基于滑坡监测数据的Elman神经网络动态预测

滑坡在我国是一种极为频发的地质灾害,且其积累位移监测曲线有着复杂的非线性特性,对此各研究者建立过许多预测模型,然而这些模型的预测精度不尽如人意。基于Elman神经网络可以任意精度逼近任意非线性函数的特征,并以sigmoid为方程的核函数,在选择隐含层数时用了试用法,通过\     

基于SSA-Adam-BP神经网络模型的堰塞坝稳定性预测

现有的堰塞坝稳定性预测模型多为线性模型,无法充分考虑堰塞坝稳定性与其形态特征和水域条件之间的复杂非线性关系.鉴于此,结合反向传播神经网络模型和樽海鞘优化算法,提出了一种新型的堰塞坝稳定性预测模型SSA-Adam-BP.该模型通过网格搜索法选取确定模型结构的最佳超参数组合,进而利用交叉验证和绘制ROC曲线的方式分别对采用...     

基于卷积神经网络的福建省区域滑坡灾害预警模型

福建省滑坡灾害频发,开展区域尺度上的滑坡灾害预警是防灾减灾的重要手段,但由于滑坡成灾机理复杂,传统的区域滑坡预警方法存在精度不足等问题。深度学习是指通过构建神经网络模型进行特征的提取、抽象、表示与学习的技术,是机器学习的一种。卷积神经网络作为一种经典的深度学习算法,具有比传统机器学习更强大的分类能力与表征能力。文章以福建省为研究区,将卷积神经网络引入滑坡灾害预警领域,构建福建省区域滑坡预警模型,过程及结果如下：（1）采用SMOTE优化算法对2010—2018年福建省滑坡灾害样本库进行优化,扩充正样本的个数,将正负样本比例从1∶3.4扩充到1∶2,样本总量达到18040个;（2）构建卷积神经网络模型结构,模型结构包括一个输入层、两个卷积层、两个最大池化层和一个全连接层以及一个输出层;（3）使用卷积神经网络对优化后的样本（2010—2018年样本的80%作为训练集）进行训练,并用贝叶斯优化算法优化模型超参数,得到福建省区域滑坡预警模型;（4）以2010—2018年样本的20%作为测试集对模型进行测试,采用混淆矩阵、ROC曲线进行模型测试,结果显示模型准确度为0.96～0.97,AUC值达到0.977,模型精度与泛化能力良好;（5）以2019年汛期滑坡灾害实况作为正样本,通过时空采样的方法采集负样本,构建2019年区域滑坡样本校验集（样本数603个）,对模型进行进一步实况校验,采用混淆矩阵、ROC曲线进行模型校验,结果显示模型准确度为0.75～0.85,AUC值为0.852。虽然仅用了2019年汛期的滑坡实况样本进行校验,但也达到较好的效果。将卷积神经网络算法应用到区域滑坡预警中,为建立区域滑坡预警模型提供了一种新的途径,初步校验表明,模型效果良好,今后将在福建省对模型进行进一步的应用与校验。     

An evolutionary approach for spatial prediction of landslide susceptibility using LiDAR and symbolic classification with genetic programming

Natural Hazards - This research examines the potential of spatial prediction of landslide susceptibility by implementing an evolutionary approach using symbolic classification with genetic...     

非饱和土吸力预测的进化神经网络方法

利用遗传算法和神经网络,基于不同类型,不同条件下非饱和土的吸力测试数据,建立了一种以含水量为主要因素,耦合密度、初始含水量、先期固结压力、孔隙比5个因素的吸力预测模型。预测结果分析表明,所建的模型能很好地拟合试验结果,从而,验证了该模型的合理性和可行性。     

An artificial neural network for groutability prediction of permeation grouting with microfine cement grouts

The use of microfine cements in permeation grouting has been growing as a strategy in geotechnical engineering because it usually provides improved groutability (N). One of the major challenges of using microfine cement grouts is the ability to estimate the N within a reasonable level of error. The suitability of traditional groutability prediction formulas, which are mostly based on the grain-size of the soil and the grout, is questionable for semi-nanometer scale grout. This study first investigated the accuracy of the current formulas; we found that the accuracy ranges from 45% to 68%, a level that is not adequate for practical engineering. An alternative approach, based on a Radial Basis Function Neural Network (RBFNN), was developed. RBFNN provides a prediction with a 95.8% accuracy within a short time frame. Several parameters were considered in our proposed network; besides the grain-size of the soil (D₁₀/D₁₅), other important parameters included the void ratio (e), the fines content (FC), the uniformity coefficient (C_u), the coefficient of gradation (C_z) and the water-to-cement ratio (w/c). A total of 240 in situ data samples were collected to support the training and testing of the network. After finding a good correlation between the field observation and the RBFNN output, it was concluded that RBFNN is a suitable and reliable tool to predict the outcome of permeation grouting when microfine cement grout is used.     

Combining neural network with fuzzy,certainty factor and likelihood ratio concepts for spatial prediction of landslides

The landslide studies can be categorized as pre- and postdisaster studies. The predisaster studies include spatial prediction of potential landslide zones known as landslide susceptibility zonation (LSZ) mapping to identify the areas/locales susceptible to landslide hazard. The LSZ maps provide an assessment of the safety of existing habitations and infrastructural/functional elements and help plan further developmental activities in the hilly regions. Landslides are one of the natural geohazards that affect at least 15% of land area of India. Different types of landslides occur frequently in geodynamical active domains of the Himalayas. In India, various techniques have been developed and adopted for LSZ mapping of different regions. However, the technique for LSZ mapping is not yet standardized. The present research is an attempt in this direction only. In our earlier work (Kanungo et al. 2006), a detailed study on conventional, artificial neural network (ANN)- black box-, fuzzy set-based and combined neural and fuzzy weighting techniques for LSZ mapping in Darjeeling Himalayas has been documented. In this paper, other techniques such as combined neural and certainty factor concept along with combined neural and likelihood ratio techniques have been assessed in comparison with combined neural and fuzzy technique for the preparation of LSZ maps of the same study area in parts of Darjeeling Himalayas. It is observed from the present study that the LSZ map produced using combined neural and fuzzy approach appears to be the most accurate one as in this case only 2.3% of the total area is found to be categorized as very high susceptibility zone and contains 30.1% of the existing landslide area. This approach can serve as one of the key objective approaches for spatial prediction of landslide hazards in hilly terrain.     

Earthquake-induced landslide prediction using back-propagation type artificial neural network: case study in northern Iran

Natural Hazards - Landslides can cause extensive damage, particularly those triggered by earthquakes. The current study used back propagation of an artificial neural network (ANN) to conduct risk...     

Study on spatial prediction and time forecast of landslide

Landslide is one of natural hazards in mountainous regions, which has resulted in a large quantity of casualties and property losses and also has absorbed high attention of the researchers and government. A considerable amount of research has been carried out in the past 30 years in the landslide field. In this paper, the contribution and existing problems on landslide are analyzed and summarized in the previous studies. Spatial prediction and zonation of the regional landslide are developed by using information content model that is a new method, with the example of landslide in Xincheng District of Badong County. On the other hand, by learning from the forecast theories and methods of earthquake forecast, probability of excess for landslide that will take place in the studied area is calculated quantitatively in next 5 and 10 years. All the calculated results are mainly accordant with the regional fact. Therefore, it may provide scientific data for landslide prevention and reduction as well as landslide management. Based on the achievement obtained in this study, it was found that 29.11% of the total area was prone to landslide due to the adverse effects of topography, reservoir water in the leading edge of bank, and improper land use. At the same time, the theory of spatial prediction and probability of excess will be example and reference for the other region of China or the world.     

基于多元非线性回归和BP神经网络的滑坡滑动距离预测模型研究

滑坡灾害严重威胁着人类的生命财产安全, 对土地资源造成了一定影响。滑坡滑动距离直接表明了滑坡的冲击、堆积范围大小, 是估算滑坡受灾面积、评估滑坡潜在风险的重要参数, 也是滑坡防灾减灾工作中需要重点关注的指标。为了更准确高效地预测滑坡危害范围, 分别采用多元非线性回归和BP神经网络2种模型对影响滑坡滑动距离的因子进行了评估和建模, 并对天水地区的滑坡实例进行研究。研究结果表明, 2种模型均可用于滑坡滑动距离的预测。相较而言, BP神经网络的预测结果与实际情况有更高的拟合度, 准确性更高。     

利用BP神经网络进行水库滑坡变形预测

滑坡变形监测与预测是滑坡预警预报中一种非常重要的途径。文章首先简单介绍了神经网络的基本原理和学习算法,然后利用某水库滑坡24期的GPS地表位移监测数据及其诱发因素即水库水位、降雨等资料,采用BP神经网络模型对该水库滑坡变形进行建模,最后将6期水库水位、降雨等资料输入模型进行滑坡变形预测,结果表明预测结果与实测数据符合性好,总体上能较好反映变形趋势。     

Validation of an artificial neural network model for landslide susceptibility mapping

The aim of this study was to validate an artificial neural network model at Youngin, Janghung, and Boeun, Korea, using the geographic information system (GIS). The factors that influence landslide occurrence, such as the slope, aspect, curvature, and geomorphology of topography, the type, material, drainage, and effective thickness of soil, the type, diameter, age, and density of forest, distance from lineament, and land cover were either calculated or extracted from the spatial database and Landsat TM satellite images. Landslide susceptibility was analyzed using the landslide occurrence factors provided by the artificial neural network model. The landslide susceptibility analysis results were validated and cross-validated using the landslide locations as study areas. For this purpose, weights for each study area were calculated by the artificial neural network model. Among the nine cases, the best accuracy (81.36%) was obtained in the case of the Boeun-based Janghung weight, whereas the Janghung-based Youngin weight showed the worst accuracy (71.72%).     

海量监测数据下分布式BP神经网络区域滑坡空间预测方法

提出BP神经网络的分布式区域滑坡预测方法,算法设计在大数据分布式处理平台Spark下实现,通过构造包含均方误差和L2正则化的代价函数,提高运算实时性和算法泛化能力。统计影响滑坡评价因子的量化指标和定义监测剖面危险级别评价值,并进行评价因子特征选取,用于三峡库区忠县区域9个滑坡11年月监测海量数据挖掘,对研究区所有滑坡监测剖面每月进行危险级别评价,实现以月为周期的区域滑坡危险程度空间预测。试验表明,采用所述方法得到的拟合精度、准确度、效率均比梯度提升决策树、随机森林算法好,预测的滑坡危险级别准确,该方法可作为区域滑坡空间预测的一种新思路。     

考虑环境变量作用的滑坡变形动态灰色-进化神经网络预测研究

滑坡变形预测对于指导灾害的预防工作、保护人民的生命和财产安全具有重大实用价值。从系统论观点出发,结合岩土体流变理论和时序分析原理,在深入研究影响滑坡变形的主控环境变量基础上,将位移时序分解为趋势项和偏离项。采用灰色系统模型提取位移时序趋势项,结合遗传算法和人工神经网络建立起进化神经网络模型,逼近主控环境变量与位移偏离项之间的非线性关系。根据蠕变阶段和变形对环境变量响应情况,实时调整模型,建立起滑坡变形预测的动态灰色-进化神经网络（GM-ENN）模型。将此预测思路和方法应用于三峡库区某滑坡变形预测研究中,证实了模型的有效性和实用性,显示了动态预测的重要性。     

红石包滑坡滑带土强度参数的神经网络预测

在分析影响滑带土强度因素的基础上,建立了滑带土强度参数的BP神经网络模型,预测滑带土在不同含水率下c,φ值的变化规律,尤其足当红石包滑坡前缘、后缘地质条件差异较大时,找出可能的工况匹配,可以为滑坡稳定性评价提供可靠依据,克服了c,φ值按峰值折减的主观性.应用表明:该模型精度很高,有应用前景.     

Use of artificial neural network for spatial rainfall analysis

In the present study, the precipitation data measured at 23 rain gauge stations over the Achaia County, Greece, were used to estimate the spatial distribution of the mean annual precipitation values over a specific catchment area. The objective of this work was achieved by programming an Artificial Neural Network (ANN) that uses the feed-forward back-propagation algorithm as an alternative interpolating technique. A Geographic Information System (GIS) was utilized to process the data derived by the ANN and to create a continuous surface that represented the spatial mean annual precipitation distribution. The ANN introduced an optimization procedure that was implemented during training, adjusting the hidden number of neurons and the convergence of the ANN in order to select the best network architecture. The performance of the ANN was evaluated using three standard statistical evaluation criteria applied to the study area and showed good performance. The outcomes were also compared with the results obtained from a previous study in the area of research which used a linear regression analysis for the estimation of the mean annual precipitation values giving more accurate results. The information and knowledge gained from the present study could improve the accuracy of analysis concerning hydrology and hydrogeological models, ground water studies, flood related applications and climate analysis studies.     

南江滑坡群体积的BP神经网络模型与预测

基于南江县境内244个典型土质滑坡统计样本,利用BP神经网络模型,采用3种不同的方案(基于不同的评价参数)对滑坡体积进行预测。方案一选取坡高、坡度、坡向、高程、植被覆盖率、岩层倾向、岩层倾角等7项评价参数;方案二选取坡高、坡度、坡向、岩层倾向、岩层倾角等5参数;方案三选取坡高、坡度、坡向等3参数。研究结果表明:3种方案建立的BP神经网络模型都具有较高的可靠性,其预测结果都可以较好地逼近真实滑坡体积值,BP神经网络能有效应用到滑坡体积预测中;3种方案预测值与实际值基本吻合,且两者间的相关系数分别为0.87083,0.90826,0.86119,评价参数的合理选择对滑坡体积预测的准确性有着重要的影响;方案二的相关系数最高,其预测准确性最好,这表明坡高、坡度、坡向、岩层倾向、岩层倾角是影响滑坡体积的重要因素,植被覆盖率和高程为其次要影响因素。     

Influences of spatial distribution of soil thickness on shallow landslide prediction

Shallow landslides usually occur during hevy rainfall and result in casualties and property losses. Thus, the possible locations where landslides are likely to occur must be identified in advance in order to avoid or reduce the harm they cause. When performing a slope-instability analysis, soil thickness is an important factor; however, soil thickness information from landslide-prone areas is rarely obtained. The objective of this study is to realize the influences of spatial distribution of soil thickness on shallow landslide prediction. Three different spatial soil-thickness distributions were applied to perform a slope-instability analysis, and uniform-distributed soil thicknesses from 0.4 m to 2.0 m were also applied for comparison. Geomorphologic information and hydrological records from a landslide-prone area in southern Taiwan were collected. Results show that the spatial distribution of soil thickness related to wetness index provides a reasonable estimation in order to avoid an over-prediction for landslide-prone areas or an under-prediction for stable areas. The analytical procedure used in this study is a simple way for assessing hillslope instability for shallow landslide prediction.