Integrating Deficit Diagnostics, Suitability Modeling, and Location–Allocation for EV Charging Infrastructure

ABSTRACT In today's rapidly evolving transportation landscape, electric vehicles (EVs) play a key role in reducing dependence on fossil fuels. Developing accessible charging infrastructure is a critical priority for sustainable regional transport planning. This study introduces an integrated methodological framework for evaluating and optimizing the spatial distribution of electric vehicle charging stations (EVCSs) at the national scale, using the Republic of Croatia as a case study. The framework combines a composite EVCS Deficit Index (IDEV) derived using CRITIC‐weighted multi‐criteria analysis, a data‐driven suitability model (EVLP) based on logistic regression (LR) with spatially stratified cross‐validation, and network‐based location–allocation optimization under explicit coverage constraints. IDEV integrates nine indicators capturing supply intensity, accessibility, impedance, load, and demand pressure. Local Indicators of Spatial Association (LISA) were used to identify statistically significant clusters of high deficits at the local government unit (LGU) level. The EVLP suitability surface was derived from 16 predictors. The LR model achieved high predictive performance, and Top‐N evaluation confirmed strong prioritization capacity. Deficit‐priority LGUs and high‐suitability pixels were subsequently integrated into a constrained location–allocation model. Scenario analysis ( k = 10, 25, 50, 100 new EVCS) revealed diminishing marginal coverage gains, while simultaneously demonstrating measurable improvements in EVCS accessibility within deficit‐priority LGUs. The proposed framework moves beyond standalone suitability mapping or coverage optimization by linking spatial equity diagnostics, probabilistic modeling, and network allocation within a unified GIS workflow. The results provide a transferable methodological template for evidence‐based EVCS planning at national scale.