A Markov Chain-Based SDDiP Method for Integrated Logistics and Hydrogen-Electric Energy Scheduling for Seaports

This article proposes a multistage stochastic programming framework for the integrated day-ahead and intraday scheduling of hydrogen–electric energy and logistics systems. This framework establishes a deep cyber-physical coupling mechanism by unlocking the demand–response potential of reefer thermal dynamics and leveraging hydrogen infrastructure as a controllable industrial buffer to synchronize power supply with port operations. To address temporal uncertainties in ship arrivals and renewable generation, a Markov chain-based stochastic dual dynamic integer programming algorithm is developed. By leveraging finite-state Markovian transitions and cut families, the proposed algorithm overcomes the curse of dimensionality, ensuring computational tractability and high solution quality for large-scale problems. Case studies based on Ningbo–Zhoushan Port real-world data demonstrate significant techno-economic benefits. The proposed coordination achieves a 33% reduction in total operating costs, a 30% cut in peak demand, and a 51% abatement in emissions compared to uncoordinated operations. Comparative benchmarks validate the method's robustness in dampening disturbance propagation and reducing cost volatility. Sensitivity analysis confirms the structural necessity of the hybrid architecture, where an optimized automated guided vehicle ratio balances charging latency against energy costs. Furthermore, results demonstrate that sufficient hydrogen capacity transforms the system into an active flexibility resource, establishing the framework as a prerequisite for energy-autonomous seaports.