共查询到20条相似文献，搜索用时 15 毫秒
1.
《山地科学学报》2021,(4)
In this study, Bayesian probability method and machine learning model are used to study the real occurrence probability of earthquake-induced landslide risk in Taiwan region. The analyses were based on the 1999 Taiwan Chi-Chi Earthquake, the largest earthquake in the history in this Region in a hundred years, thus can provide better control on the prediction accuracy of the model. This seismic event has detailed and complete seismic landslide inventories identified by polygons, including 9272 seismic landslide records. Taking into account the real earthquake landslide occurrence area, the difference in landslide area and the non-sliding/sliding sample ratios and other factors, a total of 13,656,000 model training samples were selected. We also considered other seismic landslide influencing factors, including elevation, slope, aspect, topographic wetness index, lithology, distance to fault, peak ground acceleration and rainfall. Bayesian probability method and machine learning model were combined to establish the multi-factor influence of earthquake landslide occurrence model. The model is then applied to the whole Taiwan region using different ground motion peak accelerations(from 0.1 g to 1.0 g with 0.1 g intervals) as a triggering factor to complete the real probability of earthquake landslide map in Taiwan under different peak ground accelerations, and the functional relationship between different Peak Ground Acceleration and their predicted area is obtained. 相似文献
2.
风险分析与评估是解决边坡固有不确定性的重要工具, 但同时考虑外在荷载和内在岩土力学参数的不确定性, 对边坡进行系统定量风险分析的研究较少。以西藏扎拉水电站厂后倾倒变形边坡为例, 基于场地地震峰值加速度概率密度函数和不同地震峰值加速度下边坡失稳概率拟合函数, 采用数值积分计算了边坡在设计基准期的失稳概率, 并采用离散元方法对边坡失稳后的影响范围进行了数值模拟, 在此基础上进行了承灾体易损性分析及定量风险计算, 最后采用ALARP准则进行了风险评价。研究表明, 考虑地震危险性条件下, 扎拉水电站厂后倾倒变形边坡在50 a设计基准期内失稳概率为0.061 9;边坡对水电站地面厂房存在较大威胁, 相应财产风险为5 482万元; 根据ALARP准则, 边坡风险处于不可接受区, 需采取措施防范或规避风险。研究成果对于边坡治理工程决策及风险管理具有指导意义。 相似文献
3.
Cracks in rock or soil slopes influence the stability and durability of the slopes. Seismic forces can trigger slope disasters, particularly in the cracked slopes. Considering the nonlinear characteristics of materials, the more generalized nonlinear failure criterion proposed by Baker is adopted. The influence of non-dimensional strength parameters on the stability of cracked slopes under earthquakes is performed using the upper bound limit analysis. The seismic displacement is calculated by adopting the logarithmic spiral failure surface according to the sliding rigid block model. Based on the existing studies, two methods for the stability analysis of cracked slopes under earthquakes are introduced: the pseudo-static method(with the factor of safety(Fs) as an evaluation index), and the displacement-based method(with the seismic displacement as an evaluation index). The pseudo-static method can only determine the instantaneous stability state of the cracked slope, yet the displacement-based methodreflects the stability variation of cracked slopes during earthquakes. The results indicate that the nondimensional strength parameters affect the factor of safety and seismic displacement of slopes significantly. The non-dimensional strength parameter(n) controlling the curvature of strength function shapes on the slope stability is affected by other parameters. Owing to cracks, the effect of non-dimensional strength parameters on seismic displacement becomes more significant. 相似文献
4.
In recent major earthquakes, the researchers have found the need for consideration of vertical seismic acceleration for the stability analysis of the man-made and natural slopes. However, in most past studies, the performance of slopes has been assessed by accounting only the horizontal seismic component of the ground motion, without giving due weightage to the effect of vertical component. In the present study, analytical expressions are derived to determine the factor of safety, yield seismic coefficient and consequently the seismic displacement of cohesionless soil slope under combined horizontal and vertical components of the ground motion. The derivation uses the Newmark’s sliding block approach, in which the soil slope with a planar failure surface within the framework of conventional pseudo-static analysis is assumed to follow the Mohr-Coulomb failure criterion. The effects of vertical seismic coefficient on the stability of cohesionless slope have been studied through a set of graphical presentations for a specific range of soil parameters. It is observed that overlooking the effect of the vertical component of the ground motion on factor of safety and the displacement while designing the slope may be detrimental, resulting in the slope failure. The general expressions presented in this paper may be highly useful in the field of earthquake geotechnical engineering practice for designing the cohesionless soil slopes under combined horizontal and vertical seismic loads. 相似文献
5.
利用机器学习模型进行滑坡易发性评价时, 不同的超参数设置往往会导致评价结果的不同。采用贝叶斯算法对4种常见机器学习模型(逻辑回归LR、支持向量机SVM、人工神经网络ANN和随机森林RF)的超参数进行了优化, 探索了该算法对滑坡易发性机器学习模型的优化效果。以湘中地区4县(安化县、新华县、桃江县和桃源县)滑坡易发性评价为例说明该算法的可行性与适用性。基于滑坡历史编录, 确定研究区内1 017个滑坡点, 并选定15个滑坡影响因子, 以此构建滑坡易发性模型的训练集和测试集。利用贝叶斯优化算法对4种机器学习模型的主要超参数进行了优化, 依据优化后的超参数建立了4种优化模型, 并使用AUC值等指标来比较其预测能力。结果表明: 经超参数优化后的4种机器学习模型预测性能均有所提高, 且基于贝叶斯优化的随机森林模型表现最好。 相似文献
6.
针对已有的地震检测机器学习模型训练所需的有标注的地震数据难以获取、处理成本高等问题,提出一种适用于小样本情况的地震检测机器学习新模型CCLSN。该模型联合使用连续小波时频变换和重新设计的轻量化卷积神经网络,可大幅降低训练所需的有标注的地震数据量,提高模型的适用性。实验结果表明,CCLSN仅用包含数百个地震样本的小型数据集,就可实现稳定和高精度的识别功能,精度和召回率均在98%以上。CCLSN可为我国中东部等少震弱震地区实现地震自动检测提供新的技术方法。 相似文献
7.
Interest in rock slope stability in mountainous regions has increased greatly in recent years.This issue has become a topic of major interest for geoscientists and engineering professionals,as well as for private citizens and local administrators,in many parts of the world.This paper evaluates the stability of seven rock slopes along the KamyaranMarivan tourist road,Kurdistan province,Iran,using various methods.The two main reasons for performing this research were to determine whether different methods of stability analysis provide the same results,and to determine how different factors such as the presence of water,tension cracks,and seismic forces affect the stability of these rock slopes.Firstly,field investigations were performed to obtain the engineering characteristics of the rock masses,discontinuities,and intact rocks of the slopes.Secondly,laboratory tests were carried out on rock samples obtained from the slopes,to determine the engineering properties of the intact rocks.Then for each rock slope,the contour diagram of discontinuities and slope face was drawn in the Dips v.5.1 software environment,and the failure mechanism was determined based on the kinematic or stereographic method.Next,the factors of safety of the rock slopes were calculated using the limit equilibrium method,based on the failure mechanisms resulting from the kinematic method.The accuracy of the results obtained by these two methods was investigated using SWedge v.4.0 software.The results indicated that four rock slopes have a potential for plane,wedge,and toppling failure,and three others are stable.Also,it was found that the stability of the studied rock slopes decreases greatly in the presence of water,tension cracks,and seismic forces. 相似文献
8.
PS测井技术是地震勘探方法之一,也是一种简便、快速、准确的原位测试技术。该文介绍了单孔法PS测井的原理及其在铁路工程勘察中的应用,主要包括根据等效剪切波,进行建筑抗震场地类别划分,铁路工程抗震场地类别划分。根据岩土动力学参数,达到评价岩体质量和划分围岩类别目的,利用剪切波速法估算岩土的承载力基本值。通过青岛机场隧道场地的工程实例说明了PS测井技术在岩土工程勘察设计中的应用情况及应用效果。 相似文献
9.
????????趨????????????????????????Σ??????????????????????????????????????????о???????????趨?????????????°???????????????????????趨????λ??????????????????????????? 相似文献
10.
本文以秦皇岛市青龙县城工程场地为例，采用Ａ·Ｄｅｒ·Ｋｉｕｒｅｇｈｉａｎ和Ａ·Ｈ－ＳＡｎｇ断层－破裂模型，通过潜在震源识别及震源区划分、地震动参数确定、地震动衰减公式计算模型建立及不确定性校正、场地地震危险性评定等思路和方法，对该场地进行了地震危险性分析和评定。结果认为，该场地建筑物按５０年使用期，地震烈度为６．７２度，其地面峰值加速度为０．１０３ｇ；一般建筑物按Ⅵ度烈度设防，重要建筑则按Ⅶ度烈度设防；可能产生的Ⅷ度烈度主要来自西南方向的唐山地震区。 相似文献
11.
随着大数据和以深度学习为基础的人工智能技术的快速发展,油气藏地质建模逐步从传统的两点地质统计建模、基于目标建模、多点地质统计建模和基于沉积过程建模进入智能地质建模阶段。以深度学习为基础的智能地质建模主要采用对抗生成网络建立三维地质模型,目前这些研究集中在网络结构和算法的完善,特别是对地震和测井等各类数据的条件化,少量研究侧重于样本数据的获取。目前研究中采用的训练样本大多是基于目标或基于沉积过程方法模拟得到的合成数据,为了真正将该技术应用实际地下油气藏,需要更加关注真实样本数据的获取。仅靠深度神经网络这种统计学习方法实现技术突破的难度较大,研发通用的人工智能地质建模器是未来的主要发展方向,其中统计学习与符号学习相结合可能是实现该技术的必经道路。 相似文献
12.
13.
突发性地质灾害危险性评估对灾害防治与风险管理具有重要意义。由于不同地区影响灾害发生的因子各不相同,实际评估过程中难以全面客观地选取适宜的评估因子。机器学习对处理灾害系统的高维非线性问题独具优势,但因模型难以调优而评估效果有限。本文尝试提出一种双向优化的滑坡危险性评估方法：在构建因子敏感性指数开展定量敏感性分析的基础上,结合重要性分析、相关性分析、共线性分析构建四维(Four-Dimensional, 4D)特征筛选法用于评估因子综合优选;为克服模型难以调优的问题,引入差分进化(Differential Evolution, DE)算法优化支持向量机(Support Vector Machine, SVM)与多层感知机(Multi-Layer Perceptron, MLP) 2种推广能力较强的机器学习模型。最后,以福建省滑坡为例,开展评估方法研究。研究表明：4D特征筛选法能更加客观全面地选取适宜性更高的危险性评估因子,从而降低数据维度、减少信息冗余以提升评估模型性能;DE算法对SVM与MLP具有显著的优化效果,有益于增强模型滑坡危险性的评估准确度,DE-SVM、DE-MLP相较于未优化前模型的AUC值分别提升了4.43%与4.37%;基于双向优化的滑坡危险性评估结果表明,降雨与土地利用类型对福建省滑坡发生具有重要影响作用,福建省滑坡极高危险区普遍年均降雨较高、地形复杂多变,极低危险区主要位于东南沿海一带及闽江流域两侧。本研究为滑坡危险性评估中的影响因子客观选取与机器学习模型调优提供了一定思路。 相似文献
14.
����Bayesian����BP�������GPS�߳�ת�� 总被引：2，自引：0，他引：2
?????BP?????????????????????????????????????Bayesian?????????BP??????????????????????????GPS??????????????????L-M????????????????????????Bayesian??????BP???????????????????????????????????? 相似文献
15.
针对高空间分辨率遥感影像目标提取中定位精度低、边缘粗糙等问题,提出一种融合目标边缘特征与语义信息的人工坑塘提取网络模型。方法首先利用改进的U-Net语义分割网络模块来提取遥感影像中丰富的目标语义信息,然后拓展上述语义分割网络构建边缘提取子网络来获取遥感影像的多尺度边缘特征,最后借助于编码-解码子网络融合边缘特征与语义信息,实现遥感影像目标的精准提取。将该方法运用到雷州半岛复杂背景条件下人工坑塘提取实验中,实验结果中本文提出的方法在F分数以及边界F分数等评价指标上表现最优,达到97.61%与83.01%,验证了融合高层语义信息结合低层的边缘特征在提升遥感目标提取精确度上的有效性。 相似文献
16.
现有的面实体多指标几何匹配方法在计算综合相似度和确定最终匹配实体时面临着指标权重和阈值难以科学量化的难题,集成学习算法通过构建并结合多个机器学习器来完成学习任务,在解决分类问题时体现出了较为明显的性能优势。为此,本文提出了一种基于集成学习算法CatBoost的面实体匹配方法,将匹配问题转化为分类问题。选取形状、面积、方向和位置4个几何特征作为模型分类特征;利用过采样与欠采样相结合的混合重采样技术减轻原始训练样本的类别不平衡度;借助贝叶斯优化算法确定CatBoost模型的最优超参数;引入可解释人工智能领域的SHAP解释框架从全局和局部两个角度解释各输入特征对匹配结果的影响。在青藏高原的面状湖泊数据上对本文提出的方法进行了验证,实验结果表明：对模型预测影响最大的特征是位置,然后依次是面积、形状,影响最小的特征是方向。CatBoost匹配方法在实验数据集上的查准率、查全率和F1-score分别达到0.9937、0.9753和0.9844,相比于直接使用样本不均衡的原始样本进行模型训练,分别提高了约5.8%、0.6%和3.3%。与传统的面实体多指标双向匹配方法和逻辑回归、K近邻、决策树、神经网络等常规机器学习分类算法相比,集成学习算法CatBoost性能表现更加优异,在避免指标权重和阈值设置难题的同时取得了较好的匹配结果。 相似文献
17.
如何评价弱震区弱活动断裂的最大潜在地震震级及其危险性是地震中长期预测和地震区划研究中较为重要的问题之一。借鉴闻学泽等［1］对中国大陆东部中-弱活断层潜在地震最大震级评估的思路,对丹江口核心水源区已发最大地震震级Mmax与断层小区震级-频度参数at/b值之间的关系进行统计分析。综合at/b值与最大地震震级上限Mu经验公式计算结果、老河口潜在震源区震级上限以及断裂活动性判断,丹江断裂的震级上限为MS6.0,并利用泊松模型方法评估了该断裂的地震平均复发间隔和发震概率。 相似文献
18.
《中国海洋大学学报(英文版)》2021,(3)
Marine environments have a considerable influence on the construction of the Chinese 21 st Century Maritime Silk Road. Thus, an objective and quantitative risk assessment of marine environments has become a key problem that must be solved urgently. To deal with the uncertainty in marine environmental risks caused by complex factors and fuzzy mechanisms, a new assessment technique based on a weighted Bayesian network(BN) is proposed. Through risk factor analysis, node selection, structure construction, and parameter learning, we apply the proposed weighted BN-based assessment model for the risk assessment and zonation of marine environments along the Maritime Silk Road. Results show that the model effectively fuses multisource and uncertain environmental information and provides reasonable risk assessment results, thereby offering technical support for risk prevention and disaster mitigation along the Maritime Silk Road. 相似文献
19.
手足口病是一种常见的传染病,以往的研究表明该疾病与气象、环境和社会经济等因素相关联,其影响关系复杂,而疾病本身体现出较强的区域聚集性,采用普通的线性风险建模方法无法捕捉影响因素的复杂性及空间聚集性。因此,本文以山东省为例,在前人研究的基础上,提出了采用贝叶斯网络综合风险建模方法研究手足口病的发病风险与气象、土地利用、社会经济及空气污染等要素间的关系,并通过引入空间扫描统计聚集结果,将空间聚集引入到贝叶斯网络模型加强其空间推理功能,减少模型的偏差,提高评估的精度。结果表明,本文建立的手足口病空间贝叶斯网络风险模型具有较高的估计效果,引入的空间聚集性较好地融入到贝叶斯概率推理模型中,合理建立预测因子同手足口病发病风险之间的关系。通过对建模结果的解译,分析了手足口病的发病风险影响因素,特别是气候、社会经济及空气污染的影响。本文的空间贝叶斯建模方法及研究结果对手足口病暴发的防控预警具有重要的意义。 相似文献
20.
基于主成分分析(principal component analysis, PCA)方法构建理论模型进行三分量地震偏振分析试算,获得较准确的检波器的水平方位角和垂直倾角,然后对地震响应进行坐标旋转,从而校正检波器方向。将该方法应用于跨孔井间弹性波波速测试,对实测资料进行偏振分析可知,初至P波与初至S波分别在HP分量和R分量中得到加强,不仅有利于初至起跳的拾取,还可以获得较准确的波速资料,为岩土工程设计提供更精确的动力学参数。 相似文献