地震2024,Vol.44Issue(4) :153-172.DOI:10.12196/j.issn.1000-3274.2024.04.010

基于U-Net++的建筑物抗震能力评价模型

Seismic Capacity Evaluation Model for Buildings Based on U-Net++

侯红语 窦爱霞 郭红梅
地震2024,Vol.44Issue(4) :153-172.DOI:10.12196/j.issn.1000-3274.2024.04.010

基于U-Net++的建筑物抗震能力评价模型

Seismic Capacity Evaluation Model for Buildings Based on U-Net++

侯红语 1窦爱霞 1郭红梅2
扫码查看

作者信息

  • 1. 中国地震局地震预测研究所,北京 100036
  • 2. 四川省地震局,四川成都 610041
  • 折叠

摘要

针对传统建筑物抗震评估耗时耗力、难以大范围展开的问题,本文提出了一种基于U-Net++的建筑物抗震能力评价模型.该模型以U-Net++为骨干网络,集成了多种信息源,包括高分辨率遥感影像、建筑物建造年代、高度、用途和结构类型.以四川省雅安市雨城区为研究区,设计了 17个对比实验探究不同特征对建筑物抗震能力的影响程度,为广域地震灾害风险精细化评估提供了有力的技术支撑.

Abstract

Aiming at the problem that the traditional seismic evaluation of buildings time-consuming and the difficulty in a large-scale implementation,this paper proposes a seismic capacity evaluation model for buildings based on U-Net++.The model takes U-Net++as the backbone network,integrates multiple information sources,such as high-resolution remote sensing images,building construction age,height,use and structural types.Tak-ing the Yucheng District of Ya'an City,Sichuan Province,as the study area,17 compara-tive experiments were designed to explore the degree of influence that different features have on the seismic resilience of buildings.This work provides robust technical support for refined assessments of earthquake disaster risks on a broad scale.

关键词

遥感影像/建筑物/抗震性能/语义分割/U-Net++

Key words

Remote sensing images/Buildings/Seismic capacity/Segmentation/U-Net++

引用本文复制引用

出版年

2024
地震
中国地震局地震预测研究所 中国地震学会地震预报专业委员会 中国地震学会地震流体专业委员会 中国地震学会地震电磁学专业委员会

地震

CSTPCD北大核心
影响因子:0.733
ISSN:1000-3274
段落导航相关论文