Coordinated Scheduling of Carbon Capture, Renewables, and Storage in Bulk Carriers: A Dual-Timescale LSTM-Powered Multi-Objective Energy Management System Strategy

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:Provides a novel integrated framework combining LSTM-based fuel prediction, multi-objective optimization (NSGA-II), and CCS scheduling for ship energy systems; valuable for maritime decarbonization research.

🏢実務担当者:Shipping companies can adopt the proposed scheduling strategy to reduce fuel costs and emissions; the 6-year payback suggests clear economic incentive for PV and battery investment.

🏛政策担当者:Demonstrates that CCS on ships is technically and economically feasible under certain conditions, supporting IMO and national policies promoting maritime decarbonization.

To address the challenges of energy conservation and emission reduction in the shipping industry, this study proposes an innovative scheduling strategy for the ship integrated energy system (SIES) based on data-driven fuel consumption prediction and multi-objective optimization. A multi-feature dual-time scale Long Short-Term Memory (LSTM) network is developed, integrating Automatic Identification System (AIS) data with an average resolution of 6 min, meteorological conditions, and vessel state parameters, achieving fuel consumption prediction across dual time scales. The model outperforms other machine learning models (e.g., CNN, XGBoost) in terms of R2, MAE, RMSE, and SMAPE. Dynamic simulation of annual cooling, heating, and power loads for crew accommodation areas, based on spatiotemporally matched customized meteorological data, reveals that the annual load is dominated by cooling demand, with significant seasonal fluctuations; summer loads are higher and more volatile than winter loads. A hybrid energy system integrating photovoltaic (PV) generation, energy storage, carbon capture and storage (CCS), and diesel engines is constructed. By treating the CCS load as a adjustable resource, the Non-dominated Sorting Genetic Algorithm (NSGA-II) is employed to solve the environmental–economic multi-objective optimization problem, simultaneously minimizing carbon emissions and present value of the total cost (PVC). Case studies conducted on a 79,970 DWT bulk carrier (Guangzhou–Qinhuangdao route) demonstrate the strategy’s effectiveness. The synergistic operation of solar energy and the energy storage system facilitates carbon emission reductions of 23.6% to 40.0% through fuel savings; during summer with abundant solar resources, over 95% of the CCS load can be covered. Economic analysis indicates that fuel savings from renewable energy can recover the investment in the PV and battery storage system within approximately 6 years. This integrated data-driven energy management framework mitigates CCS-induced parasitic loads and emissions, partially resolving the “carbon emissions vs. cost” dilemma, and provides a viable pathway for decarbonizing conventional diesel-powered ships, contributing to sustainable maritime operations.

• semanticscholar https://doi.org/10.3390/en19041010first seen 2026-06-29 06:24:23