
|
|
|
|
|
|
|
| Send email to the article author |
|
|
|
| Add your comments about this article |
|
|
|
