1. Planning approach for integrating charging stations and renewable energy sources in low-carbon logistics delivery.
- Author
-
Wang, Jiawei, Guo, Qinglai, and Sun, Hongbin
- Subjects
- *
DEEP reinforcement learning , *REINFORCEMENT learning , *RENEWABLE energy sources , *ENERGY storage , *SUSTAINABLE development - Abstract
To achieve green and low-carbon development in the logistics industry, logistics operators are promoting the electrification of logistics fleets, which imposes requirements for well-developed charging facilities and integrated renewable energy sources. Due to the specific characteristics of logistics activities, which are different from those of normal vehicles, the infrastructure planning for logistics delivery needs to be considered in coordination with the logistics system's operation. The coupling of planning and operation poses challenges for decision-making. This paper presents a planning–operation coupling optimization framework for low-carbon logistics delivery. The planning level optimizes the location and capacity of charging facilities, photovoltaic (PV), and energy storage systems (ESSs) based on the idea of charging demand matching. The operation level uses deep reinforcement learning (DRL) to simulate the logistics fleet's action patterns, optimize routes and charging behaviors, and extract charging demands. Benefiting from the advantages of DRL, planning and operation can interact well through charging demands, mutual coupling, and iterative adjustment. Numerical experiments based on real-world data show that the proposed framework effectively reduces charging costs and carbon emissions and also provides a benchmark for the cost evaluation of carbon reduction in logistics delivery activities. • A joint planning–operation framework suits the properties of logistics delivery. • The infrastructure planning model adopts the idea of charging demand matching. • DRL is used to extract charging demands and interact with the planning model. • Numerical experiments with real-world city data validate the method's effectiveness. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF