Landslide susceptibility assessment in Shenzhen based on multi-scale convolutional neural networks model
-
摘要:
卷积神经网络(convolutional neural networks,CNN)模型因其强大的特征提取能力被广泛应用于滑坡易发性评估,但传统CNN已难以满足要求。文章提出一种能够顾及深层与浅层特征的多尺度卷积神经网络(multi-scale convolutional neural networks,MSCNN)模型,通过增加模型深度和样本的感受野,挖掘更深层和更稳定的特征,提高复杂场景下的滑坡易发性评估可靠性。文章以深圳市为研究区,根据系统性原则和代表性原则选取了12个深圳市滑坡影响因子,构建多尺度卷积神经网络滑坡易发性评估模型,并与多层感知器(multilayer perceptron,MLP)、支持向量机(support vector machine,SVM)以及随机森林(random forest,RF)等方法进行对比。结果表明,文章构建的MSCNN模型的AUC值(0.99)较高,优于MLP(0.97)、SVM(0.91)和RF(0.85),证明提出的MSCNN模型具有优异的预测能力;深圳市极高易发性区域面积约为105.3 km2,占研究区总面积的4.98%,主要分布在坡体较陡、植被覆盖稀疏和人类工程活动频繁的龙岗区,坡度、地表粗糙度和地表起伏度成为影响深圳市滑坡的主控因子。文章实现的滑坡易发性图反映了深圳市滑坡灾害的分布现状,可为深圳市未来滑坡灾害防治提供数据支持和关键技术支撑。
Abstract:Convolutional neural network (CNN) models are widely used in landslide susceptibility assessment due to their powerful feature extraction capabilities, and traditional CNN is no longer able to meet the requirements. Therefore, this paper proposes a multi-scale convolutional neural networks (MSCNN) model that can take into account deep and shallow features. By increasing the depth of the model and expanding the receptive field of samples, the MSCNN can tap deeper and more stable features to improve the reliability of landslide susceptibility assessment in complex scenarios. In this study, Shenzhen City is selected as the research area, and 12 landslide conditioning factors of landslides in Shenzhen City were selected based on systematic and representative principles. A multi-scale convolutional neural network landslide susceptibility assessment model is constructed and compared with methods such as multilayer perceptron (MLP), support vector machine (SVM), and random forest (RF). The results show that the AUC value (0.99) of the MSCNN model constructed in this paper is higher than that of MLP (0.97), SVM (0.91), and RF (0.85), which proves that the proposed MSCNN model has excellent prediction ability. The area of extremely high susceptibility in Shenzhen City is approximately 105.3 km², accounting for 4.98% of the total area of the study area, mainly distributed in Longgang District with steep slopes, sparse vegetation cover, and frequent human engineering activities. Slope, surface roughness, and surface relief are identified as the main conditioning factors affecting landslides in Shenzhen City. The landslide susceptibility mapping implemented in this paper reflects the current distribution of landslide disasters in Shenzhen City, providing data support and key technical support for future landslide disaster prevention and control in Shenzhen City.
-
Keywords:
- MSCNN /
- landslide susceptibility assessment /
- machine learning model /
- Shenzhen
-
0. 引 言
每年由于滑坡形成的地质灾害都会给我国造成巨大的人员伤亡和财产损失。根据最新公布的《中国统计年鉴2024》统计,近5年共发生地质灾害
28109 起,其中滑坡地质灾害16290 起,占总地质灾害数量的57.67%[1],是最常见的灾害类型之一。因此,对滑坡地质灾害进行早期防控,对于减少人员伤亡和经济损失都具有十分重要的实际意义。其中,变形速率作为对滑坡地质灾害预警和防控最主要的依据之一,其预测的准确度和时效性对于滑坡地质灾害的预测和防控起着关键作用。然而,滑坡地质灾害特别是突发型滑坡在其变形过程中,可能会发生与历史变形趋势完全不符的变形趋势[2],导致现有方法在预测此类问题时产生困难。随着近年来人工智能技术的飞速发展,在线监测与深度学习相结合的超前预测逐渐成为滑坡地质灾害防控研究的热点[3 − 5]。其中,长短时记忆(long short term memory network,LSTM)神经网络由于其在处理时序数据上的优势而得到广泛关注[6 − 7]。如李丽敏等[8]将滑坡累计位移分解为趋势项与波动项,并用多项式拟合预测趋势项、LSTM网络预测波动项;张明岳等[9]预测对比了循环神经网络(recurrent neural network,RNN)和LSTM2种模型在滑坡位移预测时的精度;LI等[10]采用自回归、LSTM和支持向量机(support vector machines,SVM)建立了综合模型并分析了各模型的权重;唐宇峰等[11]采用了一种动态残差修正的LSTM进行了滑坡位移预测。然而,传统LSTM网络难以同时提取从后向前的信息,使其应用受到了一定的限制。相比于传统LSTM网络,Tengtrairat等[12]提出了一种双向长短期记忆(bi-directional long short term memory network,BiLSTM)神经网络算法,该方法采用双向重叠计算的方法,比单向LSTM可以更好地捕捉双向时序特征,因此具有更好的应用前景[13]。在滑坡领域,Cui等[14]提出一种基于语义门(semantic gate,SG)和双时长短期记忆网络(SG-BiLSTM)的方法,并识别了滑坡体图像;Wang等[15]通过BiLSTM-RNN及卷积神经网络(convolutional neural network,CNN)结合LSTM的方法,生成了基于人工智能的香港滑坡敏感性地图。Lin等[16]采用GRA-MIC融合相关计算方法选取了影响滑坡位移的因素,最后采用CNN-BiLSTM模型进行了预测。综上,BiLSTM方法在滑坡领域内已经取得许多成果。然而,对于突发型滑坡灾害,由于其在加速变形过程中的变形速率发展历程可能与历史变形速率历程完全不符,导致了现有方法在预测此类问题时存在精度及效率不足的困难。因此,建立一种动态预测且深层优化的多层耦合算法,在提高预测准确率的同时保证较快的响应速度,对于准确地进行突发型滑坡预警及增加预警后的应急响应时间是具有十分重要的实际意义的。
鉴于此,本文提出一种基于动态串联PSO- BiLSTM的滑坡变形速率预测方法,首先,通过集合经验模态分解(EEMD)将变形速率序列进行分解,得到周期项及趋势项变形速率序列;其次,设置PSO启动阈值,并分别通过多项式拟合及周期项PSO-BiLSTM预测网络,得到趋势项及周期项变形速率预测值,将预测值分别加入趋势项变形速率序列及周期项变形速率序列;再次,以趋势项变形速率序列、周期项变形速率序列及残差变形速率序列为输入,建立总PSO-BiLSTM预测网络,得到总预测变形速率,最后,由总预测变形速率和监测变形速率,相减得到下一循环计算所需的残差变形速率。通过以上方式,提高对变形速率预测的准确率及滑坡预警的响应速度,为增加滑坡预警时间提供一种新的思路。
1. PSO-BiLSTM算法理论基础
1.1 LSTM算法理论基础
LSTM神经网络是对RNN的改进算法,其工作原理见图1[17]。LSTM算法在RNN的基础上引入了单元状态c以及“门”的概念,解决了在RNN中存在的梯度消失和爆炸问题。
LSTM神经网络的基本单元称为细胞,由遗忘门、输入门及输出门构成。其中,遗忘门决定上一时刻的状态St−1保留至当前时刻的信息,其通过一个取值为0~1范围的Sigmoid函数,将输入xt与上一时刻的输出ht−1相联系来决定遗忘的信息,Sigmoid函数取0表示全部遗忘,取1代表全部记忆;输入门的作用是控制当前输入xt保存到状态单元St中的记忆量,其主要结构算法为:
(1) (2) (3) (4) (5) (6) 式中:
、 、 、 、 、 ——遗忘门、输入门、当前输 入单元状态、当前时刻单 元状态、输出、最终输出; 、 、 、 ——遗忘门、输入门、当前输入 单元和输出的权重矩阵, 、 、 、 ——遗忘门、输入门、当前输入单 元和输出的偏置项; ——2个向量连接为一个更长向量; ——Sigmoid函数; ——将实数映射到 的双正切函数[18]。1.2 BiLSTM算法理论基础
BiLSTM网络是在LSTM网络基础上发展起来的,其可在不增加数据量的前提下学习序列数据和时间步长之间的双向依赖关系[12]。BiLSTM和单向LSTM最大的区别在于,前者可以同时保存过去和未来的信息,而后者只保存过去的信息,如图2所示。
1.3 PSO-BiLSTM算法
PSO常用来求解最优化问题,其基本思路是将待求解问题的解描述为粒子,每个粒子在N维解空间中可以不断寻求,其粒子极值
代表粒子所经过位置中的最优解,而粒子群中最好的粒子位置定义为 [19]。每一个粒子都会在循环中追踪 和 从而进行位置更新,并通过计算适应度重新获取 和 ,从而不断逼近最优解。对某一个粒子 ,其速度和位置更新方法如下:式中:
——惯性权重; ——空间的维数, ; ——粒子的个数, ; 、 ——粒子的速度、位置; ——当前迭代次数; 、 ——学习因子; 、 —— 之间的随机数; 、 ——粒子 最佳位置和全部粒子的最佳位置。基于以上理论,基于PSO优化的BiLSTM网络训练流程如图3所示。
2. 动态串联PSO-BiLSTM算法
2.1 动态串联PSO-BiLSTM算法基本原理
滑坡体在随着时间的演变过程中,其滑坡变形速率趋势会呈现出“稳定型”、“渐变型”、“突发型”等不同的变化趋势。对于突发型滑坡,其从变形速率突变到产生滑坡的时间非常短(图4),现有方法在解决突变型滑坡位移预测时存在明显的精度不足、效率低下等困难。
(1)传统BiLSTM网络仅通过已有监测数据一次性建立和验证网络,并将该网络作为后续预测的依据,即“静态网络”。这种网络训练完成后不再更新,当变形速率发展趋势与前期变形速率趋势发生较大变化时,难以适应新趋势的发展,无法对变形速率进行有效预测。
(2)“动态网络”,即在每一次得到新的监测数据后对网络进行更新,可使网络具备更新后的变形速率信息,相比于静态网络可以显著提高预测精度。然而,一方面,动态BiLSTM网络预测的准确率与经验设定的网络参数有关,参数不当会严重影响其预测精度;另一方面,在变形速率产生突变时,其突变后的数据量少、突变速率变化快,动态网络的方法在预测时仍存在严重的滞后性。
(3)PSO优化可以对BiLSTM参数进行寻优从而提高准确率,但会大大增加计算成本,导致过长的预测时间而不利于工程的实际应用。
基于以上现状,本文建立了一种基于动态串联PSO-BiLSTM的滑坡变形速率预测方法。首先,设置每一次要分析的变形速率数据量N,通过“动态滑窗”方式截取待分析数据(“动态滑窗”指每获取最新一轮时序数据,将序列中最早的一轮时序数据去除,从而保持总数据量始终不变,达到减小待分析数据量和历史数据的影响的目的),并通过集合经验模态分解(EEMD)将变形速率序列进行分解,得到周期项及趋势项变形速率序列;其次,分别通过多项式拟合及周期项PSO-BiLSTM预测网络,得到趋势项及周期项变形速率预测值,并将预测值分别加入趋势项变形速率序列及周期项变形速率序列;再次,以趋势项变形速率序列、周期项变形速率序列及残差变形速率序列为输入,建立总PSO-BiLSTM预测网络,得到总预测变形速率;最后,由总预测变形速率和监测变形速率,相减得到下一循环建立总PSO-BiLSTM所需的残差变形速率。需要指出的是:(1)为提升预测效率,为PSO-BiLSTM网络仅当某一次预测的周期项残差率
( 为PSO启动阈值)时PSO才会启动,否则直接进行BiLSTM网络预测;(2)残差变形速率由下一循环监测的实际数据与当前循环的总预测变形速率之差求得,而当计算循环数i≤S(S为串联PSO-BiLSTM启动阈值)时总预测变形速率=趋势项变形速率预测值+周期项变形速率预测值,当i>S时总预测变形速率为串联PSO-BiLSTM网络求得。其流程图如图5所示。通过以上方式,在动态训练网络的基础上,考虑了不同历史时刻条件下动态网络速率预测的误差,实现了不同动态网络之间的学习,且仅当预测误差过大时才启动PSO优化,因此可在提高预测准确率的前提下保证较高的计算效率。
2.2 模型性能指标
为全面评价该滑坡变形速率预测模型精度,采用平均绝对误差(mean absolute error,MAE)、绝对百分比误差(mean absolute percentage error,MAPE)、均方根误差(root mean square error,RMSE)及拟合优度R2来作为模型性能指标,如下[20]:
(7) (8) (9) (10) 式中:
——实测值; ——预测值; ——样本数量。3. 工程实例
3.1 滑坡变形速率实测及预处理
以四川省某滑坡体实测变形速率为例。自2020年3月14日—9月17日间对该案例进行了地表裂缝监测及降雨量监测(图6),其中每小时一组数据,共得到了
11776 组滑坡变形速率监测数据。该滑坡体在前期监测中变形一直相对稳定,自9月13日12时起,该滑坡隐患点在连日降雨影响下,其变形速率有明显突变迹象。如图7所示为9月13日前500 h内变形速率变化情况。
由图7所示, 该滑坡体在前期一直处于稳定状态,仅在突变前极短时间内变形速率发生了突变。由于本文的研究目的主要是针对变形速率产生突变情况下的预测研究,稳定期的变形速率对文章研究意义不大。因此仅选取了2020年9月12日17时—14日6时间的50 h为例进行探讨分析。
在实际工程中,存在变形速率的发展不是单向放大过程的情况,而是在某些时刻内会存在阶梯性的变化,这给变形速率的预测带来了困难。在本文中,采用“相邻极大值”方法来对数据进行预处理,即每一个时刻t的变形速率
,从而平滑变形速率曲线,有利于进一步进行变形速率预测。相比于常用的“相邻平均”的方法,极大值法关注最大变形速率,因此在实际工程中是偏于安全的,如图8所示。从图8可以看出,该滑坡体的变形速率在20 h之后呈快速增加的趋势,其变化形态与历史趋势有着较大的差异,即产生了“突变”。
3.2 模型建立与参数选取
为验证本文方法的优势,分别采用动态BiLSTM网络(类型Ι)、动态PSO-BiLSTM网络(类型Ⅱ)、文中提出的动态串联PSO-BiLSTM网络在PSO启动残差率C=0(类型Ⅲ)和C=0.1(类型IV)4种情况进行预测分析。每类BiLSTM网络的输出均为一维。其中,动态BiLSTM的BiLSTM层节点数选取为50,正则化系数为0,初始学习率为0.2;PSO优化的参数为BiLSTM层节点数、初始学习率和正则化系数;PSO启动残差率C取0.1,动态滑窗截取数据为30,预测数据量为20。
3.3 预测结果及指标评价
4种类型算法均进行了3次预测,取计算结果的平均值。图9为4种方法变形速率预测结果与实测结果20轮预测对比。
其中,为获取总PSO-BiLSTM网络所需的残差变形速率,在前10个预测循环中仅采用周期项PSO-BiLSTM网络进行预测,而最后10项采用动态串联PSO-BiLSTM网络进行预测。因此,以下仅选择最后10次数据进行对比分析,如图10所示。
为进一步对比预测结果,采用MAE、MAPE、RMSE及R2共4种评价指标评价位移预测结果,表1为各模型位移预测结果的评价结果。
表 1 预测结果评价Table 1. Evaluation of prediction results预测
类型位移评价指标 计算时间/s MAE MAPE/% RMSE R2 类型Ⅰ 0.43 8.45 0.70 0.96 24.89 类型Ⅱ 0.36 7.07 0.61 0.97 294.50 类型Ⅲ 0.30 5.82 0.51 0.98 1861.87 类型Ⅳ 0.28 5.41 0.57 0.98 380.22 其中,MAE为绝对误差;MAPE为预测值与实测值的平均偏离程度,其越接近0表示效果越好;RMSE为预测值与真实值之间的偏差,越接近0表示预测值与真实值越吻合;R2越接近1说明预测越准确。
从图10及表1可以看出:(1)动态BiLSTM网络(类型Ⅰ)的绝对误差和平均偏离程度、偏差均为最大,且拟合优度R2最小,说明此时类型Ⅰ在四种类型中效果最差,但由于未采用PSO优化算法,此时拥有最佳的计算效率;(2)加入PSO算法后(类型Ⅱ),其MAE、MAPE和RMSE值相比类型Ⅰ均有明显的下降,且拟合优度上升,说明PSO优化算法对预测结果有明显的提升;(3)当加入串联算法后(类型Ⅲ和类型Ⅳ),其MAE、MAPE和RMSE值进一步下降,而拟合优度进一步提升,说明串联算法对预测精度有进一步提升;(4)类型Ⅲ与类型Ⅳ的各评价指标差距较小且各有优劣,但类型Ⅳ相比与类型Ⅲ计算效率大大提升,这对于滑坡变形速率的快速预测具有重要的实用价值。
4. 结论
(1)传统动态BiLSTM算法在进行滑坡变形速率预测时具有较高的计算效率,但在面临滑坡变形速率快速变化的情况时预测精度偏低;而相对于传统动态BiLSTM算法,PSO-BiLSTM优化算法对突发型滑坡变形速率预测结果有明显的提升。
(2)动态串联PSO-BiLSTM算法可以有效地提高突发型滑坡变形速率的预测准确率,但由于PSO优化计算时间过长,不利于工程应用;加入PSO启动机制后,其MAE、MAPE、RMSE、R2分别为0.28、5.41%、0.57、0.98,计算时间为380.22 s,在具有较高的精度的同时保证了计算效率,对于滑坡预测的快速响应、提高工程实用价值都有重要的意义。
-
表 1 滑坡影响因子数据来源
Table 1 Data sources for landslide conditioning factors
数据源 分辨率 滑坡影响因子 数据来源 数字高程模型 30 m 高程 https://www.gscloud.cn/ 坡度 坡向 曲率 地表粗糙度 地表起伏度 断层 1∶ 2500000 到断层距离 https://www.cgs.gov.cn/ 道路和河流 1∶ 1000000 到河流距离 OpenStreetMap 到道路距离 土地利用类型 30 m 土地利用类型 http://data.ess.tsinghua/ 土壤类型 1000 m土壤类型 https://www.fao.org/ 沙含量 1000 m沙含量 http://www.geodata.cn/ 表 2 影响因子共线性评价表
Table 2 Evaluation of factor collinearity among conditioning factors
序号 影响因子 VIF 容差 1 高程 1.514 0.660 2 坡度 6.666 0.150 3 坡向 1.007 0.993 4 曲率 1.003 0.997 5 地表粗糙度 5.012 0.200 6 地表起伏度 2.148 0.466 7 到断层距离 1.082 0.924 8 土壤类型 1.071 0.993 9 沙含量 1.029 0.972 10 到河流距离 1.110 0.901 11 土地利用类型 1.285 0.778 12 到道路距离 1.070 0.935 表 3 影响因子地理探测器结果
Table 3 Results of geodetector analysis for conditioning factors
序号 影响因子 q值 1 高程 0.322 2 坡度 0.185 3 坡向 0.058 4 曲率 0.073 5 地表粗糙度 0.118 6 地表起伏度 0.193 7 到断层距离 0.107 8 土壤类型 0.144 9 沙含量 0.179 10 到河流距离 0.172 11 土地利用类型 0.128 12 到道路距离 0.093 -
[1] 葛大庆,戴可人,郭兆成,等. 重大地质灾害隐患早期识别中综合遥感应用的思考与建议[J]. 武汉大学学报(信息科学版),2019,44(7):949 − 956. [GE Daqing,DAI Keren,GUO Zhaocheng,et al. Reflections and suggestions on the application of integrated remote sensing in the early identification of major geological hazard potential[J]. Geomatics and Information Science of Wuhan University,2019,44(7):949 − 956. (in Chinese with English abstract)] GE Daqing, DAI Keren, GUO Zhaocheng, et al. Reflections and suggestions on the application of integrated remote sensing in the early identification of major geological hazard potential[J]. Geomatics and Information Science of Wuhan University, 2019, 44(7): 949 − 956. (in Chinese with English abstract)
[2] 刘帅,王涛,曹佳文,等. 基于优化随机森林模型的降雨群发滑坡易发性评价研究——以西秦岭极端降雨事件为例[J/OL]. 地质通报(2024-01-19)[2024-03-15]. [LIU Shuai,WANG Tao,CAO Jiawen,et al. A case study on susceptibility assessment of precipitation-induced mass landslides based on optimal random forest model, west Qinling Mountains[J/OL]. Geological Bulletin of China,(2024-01-19)[2024-03-15]. http://kns.cnki.net/kcms/detail/11.4648.P.20240118.1821.002.html. (in Chinese with English abstract)] LIU Shuai, WANG Tao, CAO Jiawen, et al. A case study on susceptibility assessment of precipitation-induced mass landslides based on optimal random forest model, west Qinling Mountains[J/OL]. Geological Bulletin of China, (2024-01-19)[2024-03-15]. http://kns.cnki.net/kcms/detail/11.4648.P.20240118.1821.002.html. (in Chinese with English abstract)
[3] 冉涛,徐如阁,周洪福,等. 雅砻江流域深切河谷区滑坡类型、成因及分布规律——以子拖西―麻郎错河段为例[J]. 中国地质,2024,51(2):511 − 524. [RAN Tao,XU Ruge,ZHOU Hongfu,et al. Type, formation mechanism and distribution regularity of landslides in the deeply incised valley area of Yalong River Basin:A case study of Zituoxi–Malangcuo river section[J]. Geology in China,2024,51(2):511 − 524. (in Chinese with English abstract)] RAN Tao, XU Ruge, ZHOU Hongfu, et al. Type, formation mechanism and distribution regularity of landslides in the deeply incised valley area of Yalong River Basin: A case study of Zituoxi–Malangcuo river section[J]. Geology in China, 2024, 51(2): 511 − 524. (in Chinese with English abstract)
[4] 朱庆,曾浩炜,丁雨淋,等. 重大滑坡隐患分析方法综述[J]. 测绘学报,2019,48(12):1551 − 1561. [ZHU Qing,ZENG Haowei,DING Yulin,et al. A review of major potential landslide hazards analysis[J]. Acta Geodaetica et Cartographica Sinica,2019,48(12):1551 − 1561. (in Chinese with English abstract)] ZHU Qing, ZENG Haowei, DING Yulin, et al. A review of major potential landslide hazards analysis[J]. Acta Geodaetica et Cartographica Sinica, 2019, 48(12): 1551 − 1561. (in Chinese with English abstract)
[5] 陶伟,胡晓波,姜元俊,等. 颗粒粒径对滑坡碎屑流动力特征及能量转化的影响——以四川省三溪村滑坡为例[J]. 地质通报,2023,42(9):1610 − 1619. [TAO Wei,HU Xiaobo,JIANG Yuanjun,et al. Influence of particle size on dynamic characteristics and energy conversion of debris flow in landslide:A case study of Sanxicun landslide in Sichuan Province[J]. Geological Bulletin of China,2023,42(9):1610 − 1619.] TAO Wei, HU Xiaobo, JIANG Yuanjun, et al. Influence of particle size on dynamic characteristics and energy conversion of debris flow in landslide: A case study of Sanxicun landslide in Sichuan Province[J]. Geological Bulletin of China, 2023, 42(9): 1610 − 1619.
[6] 桑凯. 近60年中国滑坡灾害数据统计与分析[J]. 科技传播,2013,5(10):129. [SANG Kai. Statistics and analysis of landslide disasters in China in the last 60 years [J]. Public Communication of Science & Technology,2013,5(10):129. (in Chinese with English abstract)] SANG Kai. Statistics and analysis of landslide disasters in China in the last 60 years [J]. Public Communication of Science & Technology, 2013, 5(10): 129. (in Chinese with English abstract)
[7] 李振洪,宋闯,余琛,等. 卫星雷达遥感在滑坡灾害探测和监测中的应用:挑战与对策[J]. 武汉大学学报(信息科学版),2019,44(7):967 − 979 [LI Zhenhong,SONG Chuang,YU Chen,et al. Satellite radar remote sensing in landslide hazard detection and monitoring:Challenges and countermeasures[J]. Geomatics and Information Science of Wuhan University,2019,44(7):967 − 979. (in Chinese with English abstract)] LI Zhenhong, SONG Chuang, YU Chen, et al. Satellite radar remote sensing in landslide hazard detection and monitoring: Challenges and countermeasures[J]. Geomatics and Information Science of Wuhan University, 2019, 44(7): 967 − 979. (in Chinese with English abstract)
[8] FANG Zhice,WANG Yi,PENG Ling,et al. Integration of convolutional neural network and conventional machine learning classifiers for landslide susceptibility mapping[J]. Computers and Geosciences,2020,139:104470.
[9] 杨金山,王常效. 韧性城市建设的深圳实践与展望[J]. 特区实践与理论,2022(2):79 − 84. [YANG Jinshan,WANG Changxiao. Shenzhen practice and prospects for resilient city building[J]. Practice and Theory of Sezs,2022(2):79 − 84. (in Chinese with English abstract)] YANG Jinshan, WANG Changxiao. Shenzhen practice and prospects for resilient city building[J]. Practice and Theory of Sezs, 2022(2): 79 − 84. (in Chinese with English abstract)
[10] CHEN Wei,POURGHASEMI H R,PANAHI M,et al. Spatial prediction of landslide susceptibility using an adaptive neuro-fuzzy inference system combined with frequency ratio,generalized additive model,and support vector machine techniques[J]. Geomorphology,2017,297:69 − 85.
[11] 刘传正. 深圳红坳弃土场滑坡灾难成因分析[J]. 中国地质灾害与防治学报,2016,27(1):1 − 5. [LIU Chuanzheng. Genetic mechanism of landslide tragedy happened in Hong’ao dumping place in Shenzhen, China[J]. The Chinese Journal of Geological Hazard and Control,2016,27(1):1 − 5. (in Chinese with English abstract)] LIU Chuanzheng. Genetic mechanism of landslide tragedy happened in Hong’ao dumping place in Shenzhen, China[J]. The Chinese Journal of Geological Hazard and Control, 2016, 27(1): 1 − 5. (in Chinese with English abstract)
[12] 高杨, 卫童瑶, 李滨, 等. 深圳“12•20” 渣土场远程流化滑坡动力过程分析[J]. 水文地质工程地质,2019,46(1):129 − 138. [GAO Yang, WEI Tongyao, LI Bin, et al. Dynamics process simulation of long Run-out catastrophic landfill flowslide on December 20th, 2015 in Shenzhen, China[J]. Hydrogeology & Engineering Geology,2019,46(1):129 − 138. (in Chinese with English abstract)] GAO Yang, WEI Tongyao, LI Bin, et al. Dynamics process simulation of long Run-out catastrophic landfill flowslide on December 20th, 2015 in Shenzhen, China[J]. Hydrogeology & Engineering Geology, 2019, 46(1): 129 − 138. (in Chinese with English abstract)
[13] 朱庆,张曼迪,丁雨淋,等. 环境因子空间特征约束的区域滑坡敏感性模糊逻辑分析方法[J]. 武汉大学学报(信息科学版),2021,46(10):1431 − 1440. [ZHU Qing,ZHANG Mandi,DING Yulin,et al. A fuzzy logic analysis method for regional landslide sensitivity with spatial feature constraints of environmental factors[J]. Geomatics and Information Science of Wuhan University,2021,46(10):1431 − 1440. (in Chinese with English abstract)] ZHU Qing, ZHANG Mandi, DING Yulin, et al. A fuzzy logic analysis method for regional landslide sensitivity with spatial feature constraints of environmental factors[J]. Geomatics and Information Science of Wuhan University, 2021, 46(10): 1431 − 1440. (in Chinese with English abstract)
[14] 刘福臻,王灵,肖东升. 机器学习模型在滑坡易发性评价中的应用[J]. 中国地质灾害与防治学报,2021,32(6):98 − 106. [LIU Fuzhen,WANG Ling,XIAO Dongsheng. Application of machine learning model in landslide susceptibility evaluation[J]. The Chinese Journal of Geological Hazard and Control,2021,32(6):98 − 106. (in Chinese with English abstr-act)] LIU Fuzhen, WANG Ling, XIAO Dongsheng. Application of machine learning model in landslide susceptibility evaluation[J]. The Chinese Journal of Geological Hazard and Control, 2021, 32(6): 98 − 106. (in Chinese with English abstr-act)
[15] 何书,鲜木斯艳·阿布迪克依木,胡萌,等. 基于自组织特征映射网络-随机森林模型的滑坡易发性评价——以江西大余县为例[J]. 中国地质灾害与防治学报,2022,33(1):132 − 140. [HE Shu,ABUDIKEYIMU XMSY,HU Meng,et al. Evaluation on landslide susceptibility based on self-organizing feature map network and random forest model:A case study of Dayu County of Jiangxi Province[J]. The Chinese Journal of Geological Hazard and Control,2022,33(1):132 − 140. (in Chinese with English abstract)] HE Shu, ABUDIKEYIMU XMSY, HU Meng, et al. Evaluation on landslide susceptibility based on self-organizing feature map network and random forest model: A case study of Dayu County of Jiangxi Province[J]. The Chinese Journal of Geological Hazard and Control, 2022, 33(1): 132 − 140. (in Chinese with English abstract)
[16] 贾雨霏,魏文豪,陈稳,等. 基于SOM-I-SVM耦合模型的滑坡易发性评价[J]. 水文地质工程地质,2023,50(3):125 − 137. [JIA Yufei,WEI Wenhao,CHEN Wen,et al. Landslide susceptibility assessment based on the SOM-I-SVM model[J]. Hydrogeology& Engineering Geology,2023,50(3):125 − 137. (in Chinese with English abstract)] JIA Yufei, WEI Wenhao, CHEN Wen, et al. Landslide susceptibility assessment based on the SOM-I-SVM model[J]. Hydrogeology& Engineering Geology, 2023, 50(3): 125 − 137. (in Chinese with English abstract)
[17] HUANG Faming,CAO Zhongshan,GUO Jianfei,et al. Comparisons of heuristic,general statistical and machine learning models for landslide susceptibility prediction and mapping[J]. Catena,2020,191:104580.
[18] 王毅,方志策,牛瑞卿,等. 基于深度学习的滑坡灾害易发性分析[J]. 地球信息科学学报,2021,23(12):2244 − 2260. [WANG Yi,FANG Zhice,NIU Ruiqing,et al. Landslide hazard susceptibility analysis based on deep learning[J]. Journal of Geo-Information Science,2021,23(12):2244 − 2260. (in Chinese with English abstract)] WANG Yi, FANG Zhice, NIU Ruiqing, et al. Landslide hazard susceptibility analysis based on deep learning[J]. Journal of Geo-Information Science, 2021, 23(12): 2244 − 2260. (in Chinese with English abstract)
[19] 蒋万钰,陈冠,孟兴民,等. 基于卷积神经网络模型的区域滑坡敏感性评价——以川藏铁路沿线为例[J]. 兰州大学学报(自然科学版),2022,58(2):203 − 211. [JIANG Wanyu,CHEN Guan,MENG Xingmin,et al. Regional landslide sensitivity evaluation based on convolutional neural network model:An example along the Sichuan-Tibet Railway[J]. Journal of Lanzhou University(Natual Sciences),2022,58(2):203 − 211. (in Chinese with English abstract)] JIANG Wanyu, CHEN Guan, MENG Xingmin, et al. Regional landslide sensitivity evaluation based on convolutional neural network model: An example along the Sichuan-Tibet Railway[J]. Journal of Lanzhou University(Natual Sciences), 2022, 58(2): 203 − 211. (in Chinese with English abstract)
[20] 赵占骜,王继周,毛曦,等. 多维CNN耦合的滑坡易发性评价方法[J]. 武汉大学学报(信息科学版),(2023-04-12)[2023-04-21]. [ZHAO Zhanao,WANG Jizhou,MAO Xi,et al. A multi-dimensional CNN coupled landslide susceptibility assessment method[J]. Geomatics and Information Science of Wuhan University,(2023-04-12)[2023-04-21]. https://doi.org/10.13203/j.whugis20220325. (in Chinese with English abstract)] ZHAO Zhanao, WANG Jizhou, MAO Xi, et al. A multi-dimensional CNN coupled landslide susceptibility assessment method[J]. Geomatics and Information Science of Wuhan University, (2023-04-12)[2023-04-21]. https://doi.org/10.13203/j.whugis20220325. (in Chinese with English abstract)
[21] HE Yi,ZHAO Zhan’ao,YANG Wang,et al. A unified network of information considering superimposed landslide factors sequence and pixel spatial neighbourhood for landslide susceptibility mapping[J]. International Journal of Applied Earth Observation and Geoinformation,2021,104:102508.
[22] 李粮纲,徐玉胜,江辉煌,等. 深圳地区地质环境特征与地质灾害防治[J]. 安全与环境工程,2007(4):28 − 31. [LI Lianggang,XU Yusheng,JIANG Huihuang,et al. Characteristics of the geological environment and prevention of geological hazards in Shenzhen[J]. Safety and Environmental Engineering,2007(4):28 − 31. (in Chinese with English abstract)] LI Lianggang, XU Yusheng, JIANG Huihuang, et al. Characteristics of the geological environment and prevention of geological hazards in Shenzhen[J]. Safety and Environmental Engineering, 2007(4): 28 − 31. (in Chinese with English abstract)
[23] 李丹彤. 深圳市地质灾害防治管理研究[D]. 深圳:深圳大学,2017. [LI Dantong. Study on the management of geological disaster prevention and control in Shenzhen[D]. Shenzhen:Shenzhen University,2017. (in Chinese with English abstract)] LI Dantong. Study on the management of geological disaster prevention and control in Shenzhen[D]. Shenzhen: Shenzhen University, 2017. (in Chinese with English abstract)
[24] 崔玉龙,傅贵,郭鸿,等. 基于谷歌影像的新疆典型黄土区滑坡分布分析[J]. 陕西理工大学学报(自然科学版),2020,36(2):51 − 56. [CUI Yulong,FU Gui,GUO Hong,et al. Analysis of landslide distribution in typical loess areas of Xinjiang based on Google images[J]. Journal of Shaanxi University of Technology(Natural Science Edition),2020,36(2):51 − 56. (in Chinese with English abstract)] CUI Yulong, FU Gui, GUO Hong, et al. Analysis of landslide distribution in typical loess areas of Xinjiang based on Google images[J]. Journal of Shaanxi University of Technology(Natural Science Edition), 2020, 36(2): 51 − 56. (in Chinese with English abstract)
[25] BUI D T,PRADHAN B,LOFMAN O,et al. Spatial prediction of landslide hazards in Hoa Binh Province (Vietnam):A comparative assessment of the efficacy of evidential belief functions and fuzzy logic models[J]. Catena,2012,96:28 − 40.
[26] 陈学兄. 基于遥感与GIS的中国水土流失定量评价[D]. 杨凌:西北农林科技大学,2013. [CHEN Xuexiong. Quantitative evaluation of soil erosion in China based on remote sensing and GIS[D]. YANG Ling:Northwest Agriculture and Forestry University,2013. (in Chinese with English abstract)] CHEN Xuexiong. Quantitative evaluation of soil erosion in China based on remote sensing and GIS[D]. YANG Ling: Northwest Agriculture and Forestry University, 2013. (in Chinese with English abstract)
[27] 王文辉. 联合SBAS-InSAR与机器学习的滑坡隐患识别——以兰州市为例[D]. 兰州:兰州交通大学,2021. [WANG Wenhui. Combined SBAS-InSAR and machine learning for landslide hazard identification[D]. Lanzhou:Lanzhou Jiaotong University,2021. (in Chinese with English abstract)] WANG Wenhui. Combined SBAS-InSAR and machine learning for landslide hazard identification[D]. Lanzhou: Lanzhou Jiaotong University, 2021. (in Chinese with English abstract)
[28] HE Sanwei,PAN Peng,DAI Lan,et al. Application of kernel-based Fisher discriminant analysis to map landslide susceptibility in the Qinggan River delta,Three Gorges,China[J]. Geomorphology,2012,171/172:30 − 41.
[29] 张玉芳,齐明柱,马华. 深圳市边坡病害及其防治[J]. 岩石力学与工程学报,2006,(增刊2):3412 − 3421. [ZHANG Yufang,QI Mingzhu,MA Hua. Slope diseases and prevention in Shenzhen[J]. Chinese Journal of Rock Mechanics and Engineering,2006,(Sup 2):3412 − 3421. (in Chinese with English abstract)] ZHANG Yufang, QI Mingzhu, MA Hua. Slope diseases and prevention in Shenzhen[J]. Chinese Journal of Rock Mechanics and Engineering, 2006, (Sup 2): 3412 − 3421. (in Chinese with English abstract)
[30] 李利峰,张晓虎,邓慧琳,等. 基于熵指数与逻辑回归耦合模型的滑坡灾害易发性评价——以蓝田县为例[J]. 科学技术与工程,2020,20(14):5536 − 5543. [LI Lifeng,ZHANG Xiaohu,DENG Huilin,et al. Landslide hazard susceptibility evaluation based on entropy index and logistic regression coupled model:Lantian County as an example[J]. Science Technology and Engineering,2020,20(14):5536 − 5543. (in Chinese with English abstract)] LI Lifeng, ZHANG Xiaohu, DENG Huilin, et al. Landslide hazard susceptibility evaluation based on entropy index and logistic regression coupled model: Lantian County as an example[J]. Science Technology and Engineering, 2020, 20(14): 5536 − 5543. (in Chinese with English abstract)
[31] 何书,胡萌,杨志华,等. 基于模糊频率比与熵指数的滑坡易发性评价——以崇义县为例[J]. 有色金属科学与工程,2022,13(4):80 − 90. [HE Shu,HU Meng,YANG Zhihua,et al. Landslide susceptibility evaluation based on fuzzy frequency ratio and entropy index:An example from Chongyi County[J]. Nonferrous Metals Science and Engineering,2022,13(4):80 − 90. (in Chinese with English abstract)] HE Shu, HU Meng, YANG Zhihua, et al. Landslide susceptibility evaluation based on fuzzy frequency ratio and entropy index: An example from Chongyi County[J]. Nonferrous Metals Science and Engineering, 2022, 13(4): 80 − 90. (in Chinese with English abstract)
[32] 王劲峰,徐成东. 地理探测器:原理与展望[J]. 地理学报,2017,72(1):116 − 134. [WANG Jinfeng,XU Chengdong. Geodetector:Principles and prospects[J]. Acta Geographi-ca Sinica,2017,72(1):116 − 134. (in Chinese with English abstract)] WANG Jinfeng, XU Chengdong. Geodetector: Principles and prospects[J]. Acta Geographi-ca Sinica, 2017, 72(1): 116 − 134. (in Chinese with English abstract)
[33] 王志恒,胡卓玮,赵文吉,等. 基于多层感知器模型的区域滑坡敏感性评价研究——以四川低山丘陵区为例[J]. 防灾减灾工程学报,2015,35(5):691 − 698. [WANG Zhiheng,HU Zhuowei,ZHAO Wenji,et al. A study on regional landslide sensitivity evaluation based on multi-layer perceptron model:An example of low mountain hilly area in Sichuan[J]. Journal of Disaster Prevention and Mitigation Engineering,2015,35(5):691 − 698. (in Chinese with English abstract)] WANG Zhiheng, HU Zhuowei, ZHAO Wenji, et al. A study on regional landslide sensitivity evaluation based on multi-layer perceptron model: An example of low mountain hilly area in Sichuan[J]. Journal of Disaster Prevention and Mitigation Engineering, 2015, 35(5): 691 − 698. (in Chinese with English abstract)
[34] 王卫东,刘攀,龚陆. 基于支持向量机模型的四川省滑坡灾害易发性区划[J]. 铁道科学与工程学报,2019,16(5):1194 − 1200. [WANG Weidong,LIU Pan,GONG Lu. Landslide hazard susceptibility zoning in Sichuan Province based on support vector machine model[J]. Journal of Railway Science and Engineering,2019,16(5):1194 − 1200. (in Chinese with English abstract)] WANG Weidong, LIU Pan, GONG Lu. Landslide hazard susceptibility zoning in Sichuan Province based on support vector machine model[J]. Journal of Railway Science and Engineering, 2019, 16(5): 1194 − 1200. (in Chinese with English abstract)
[35] 黄发明,胡松雁,闫学涯,等. 基于机器学习的滑坡易发性预测建模及其主控因子识别[J]. 地质科技通报,2022,41(2):79 − 90. [HUANG Faming,HU Songyan,YAN Xueya,et al. Machine learning-based predictive modeling of landslide susceptibility and its master control factor identification[J]. Bulletin of Geological Science and Technology,2022,41(2):79 − 90. (in Chinese with English abstract)] HUANG Faming, HU Songyan, YAN Xueya, et al. Machine learning-based predictive modeling of landslide susceptibility and its master control factor identification[J]. Bulletin of Geological Science and Technology, 2022, 41(2): 79 − 90. (in Chinese with English abstract)
[36] HONG H,TSANGARATOS P,ILIA I,et al. Introducing a novel multi-layer perceptron network based on stochastic gradient descent optimized by a meta-heuristic algorithm for landslide susceptibility mapping[J]. Sci Total Environ,2020,742:140549.
-
期刊类型引用(7)
1. 蔡小东. 暴雨条件下赣南山区高速公路高边坡稳定性研究. 价值工程. 2025(04): 139-141 . 百度学术
2. 赖波,江金进,江山,江宁,赵风顺. 珠海市南水镇金龙边坡变形破坏特征及稳定性评价. 城市地质. 2024(01): 20-28 . 百度学术
3. 王玉刚,袁伟. 基于遗传算法的地下暗挖工程边坡稳定性分析. 大众标准化. 2023(15): 103-105 . 百度学术
4. 林逸晖,李广涛,杨天雨,乔登攀,王俊,张希,赵怀军. 基于PCA-RF的边坡稳定性预测. 化工矿物与加工. 2023(12): 59-65 . 百度学术
5. 王猛,何德伟,贾志宏,胡至华. 基于多源遥感数据的高位滑坡特征分析——以广元市利州区荣山镇岩窝村滑坡为例. 中国地质灾害与防治学报. 2023(06): 57-68 . 本站查看
6. 刘顺青,蔡宇宸,程涛,周萍,王旭畅. 降雨入渗条件下下蜀土边坡稳定性分析. 地质灾害与环境保护. 2022(02): 32-37 . 百度学术
7. 郭世兴. 关于崩滑地质灾害抢险治理工程的斜坡稳定性分析. 世界有色金属. 2021(21): 203-204 . 百度学术
其他类型引用(1)