A power-traffic graph embedding distributed energy storage planning model for energy-efficient and reliable urban rail transit

The escalating energy demands and limited flexibility of urban rail transit systems to integrate renewable generation impose significant economic and carbon burdens. The interdependencies of power and transportation system presents critical challenges for planning. This paper proposes a novel graph-based power-traffic planning framework to develop low-carbon, reliable urban rail transit by capturing spatiotemporal dynamics and cross-network interactions. First, power and traffic subsystem models establish topology and flow characteristics through graph-theoretic representations. Hypergraph-based modeling precisely reflects structural and functional dependencies between transportation and energy infrastructures. Then, a bi-level planning model of urban rail transit with normal and emergency scenarios is proposed to determine the locations and sizes for distributed energy storage systems. The proposed model can not only consider the initial investment and operation costs but also consider the expected response cost to emergency scenarios that are generated in line the traffic flow levels, power load levels, and renewable-based power generation levels. Furthermore, the robust optimization with bounded uncertainty sets addressing renewable and load variability. Finally, the results of a 6-line 114-station network demonstrates over 30% cost savings and significant carbon emission reductions, proving the necessity of subsystem interdependency considerations. The framework establishes new foundations for resilient, efficient electrified transit planning.