期刊文献+
共找到3篇文章
< 1 >
每页显示 20 50 100
基于KNN算法的隧道地表变形监测数据异常检测分析 被引量:5
1
作者 邹彤彤 杨涛 +5 位作者 周勇 王仲平 余来斌 王东凯 肖荣 出晓娟 《科学技术创新》 2021年第28期95-97,共3页
随着城市地下工程的不断增加,各种地面问题日趋严重。在隧道开挖中,由于受地下复杂的地质工程环境、周边条件等因素的影响,会产生渗漏水、裂缝、断面变形侵限等各种病害,造成人员伤亡及经济损失。所以对地表变形监测数据进行异常监测是... 随着城市地下工程的不断增加,各种地面问题日趋严重。在隧道开挖中,由于受地下复杂的地质工程环境、周边条件等因素的影响,会产生渗漏水、裂缝、断面变形侵限等各种病害,造成人员伤亡及经济损失。所以对地表变形监测数据进行异常监测是保证地基稳定性鉴别和预测、预警依据的必要前提。针对无数据标签的隧道地表变形监测数据的异常检测问题,本文提出了一种无监督式的k近邻(K-Nearest Neighbor,KNN)异常监测算法,并将其应用于岑溪至水汶高速公路岑溪大隧道的地表变形监测数据中,对23个监测点分别进行异常检测,并对不同区域的异常检测结果进行分析,确保监测结果真实有效。 展开更多
关键词 隧道地表变形数据 KNN 异常检测
下载PDF
顾及邻近点变形因素项的动态模型辨识及预测 被引量:16
2
作者 潘国荣 刘大杰 《测绘学报》 EI CSCD 北大核心 2001年第1期32-35,共4页
在动态建模预测中 ,为考虑其他因素 (如相邻点变形量 )对某点变形的影响 ,本文将AR模型推广到建立考虑邻近点变形因素项的自回归模型。最后以上海地铁二号线东方路站至东昌路站区间隧道相对盾构开挖面一定距离的某点地表变形数据为例 ,... 在动态建模预测中 ,为考虑其他因素 (如相邻点变形量 )对某点变形的影响 ,本文将AR模型推广到建立考虑邻近点变形因素项的自回归模型。最后以上海地铁二号线东方路站至东昌路站区间隧道相对盾构开挖面一定距离的某点地表变形数据为例 ,进行了建模及预测 ,并将结果与 AR模型作了比较 ,验证了模型的有效性。 展开更多
关键词 动态建模 预测 模型修正 隧道 地铁施工 地表变形数据 工程测量
下载PDF
Fractal Measures of Drainage Network to Investigate Surface Deformation from Remote Sensing Data: A Paradigm from Hindukush (NE-Afghanistan) 被引量:3
3
作者 Syed Amer MAHMOOD Richard GLOAGUEN 《Journal of Mountain Science》 SCIE CSCD 2011年第5期641-654,共14页
This approach represents the relative susceptibility of the topography of the earth to active deformation by means of geometrical distinctiveness of the river networks. This investigation employs the fractal analysis ... This approach represents the relative susceptibility of the topography of the earth to active deformation by means of geometrical distinctiveness of the river networks. This investigation employs the fractal analysis of drainage system extracted from ASTER Global Digital Elevation Model (GDEM-30m resolution). The objective is to mark active structures and to pinpoint the areas robustly influenced by neotectonics. This approach was examined in the Hindukush, NE-Afghanistan. This region is frequently affected by deadly earthquakes and the modern fault activities and deformation are driven by the collision between the northward-moving Indian subcontinent and Eurasia. This attempt is based on the fact that drainage system is strained to linearize due to neotectonic deformation. Hence, the low fractal dimensions of the Kabul, Panjsher, Laghman, Andarab, Alingar and Kocha Rivers are credited to active tectonics. A comprehensive textural examination is conducted to probe the linearization, heterogeneity and connectivity of the drainage patterns. The aspects for these natural textures are computed by using the fractal dimension (FD), lacunarity (LA) and succolarity (SA) approach. All these methods are naturally interrelated, i.e. objects with similar FD can be further differentiated with LA and/or SA analysis. The maps of FD, LA and SA values are generated by using a sliding window of 50 arc seconds by 50 arc seconds (50" × 50"). Afterwards, the maps are interpreted in terms of regional susceptibility to neotectonics. This method is useful to pinpoint numerous zones where the drainage system is highly controlled by Hindukush active structures. In the North-Northeast of the Kabul block, we recognized active tectonic blocks. The region comprising, Kabul, Panjsher, Andrab, Alingar and Badakhshan is more susceptible to damaging events. This investigation concludes that the fractal analysis of the river networks is a bonus tool to localize areas vulnerable to deadly incidents influencing the Earth’s topography and consequently intimidate human lives. 展开更多
关键词 FRACTAL Drainage network LACUNARITY Succolarity Surface deformation and Hindukush
下载PDF
上一页 1 下一页 到第
使用帮助 返回顶部