Abstract:

Earthquakes are prone to causing building damage and casualties. While satellite remote sensing fails to deliver sufficient data accuracy, unmanned aerial vehicles (UAVs) have emerged as a pivotal tool for post-earthquake rescue information search, leveraging their rapid response, high mobility, and the precise data collection capability of onboard equipment. To address the existing issues such as over-reliance on manual experience for UAV search routes and low data collection efficiency, this study develops a rapid search route planning method for post-earthquake UAVs. A four-stage technical framework is proposed: "high-value POI screening - spatial clustering zoning - rescue priority evaluation - ant colony algorithm-based path optimization". Six categories of high-value POI data are selected, a two-layer clustering algorithm is employed for zoning and calculating the required number of UAVs, a multi-index priority model is constructed, and the ant colony algorithm is introduced to optimize routes. Taking Fengnan District of Tangshan City as the research area, 1,443 high-value POIs across six categories from December 2024 are selected for experiments using DJI M300RTK UAVs. The results show that the two-layer clustering algorithm reduces blank areas by 61.6%, and 4 constraint-compliant routes are generated for Zones 0-4. Compared with the case of similar area in Jishishan, Gansu Province, the number of flight sorties is reduced by 13. This method can generate efficient UAV flight routes and provide robust support for post-earthquake UAV rescue information search operations.

Key words: information technology, post-earthquake information search, spatial clustering, ant colony algorithm, low-altitude economy