EUVIMEDEUVIMED entdecken
Crossref

A GIS-AHP and Machine Learning-Based Decision Support Framework for UAV Emergency Landing Site Selection in Urban Areas

Mehmet Nacar

Türkiye İnsansız Hava Araçları Dergisi · 17.11.2026 · (8)

DOI: 10.51534/tiha.1963797 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Türkiye İnsansız Hava Araçları Dergisi, online erschienen am 17.11.2026. Verfasst von Mehmet Nacar. Die Arbeit nennt 12 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Urban unmanned aerial vehicle operations require pre-defined emergency landing alternatives to reduce potential risks during abnormal flight conditions. This study proposes a Geographic Information System and Analytic Hierarchy Process-based decision support framework integrated with machine learning for identifying potential emergency landing zones for unmanned aerial vehicles in urban areas. Gaziantep city center, Türkiye, was selected as the case study area, and a 10 km radius analysis zone was divided into 200 m grid cells. OpenStreetMap data were used to derive spatial criteria, including building area ratio, open-space ratio, cemetery area ratio, distance to roads, distance to sensitive areas, distance to power infrastructure, distance to cemeteries, and distance to the city center. A GIS-AHP suitability score was calculated for each grid cell, and this score was modelled using Linear Regression, Support Vector Regression, Random Forest, Gradient Boosting, and XGBoost algorithms. The results showed that 153 of 8016 grid cells satisfied the candidate filtering criteria. After DBSCAN-based spatial clustering with a 400 m distance threshold, 46 independent candidate emergency landing zones were identified. Among the tested models, XGBoost achieved the best approximation of the AHP-derived suitability score with an R^2 value of 0.9916. Feature importance analysis indicated that building area ratio, distance to roads, distance to sensitive areas, and open-space ratio were the most influential variables. The proposed framework provides a low-cost and repeatable preliminary decision support approach for urban UAV emergency landing site assessment. However, the identified candidate zones should be validated through field inspection, high-resolution imagery, and UAV-specific operational safety analysis before practical use.

Zitieren

APA

Nacar, M. (2026). A GIS-AHP and Machine Learning-Based Decision Support Framework for UAV Emergency Landing Site Selection in Urban Areas. Türkiye İnsansız Hava Araçları Dergisi. https://doi.org/10.51534/tiha.1963797

Vancouver

Nacar M. A GIS-AHP and Machine Learning-Based Decision Support Framework for UAV Emergency Landing Site Selection in Urban Areas. Türkiye İnsansız Hava Araçları Dergisi. 2026. doi:10.51534/tiha.1963797

BibTeX
@article{Nacar2026,
  author = {Nacar, Mehmet},
  title = {A GIS-AHP and Machine Learning-Based Decision Support Framework for UAV Emergency Landing Site Selection in Urban Areas},
  journal = {Türkiye İnsansız Hava Araçları Dergisi},
  year = {2026},
  number = {8},
  doi = {10.51534/tiha.1963797}
}
Kostenlos starten →