Original scientific paper

https://doi.org/10.17559/TV-20251020003077

Research on Planning Model of Urban and Rural Emergency Shelters Based on DBSCAN Spatial Clustering Algorithm

Jingkai Zhou ; School of Environmental and Chemical Engineering, Shanghai University, Shanghai, 200444, China *
Ying Sun ; The Control Centre of the "Henan Province People's Victory Cana", Henan Xinxiang, 453000, China

* Corresponding author.

Abstract

To optimize the clustering effect, this study proposes an improved strategy of the DBSCAN algorithm based on the core focus. The innovation of this method lies in: taking the core object as the coordinate origin, the points within its neighborhood are divided into different quadrant regions. Subsequently, based on the distribution of the number of adjacent points within each quadrant and their spatial deviation from the core object, the centroids of the points within each quadrant are calculated and selected as representative points. Ultimately, these representative points are utilized to drive the expansion process of the clustering clusters. This improvement aims to capture the directional distribution characteristics of spatial point sets more precisely and enhance the accuracy of cluster pattern recognition. In view of the particularity of the location selection of emergency shelters and evacuation sites and their positioning in the urban planning process, this study constructs a planning model for urban and rural emergency shelters based on the DBSCAN spatial clustering algorithm for urban planning application scenarios. Based on the existing research framework, this model systematically integrates key elements such as the quantity and scale of shelters, spatial location selection, service path length limitations, and population capacity carrying capacity, and establishes a mathematical model covering multiple optimization objectives and constraints. A set of algorithm processes centered on iterative calculation was designed for model solving. Especially with the support of the network space analysis function provided by the DBSCAN algorithm, the theoretical service coverage range of each candidate refuge was accurately calculated, thereby scientifically evaluating the rationality and superiority of their spatial layout schemes.

Keywords

DBSCAN spatial clustering algorithm; site selection model refuge and evacuation site; urban and rural emergency; urban planning multi-condition constraints

Hrčak ID:

353073

URI

https://hrcak.srce.hr/353073

Publication date:

7.11.2026.

Visits: 0 *