Multi-stage adaptive expansion of EV charging stations considering impacts from the transportation network and power grid

The increasing popularity of electric vehicles (EVs) necessitates the expansion of existing EV charging stations to meet rising demand. This paper tackles the complex challenge of accurately estimating charging demand over time while considering the impacts of transportation networks and power grids. We first develop a traffic assignment model and design an algorithm based on the Frank-Wolfe method to calculate user equilibrium traffic flow for each link in the transportation network. By leveraging this flow, we enhance demand estimation through a mapping function that provides more precise estimates of EV charging demand. Recognizing the uncertainties in charging demand, we formulate a multi-stage robust adaptive expansion model for EV charging stations to allow for adjustments as demand evolves. This model aims to maximize the total profit of charging station operators by determining the optimal number of charging piles across multiple stages. To solve this optimization problem, we design an algorithm based on the partition-and-bound method with finite adaptability to effectively balance computational efficiency with optimization performance. Extensive case studies validate the effectiveness of our models and algorithms, demonstrating their applicability in real-world scenarios and underscoring the need for coordinated planning to support the growing EV infrastructure.