Prioritizing Urban Bus Electrification with Route-Level Energy Profiling: Two-Stage Decision Framework

Abstract The transition to electric buses (EBs) features prominently in urban transport decarbonization strategies. Yet, existing EB deployment plans usually follow first-available or budget-driven approaches, overlooking substantial route-level variations in energy consumption. This paper develops a two-stage, data-driven framework for prioritizing EB deployment: (1) ranking routes for electrification; and (2) specifying the EB coverage level on each route. Based on 5 days of high-resolution global positioning system (GPS) trajectories and 5.2 million smart-card records from 185 bus routes in Xi’an, China, this study extracts a variety of energy-consumption indicators covering vehicle dynamics, route attributes, and operational factors. A gradient boosting decision trees model identifies the relative influence of each indicator, and its feature importance feeds a multicriteria decision-making module (MOORA + TOPSIS) that ranks routes for electrification. The analysis reveals that high-priority routes reside in high-demand, pollution-sensitive core districts, while low-priority routes lie in dispersed suburban areas with weak charging support. Partial-dependency analysis then informs the appropriate operational coverage level for EBs on different routes. The resulting strategy fully electrifies roughly the top one-third of routes and assigns tiered EB coverage levels of 90%, 80%, and 70% to the remainder. This research provides a practical pathway for transit operators to replace diesel buses with EBs, enabling more targeted bus fleet management to conserve energy.