Two-layer deep reinforcement learning based port energy management strategy considering transportation-energy coupling characteristics

The coupling between energy and logistics systems in port microgrids necessitates an integrated energy optimization management strategy. In the existing literature, port microgrid energy management problems are typically formulated as mixed-integer programming models. However, their effectiveness is limited by uncertainties in renewable energy sources, as well as ship arrival times and demands, and they lack real-time adjustability. To address these challenges, a two-layer deep reinforcement learning (DRL)-based energy management strategy is proposed, incorporating transportation-energy coupling characteristics and battery dispatching behavior. Firstly, an integrated energy management problem is formulated, which encompasses berth allocation, energy dispatch, and battery swapping station scheduling to enhance the synergy between logistics and energy systems. Secondly, the optimization problem is reformulated as a Markov decision process (MDP). Finally, the proposed two-layer DRL-based optimization framework is employed to solve this energy management problem with hybrid and variable action spaces introduced by berth allocation in a real-time manner. The simulation results indicate that the proposed method can reduce operating costs and demonstrate the superiority and scalability of the algorithm when compared to other methods.