Optimizing electric vehicle charging infrastructure deployment in nordic regions: A hybrid GIS and K-means approach

The strategic deployment of electric vehicle charging stations (EVCS) is essential for accelerating transport decarbonization, particularly in regions characterized by sparse populations, long travel distances, and severe winter conditions. This study proposes a hybrid methodology for EVCS location planning in the Nordic context by integrating a weighted K-means clustering algorithm with GIS-based feasibility screening. The method combines demand-driven optimization with practical spatial constraints, including road accessibility, land-use suitability, inter-station spacing, grid proximity, and cold-climate service considerations. A penalty term is introduced to reduce low-utilization bias and improve centroid stability by limiting the influence of weak-demand outliers on candidate station placement. The workflow is demonstrated through a synthetic but geographically representative Nordic case study using six principal urban nodes and corridor-based demand patterns. Results show that the baseline K = 6 configuration provides an efficient demand-oriented network, while a sensitivity test with K = 7 reduces the largest northern service gap at the expense of slightly lower overall system efficiency. Beyond methodological development, the study provides practical implementation guidance by showing how candidate sites can be refined through GIS layers and ranked according to real-world deployment criteria. The proposed framework is transparent, adaptable, and suitable for planning EV charging infrastructure in cold-region transport systems where both operational efficiency and spatial accessibility are critical.