光储充一体化充电站选址、定容与服务分配协同优化研究

Research on the Collaborative Optimization of Siting, Sizing, and Service Allocation for Integrated Photovoltaic-Energy Storage-Charging Stations

摘要: 在“双碳”战略背景下，电动汽车的迅猛发展对城市配电网构成挑战。光储充一体化充电站作为一种能够就地消纳可再生能源、缓解电网冲击的有效解决方案，其科学规划至关重要。然而，现有研究较少将场站选址、各子系统容量配置及服务分配在统一的确定性框架下进行协同优化。针对此问题，本文构建了一个混合整数线性规划(MILP)模型。该模型以年化总社会成本最小化为目标，在一个统一的框架内对充电网络的选址、各子系统包括光伏、储能和充电容量以及服务范围进行协同优化。模型全面考虑了充电站功率平衡、储能系统物理特性以及与电网的交互约束。以上海市嘉定区为例的实证分析表明：该确定性模型具备高效的求解性能，能够在短时间内为城市级规划问题提供最优解。决策结果揭示了在理想条件下，为最大化用户便利性，模型倾向于进行广覆盖的站点布局，并根据各区域负荷特性进行高度差异化的容量配置。本研究不仅为确定性环境下的新型充电基础设施规划提供了一套科学的理论方法与决策基准，也为后续考虑不确定性风险的复杂规划研究奠定了基础。

Abstract: Under the “Dual Carbon” goals strategy, the rapid development of electric vehicles poses significant challenges to urban power distribution networks. Integrated photovoltaic-energy storage-charging (PSC) stations, which effectively consume local renewable energy and mitigate grid impact, require scientifically sound planning. However, existing research seldom addresses the collaborative optimization of station siting, subsystem capacity sizing (for photovoltaic, energy storage, and charging facilities), and service allocation within a unified deterministic framework. To address this gap, this paper develops a mixed-integer linear programming (MILP) model. Aiming to minimize the total annualized social cost, the model collaboratively optimizes the siting of charging stations, the capacity sizing of subsystems—including photovoltaic, energy storage, and charging infrastructure—and service allocation within a unified framework. It comprehensively incorporates constraints such as station power balance, the physical characteristics of the energy storage system, and interaction with the main grid. An empirical analysis conducted in Jiading District, Shanghai, demonstrates that this deterministic model achieves high computational efficiency, providing optimal solutions for city-scale planning problems within a short timeframe. The results indicate that, under ideal conditions, to maximize user convenience, the model favors a widely distributed station layout with highly differentiated capacity configurations tailored to the load characteristics of each zone. This study not only provides a scientific theoretical method and a decision-making benchmark for planning new charging infrastructure in deterministic environments but also establishes a foundation for future research incorporating uncertainty and risk.