Random forest-based speed forecasting and carbon capture optimization for shipboard integrated energy systems

In response to the demand for low-carbon transformation in the shipping industry, this study proposes an integrated ship energy system that combines carbon capture and storage (CCS) with power-to-gas (P2G) technology. It develops a multi-energy coupling architecture incorporating liquefied natural gas (LNG), hydrogen, carbon dioxide, and diesel. Considering the significant impact of dynamic changes in ship speed on energy consumption and carbon emissions, this study innovatively introduces the Random Forest (RF) algorithm, utilizing historical speed data for high-precision prediction. The predicted speed load curve is used as a key input parameter and is incorporated into the Integrated Energy System(IES). Based on this, an optimal dispatch model for the IES is established,the coupling characteristics of electricity, heat, gas, and carbon are analyzed. Comparative analysis across different scenarios demonstrates that this model significantly enhances system flexibility while reducing carbon emissions and operational costs of the IES. Finally, a hierarchical carbon trading model incorporating carbon quota constraints is proposed to refine carbon trading cost modeling and analyze transaction prices. Simulation results indicate that, compared with traditional ship energy systems, the proposed approach reduces costs by 2.9%–9.5% and cuts carbon emissions by 12%–15%. This study provides a theoretical foundation for the low-carbon design and optimization of ship energy systems, offering critical insights for achieving the ”dual carbon” goals in the shipping industry. • Innovative multi-energy coupling architecture enables carbon-energy synergy. • Random forest algorithm achieves high-accuracy speed prediction for optimized dispatch. • Integrates CCS and P2G in ship IES for optimal renewable use. • Demonstrates CCS-P2G synergy in minimizing carbon emissions.