基于多算法参数优化与SVR模型的白水河滑坡位移预测

滑坡预测对于减轻地质灾害的危害十分重要,但对科学研究却很有挑战性。基于变形特征和位移监测数据,建立了三峡库区白水河滑坡的时间序列加法模型。在模型中,累计位移分为3个部分:趋势、周期和随机项,解释了由内部因素（地质环境,重力等）、外部因素（降雨,水库水位等）、随机因素（不确定性）共同作用的影响。在对位移数据进行统计分析后,提出了一个3次多项式模型对趋势项进行学习,并利用多算法寻优的支持向量回归机（SVR）模型对周期项进行训练与预测。结果表明,在预测精度上,基于时间序列与遗传算法-支持向量回归机（GA-SVR）耦合的位移预测模型要明显优于网格寻优（GS）以及粒子群算法（PSO）优化的支持向量回归机模型。因此,GA-SVR模型在滑坡位移预测方面可以得到较好的应用。在“阶跃型”滑坡位移预测中,GA-SVR将具有广阔的应用前景。     

Displacement prediction of step-like landslide by applying a novel kernel extreme learning machine method

Landslide displacement prediction is an essential component for developing landslide early warning systems. In the Three Gorges Reservoir area (TGRA), landslides experience step-like deformations (i.e., periods of stability interrupted by abrupt accelerations) generally from April to September due to the influence of precipitation and reservoir scheduled level variations. With respect to many traditional machine learning techniques, two issues exist relative to displacement prediction, namely the random fluctuation of prediction results and inaccurate prediction when step-like deformations take place. In this study, a novel and original prediction method was proposed by combining the wavelet transform (WT) and particle swarm optimization-kernel extreme learning machine (PSO-KELM) methods, and by considering the landslide causal factors. A typical landslide with a step-like behavior, the Baishuihe landslide in TGRA, was taken as a case study. The cumulated total displacement was decomposed into trend displacement, periodic displacement (controlled by internal geological conditions and external triggering factors respectively), and noise. The displacement items were predicted separately by multi-factor PSO-KELM considering various causal factors, and the total displacement was obtained by summing them up. An accurate prediction was achieved by the proposed method, including the step-like deformation period. The performance of the proposed method was compared with that of the multi-factor extreme learning machine (ELM), support vector regression (SVR), backward propagation neural network (BPNN), and single-factor PSO-KELM. Results show that the PSO-KELM outperforms the other models, and the prediction accuracy can be improved by considering causal factors.     

Using an extreme learning machine to predict the displacement of step-like landslides in relation to controlling factors

In the evolution of landslides, besides the geological conditions, displacement depends on the variation of the controlling factors. Due to the periodic fluctuation of the reservoir water level and the precipitation, the shape of cumulative displacement-time curves of the colluvial landslides in the Three Gorges Reservoir follows a step function. The Baijiabao landslide in the Three Gorges region was selected as a case study. By analysing the response relationship between the landslide deformation, the rainfall, the reservoir water level and the groundwater level, an extreme learning machine was proposed in order to establish the landslide displacement prediction model in relation to controlling factors. The result demonstrated that the curves of the predicted and measured values were very similar, with a correlation coefficient of 0.984. They showed a distinctive step-like deformation characteristic, which underlined the role of the influencing factors in the displacement of the landslide. In relation to controlling factors, the proposed extreme learning machine (ELM) model showed a great ability to predict the Baijiabao landslide and is thus an effective displacement prediction method for colluvial landslides with step-like deformation in the Three Gorges Reservoir region.     

基于时间序列与人工蜂群支持向量机的滑坡位移预测研究

总结以往滑坡预测方法存在的诸多不足,针对滑坡监测位移-时间曲线特点,本文提出了一种基于时间序列的人工蜂群算法（ABC）与支持向量回归机（SVR）相结合的滑坡位移预测方法。以三峡库区白水河滑坡为例,通过对滑坡位移、降雨、库水位等因素的分析,研究影响滑坡位移变化的因素。用时间序列加法模型和移动平均法将滑坡位移分解为趋势项和周期项。以多项式最小二乘法拟合滑坡位移趋势项,用人工蜂群支持向量机模型对滑坡位移周期项进行训练和预测。通过灰色系统关联分析法计算多项因子与滑坡位移周期项之间的关联性。最终的滑坡总位移预测值为周期项预测值与趋势项预测值之和。与BP神经网络、PSO-SVR模型方法相比,该方法在滑坡位移预测中有更高的精度,在防灾减灾工作中有较好的推广应用前景。     

考虑监测离群值朱家店滑坡位移预测研究

滑坡位移预测预报是滑坡防灾减灾的重要组成部分,提高滑坡位移预测的准确性与精确度是该项研究的重点与难点。本文在滑坡位移预测中考虑了监测样本的离群值,通过忽略、指定与修正离群值3种方式,研究滑坡位移预测样本离群值的最优处理方式。以三峡库区朱家店滑坡为例,基于ARIMA(p,d,q)模型,分别对累积位移与位移速率时间序列开展了预测研究。研究结果表明:修正离群值的预测结果介于忽略和指定离群值两者之间,更适用于存在监测离群值的滑坡位移预测;对于ARIMA模型,更适合采用位移速率进行预测预报;使用位移速率时间序列ARIMA(1,0,1)并修正离群值的预测结果为:2016年和2017年6月份滑坡前缘GP3"阶跃"位移分别为79. 0 mm和70. 2 mm,截止2017年8月,GP3累积位移将达1647. 7 mm。     

滑坡滑动面位置的确定及超前预测

依据滑坡深部位移信息,分析滑坡“B”型、“D”型、“r”型变形特征，确定滑坡滑动面位置，提出了滑坡滑动面位置超前预测的位移矢量法。为系统分析及超前预测滑坡滑动面位置提供依据，为深入研究滑坡运动特征提供信息。     

基于APSO-SVR-GRU模型的白水河滑坡周期项位移预测

滑坡周期项位移的预测,是研究地质灾害中滑坡变形至关重要的一步。由于单一模型易受偶然因素影响,且无法充分利用有效信息,导致其预测精度不高,适用性不强。基于此,文中提出了一种结合自适应粒子群算法（APSO）、支持向量机回归算法（SVR）、门控神经网络算法(GRU)的组合模型。该模型通过自适应粒子群优化算法对支持向量机回归算法进行参数寻优,确定最优参数组合,然后利用最小二乘法对APSO-SVR模型与GRU模型赋权建立最优权重比组合模型。以三峡白水河滑坡作为研究对象,选取降雨量、库水位及位移量作为周期项位移的影响因子,对模型进行训练验证,结果表明:在白水河滑坡周期项位移预测中,文中所提出的APSO-SVR-GRU组合模型与单一模型相比,具有更高的预测精度和稳定性。     

基于时间序列分解和多变量混沌模型的滑坡阶跃式位移预测

三峡库区某些库岸滑坡在强降雨、库水位涨落等诱发因素影响下,其位移时间序列表现出阶跃式变化特征且可能存在混沌特性.但目前常用于滑坡位移预测的混沌模型,均建立在单变量混沌理论的基础之上.且已有的考虑了诱发因素的常规多变量模型,大都采用经验性的方法来选取输入变量;常规多变量模型对滑坡位移序列的非线性特征,及其与诱发因素间的动态响应关系缺乏数学理论上的深入分析.因此,提出一种基于指数平滑法、多变量混沌模型和极限学习机（extreme learing machine,ELM）的滑坡位移组合预测模型.指数平滑多变量混沌ELM模型首先对滑坡累积位移序列的混沌特性进行识别;然后用指数平滑法对累积位移进行预测,得到趋势项位移,并用累积位移减去趋势项位移得到剩余的波动项位移;之后对波动项位移及降雨量、库水位变化量这3个因子进行多变量相空间重构,并用ELM模型对多变量重构后的波动项位移进行预测;最后将预测得到的趋势项和波动项位移值相加,得到最终的累积位移预测值.以三峡库区白水河滑坡ZG93监测点的累积位移作为实例进行分析,并将模型与指数平滑多变量混沌粒子群-支持向量机（PSO-SVM）模型、指数平滑单变量混沌ELM模型作对比.结果表明滑坡位移序列存在混沌特性,模型能有效预测滑坡位移,其预测效果优于对比模型.且本文模型从混沌理论的角度将波动项位移与降雨量、库水位变化量的动态响应关系进行综合分析,更能反映滑坡位移系统演化的物理本质.      

基于H-P滤波法、ARIMA和VAR模型的库区滑坡位移综合预测

受库水位涨落及降雨等影响,库区滑坡位移表现出明显的周期性。基于位移时间序列分析,将滑坡监测位移分解为趋势项与周期项之和。趋势项反映滑坡变形的长期趋势,其主要受滑坡本身地质结构等因素影响。周期项反映滑坡变形的波动性,其主要受外部因素影响。以三峡库区巫山塔坪滑坡为例,考虑长江水位与降雨量影响,采用H-P滤波法从滑坡位移中分解出趋势项及周期项,利用差分自回归滑动平均模型（ARIMA）对趋势项进行平稳处理并计算趋势项预测值,利用向量自回归模型（VAR）计算周期项预测值。趋势项预测值与周期项预测值之和为滑坡位移预测值。与实际监测值及多种方法分析比较,表明综合预测所得结果能较好反映滑坡变形的趋势性和波动性,位移预测效果较好。     

基于时间序列与混合核函数SA-SVR的滑坡位移预测模型研究

本文针对阶跃型滑坡变形定量预测困难,提出一种基于时间序列分解与混合核函数SA-SVR的滑坡位移预测模型.首先基于时间序列分解原理,反复使用指数平滑法将滑坡累积位移分解为趋势项位移和周期项位移,使分解后的趋势项位移较平滑且能保证周期项位移的预测精度.同时针对多项式预测容易过拟合造成预测值偏离真实值的问题,采用K-flod...     

基于优化经验模态分解和聚类分析的滑坡位移智能预测研究

针对三峡库区"阶跃式"滑坡的变形特征,提出了一种新的滑坡位移预测方法。以白水河滑坡ZG118和XD-01监测点位移数据为例,采用基于软筛分停止准则的经验模态分解(SSSC-EMD)将累计位移-时间曲线和影响因子时间序列自适应地分解为多个固有模态函数(IMF),并采用K均值(K-Means)聚类法对其进行聚类累加,得到有物理含义的位移分量(趋势性位移、周期性位移以及随机性位移)和影响因子分量(高频影响因子和低频影响因子)。使用最小二乘法对趋势性位移进行拟合预测;采用果蝇优化-最小二乘支持向量机(FOA-LSSVM)模型对周期性位移和随机性位移进行预测。将各位移分量预测值进行叠加处理,实现滑坡累计位移的预测。研究结果表明,所提出的(SSSC-EMD)-K-Means-(FOA-LSSVM)模型能够预测"阶跃式"滑坡的位移变化规律,且预测精度高于传统的支持向量机回归(SVR)、最小二乘支持向量机(LSSVM)模型;并通过改变训练集长度,进行单因素分析,发现其与预测精度之间呈正相关关系。     

基于非平稳时间序列分析的滑坡变形预测

滑坡的位移监测资料通常可用来预测滑坡的变形发展趋势,位移的发展反映了滑坡的变形过程.为了预测在现有条件持续情况下的滑坡变形趋势,将滑坡位移监测数据视为非平稳时间序列,应用时间序列分析方法,建立了滑坡变形趋势的预测模型.以三峡库区秭归县白水河滑坡为例,通过对变形预警区监测点位移实测时间序列的分析,取监测点ZG93和XD-04为代表,建立了时间序列预测模型,从第17个月开始向前做6步预测,分析预测曲线与实测曲线之间的关系,并计算预测误差,结果显示除个别数据点之外,预测误差均在±9%以内,曲线吻合较好,说明所建模型效果良好,从而为判断白水河滑坡未来的变形发展趋势提供了可靠的理论依据.     

联合多种数据驱动建模方法的滑坡位移预测研究

边坡位移是滑坡演化的宏观体现,分析并预测滑坡位移发展态势对于防灾减灾具有重要意义。由于滑坡位移曲线具有明显的非线性特征,单一模型往往难以刻画其非线性与复杂性。为发展一种普遍适用于滑坡位移的预测方法,提出了一种联合多种数据驱动模型的新方法。该方法根据时间序列分析理论,将滑坡位移序列分解为趋势项和周期项,趋势项采用并联型灰色神经网络处理,周期项则采用人工蜂群算法（ABC）优化后的极限学习机模型（ELM）处理,从而充分应用各种模型的优点。以三峡库区白水河和八字门滑坡为例,对位移数据进行分析处理后,灰色神经网络模型预测其趋势性位移,改进后的极限学习机模型对周期性位移进行训练及预测。结果表明:在预测精度上,优化后的极限学习机模型准确度高于极限学习机模型及小波神经网络等方法,提出的灰色神经网络与ABC-ELM的组合模型可作为实际工程的一个参考。     

基于一元线性回归模型八字门滑坡累积位移预测

我国滑坡灾害发生频繁,但滑坡的变形预测预报一直是难题,因此每年都因滑坡的变形破坏导致重大的人员伤亡和财产损失。以三峡库区八字门滑坡为研究对象,基于十多年的监测数据分析,研究分析了该滑坡的变形特征:八字门滑坡变形的主要影响因素为降雨和库水位下降,并且累积位移曲线具有"阶跃型"的变形特征。当外界因素去除或者减小的情况下,累积位移-时间曲线将变得平稳。根据此特性,选取每年变形曲线"阶跃段"（6~8月份）的监测数据,以累积位移为目标函数,基于一元线性回归模型,对八字门滑坡2004年到2017年同期的滑坡监测数据进行分析。结果表明:一元线性回归模型能够很好地模拟八字门滑坡"阶跃段"的变形过程,此变形阶段累积位移与时间呈线性关系,直线斜率基本相同。根据此线性关系,对滑坡的累积位移进行了预测,结果表明与实际监测数据相比较,预测误差在±5 mm以内,相对误差在1%以下,精度可以满足滑坡监测预警要求,可以为八字门滑坡的防治工作提供参考。     

多元时间序列分析的滑坡演化阶段划分

滑坡的时间-位移曲线一般具有3个阶段特征,即初始变形阶段、等速变形阶段和加速变形阶段,不同演化阶段加速度具有不同的变化特点.目前往往是依据对加速度曲线特征的分析来人为划分演化阶段,缺少相应的理论支持和定量计算.针对上述问题,选取月降雨量、月库水位高程变化量对滑坡的累计位移建立多因素的时间序列预测模型.然后利用Chow分割点检验理论,以所建模型中F和LR统计量最大值点作为分割点对滑坡演化阶段进行划分.以新滩滑坡和三峡库区白水河滑坡为例,利用累计位移、降雨及库水位变化数据进行计算验证.结果表明,对多元时间序列模型进行Chow分割点检验可对滑坡的演化阶段进行准确划分,为滑坡的临滑预警预报提供重要判据.      

Displacement prediction in colluvial landslides, Three Gorges Reservoir, China

The prediction of active landslide displacement is a critical component of an early warning system and helps prevent property damage and loss of human lives. For the colluvial landslides in the Three Gorges Reservoir, the monitored displacement, precipitation, and reservoir level indicated that the characteristics of the deformations were closely related to the seasonal fluctuation of rainfall and reservoir level and that the displacement curve versus time showed a stepwise pattern. Besides the geological conditions, landslide displacement also depended on the variation in the influencing factors. Two typical colluvial landslides, the Baishuihe landslide and the Bazimen landslide, were selected for case studies. To analyze the different response components of the total displacement, the accumulated displacement was divided into a trend and a periodic component using a time series model. For the prediction of the periodic displacement, a back-propagation neural network model was adopted with selected factors including (1) the accumulated precipitation during the last 1-month period, (2) the accumulated precipitation over a 2-month period, (3) change of reservoir level during the last 1 month, (4) the average elevation of the reservoir level in the current month, and (5) the accumulated displacement increment during 1 year. The prediction of the displacement showed a periodic response in the displacement as a function of the variation of the influencing factors. The prediction model provided a good representation of the measured slide displacement behavior at the Baishuihe and the Bazimen sites, which can be adopted for displacement prediction and early warning of colluvial landslides in the Three Gorges Reservoir.     

黄河小浪底斜坡变形的“反迭瓦”式构造机制研究

针对小浪底地区斜坡中普遍发育的“返迭瓦”式构造,结合该区的工程地质条件,对其成因机制进行分析。提出了“反迭瓦”式构造的形成是由于边坡蠕变形的结果,并非造成滑坡的原因。     

基于数值模拟的三峡库区典型堆积层滑坡变形预测方法

目前对堆积层滑坡的变形预测大多基于数学模型或方法,忽略了引起滑坡位移显著变化的动力外因及滑坡自身的地质特征,因此,预报准确度和可信度较低。以三峡库区典型堆积层滑坡--鹤峰场镇滑坡为例,通过4组主要控制因素科学组合构建了滑坡的基本地质模型;以此为基础,重点考虑引起滑坡发生变形的库水作用动力因素,建立滑坡的数值-力学模型。通过实际监测点的变形监测结果与数值-力学模型中模型监测点的变形进行拟合分析,获取了实际时间与数值-力学模型中时步的等效关系;基于时间-时步等效关系及三峡水库设计水位调度曲线,得到了不同时步水位的波动特征;通过时步的外延,并在相应的时步段对数值-力学模型施加等效时间的库水作用,预测了滑坡在未来库水位变动条件下的变形。该预测方法既考虑了滑坡的工程地质模型又考虑了地下水作用效应,克服了纯数学方法预测的不足。     

基于光纤监测和PSO-SVM模型的马家沟滑坡深部位移预测研究

滑坡位移预测效果一方面取决于预测模型的优劣,另一方面取决于野外监测数据的质量。针对目前滑坡常规监测技术与评价方法的不足,本文采用光纤监测技术、监测数据与PSO-SVM预测模型相结合的评价方法,对三峡马家沟Ⅰ号滑坡的深部位移进行了预测;通过对320个滑坡深部位移光纤监测数据分析,基于时间序列法,将滑坡位移分为趋势性位移和波动性位移;趋势性位移采用拟合法进行预测,波动性位移采用PSO-SVM模型进行预测;最后将趋势项和波动项位移预测值叠加得到累积位移的预测值。研究结果表明,PSO-SVM模型对波动性位移预测的均方根误差0.51 mm,平均绝对百分误差0.37 mm,能准确预测滑坡波动项位移;累积位移预测值与实测值的相关系数为0.98,均方根误差为0.54 mm,预测效果较好,可以用来对滑坡深部位移进行短期预测。     

Probabilistic forecasting of landslide displacement accounting for epistemic uncertainty: a case study in the Three Gorges Reservoir area,China

Accurate and reliable displacement forecasting plays a key role in landslide early warning. However, due to the epistemic uncertainties associated with landslide systems, errors are unavoidable and sometimes significant in traditional methods of deterministic point forecasting. Transforming traditional point forecasting into probabilistic forecasting is essential for quantifying the associated uncertainties and improving the reliability of landslide displacement forecasting. This paper proposes a hybrid approach based on bootstrap, extreme learning machine (ELM), and artificial neural network (ANN) methods to quantify the associated uncertainties via probabilistic forecasting. The hybrid approach consists of two steps. First, a bootstrap-based ELM is applied to estimate the true regression mean of landslide displacement and the corresponding variance of model uncertainties. Second, an ANN is used to estimate the variance of noise. Reliable prediction intervals (PIs) can be computed by combining the true regression mean, variance of model uncertainty, and variance of noise. The performance of the proposed hybrid approach was validated using monitoring data from the Shuping landslide, Three Gorges Reservoir area, China. The obtained results suggest that the Bootstrap-ELM-ANN approach can be used to perform probabilistic forecasting in the medium term and long term and to quantify the uncertainties associated with landslide displacement forecasting for colluvial landslides with step-like deformation in the Three Gorges Reservoir area.