Integrating user preferences and demand uncertainty in electric micro-mobility battery-swapping station planning: A data-driven three-stage model

Abstract

With the rapid adoption of electric micro-mobility vehicles (EMVs), the demand for efficient, user-centered battery-swapping infrastructure is rising. However, existing battery-swapping station (BSS) planning often falls short by neglecting critical elements such as user preferences and demand uncertainty. This study introduces a three-stage BSS planning framework that holistically addresses demand allocation, location-capacity optimization, and deployment adaptability under fluctuating demand. First, EMV users' preferences are integrated into a demand allocation model, capturing range anxiety and individual station selection criteria. This demand-sensitive allocation then informs a multi-objective bi-level planning model, balancing construction costs with user travel distances to BSS facilities. Finally, the model incorporates a demand uncertainty layer, supported by simulation scenarios, to create a robust facility deployment strategy that anticipates various demand fluctuations. An improved Non-dominated Sorting Genetic Algorithm II (NSGA-II) effectively optimizes this model, validated through a case study in Nanjing, producing 41 Pareto-efficient solutions. Sensitivity analysis highlights how factors like range anxiety and charging time impact BSS planning outcomes, underscoring the value of this data-driven approach. This work demonstrates that a comprehensive planning approach, integrating user behavior and uncertainty considerations, can significantly enhance the effectiveness and adaptability of EMV battery-swapping networks.