| 基于DInSAR-BP神经网络的震后区域滑坡危险性综合评价研究 |
| |
| 引用本文： | 朱崇浩,张建经,马东华,刘阳,向波.基于DInSAR-BP神经网络的震后区域滑坡危险性综合评价研究[J].工程地质学报(英文版),2020,28(3):530-540. |
| |
| 作者姓名： | 朱崇浩 张建经 马东华 刘阳 向波 |
| |
| 作者单位： | ①.西南交通大学土木工程学院, 成都 610031, 中国 |
| |
| 基金项目： | 国家重点研发计划2017YFC0504901四川省自然资源厅科研项目KJ-2019-5四川省交通运输科技项目2016BZ-2 |
| |
| 摘 要： | 地震对人类的威胁不仅是发生时直接造成的人员伤亡和财产损失,更体现在地震所产生的高隐蔽性、高危险性滑坡隐患体带来的危害,震后区域滑坡隐患体的快速识别和科学评价在震后抢险、排险工作中至关重要。以九寨沟地区"川主寺-九寨沟"公路沿线区域为研究对象,建立了基于DIn SAR-BP神经网络技术的震后区域滑坡危险性综合评价模型。研究结果显示,九寨沟地区震后的滑坡高危险性区域面积约为2602.35 km2,是震前的3.4倍,并且这些区域主要分布在震中东北方向约20 km附近、九寨沟景区内以及川九路前70 km,符合震后调查情况;使用多元非线性回归法可以有效计算震后地表形变值对滑坡危险性的影响,使震后危险性评价结果精度提高了13.9%,证明了模型在研究区域内具有良好的适用性。
|
| 关 键 词： | DIn SAR BP神经网络 滑坡危险性 滑坡隐患体 震后综合分析 |
| 收稿时间： | 2019-03-25 |
COMPREHENSIVE ANALYSIS ON RISK OF LANDSLIDES IN POST-EARTHQUAKE AREA BASED ON DINSAR-BP NEURAL NETWORKS |
| |
| ZHU Chonghao,ZHANG Jianjing,MA Donghua,LIU Yang,XIANG Bo.COMPREHENSIVE ANALYSIS ON RISK OF LANDSLIDES IN POST-EARTHQUAKE AREA BASED ON DINSAR-BP NEURAL NETWORKS[J].Journal of Engineering Geology,2020,28(3):530-540. |
| |
| Authors: | ZHU Chonghao ZHANG Jianjing MA Donghua LIU Yang XIANG Bo |
| |
| Institution: | ①.Civil Engineering, Southwest Jiaotong University, Chengdu 610031, China②.Sichuan Provincial Communications Department Highway Planning Survey Design Institute, Chengdu 610041, China |
| |
| Abstract: | The threats of earthquakes to humans are not only the casualties and property losses, but also the hidden danger of geological hazards. To assess the risk of landslides in the area after earthquakes, a quick and reliable model for the landslide risk assessment is necessary. Based on the study area along the "Chuanzhusi-Jiuzhaigou" road, we proposes a comprehensive analysis model for the landslide hazard in the post-earthquake area. It can be called the DInSAR-BP model. The model results show that the area of the high-risk landsides in Jiuzhaigou after the earthquake is about 2602.35 km2 and 3.4 times larger than that before the earthquake. And these areas are mainly distributed in the Jiuzhaigou scenic area and the slope of the first 70 km along the "Chuanzhusi-Jiuzhaigou" road. These results match the survey results after the earthquake. The Multivariate nonlinear regression can consider the impact from the earthquake to the risk of landslides, which improves the accuracy of post-earthquake risk assessment results by 13.9%. And it proves that the model has good applicability in the study area. |
| |
| Keywords: | DInSAR BP neural network Landslide hazard Hidden dangers of landslides Comprehensive analysis after earthquake |
| 本文献已被 CNKI 维普 等数据库收录！ |
| 点击此处可从《工程地质学报(英文版)》浏览原始摘要信息 |
| 点击此处可从《工程地质学报(英文版)》下载免费的PDF全文 |
