首页 > 范文大全 > 正文

拱坝变形监测预报的随机森林模型及应用

开篇:润墨网以专业的文秘视角,为您筛选了一篇拱坝变形监测预报的随机森林模型及应用范文,如需获取更多写作素材,在线客服老师一对一协助。欢迎您的阅读与分享!

摘要:大坝变形预报对大坝运行安全评估起着至关重要的作用。传统模型预报精度不够、模拟效果不稳定;若大坝变形数据有异常值时,传统机器算法模型识别和处理异常值的灵活性很小,导致预报结果有偏差。为了解决这些问题,首次将随机森林算法运用到大坝变形监测领域,将大坝测点根据随机森林相似性矩阵分成若干个子集,针对每一个子集建立随机森林预测模型,分区建立预测模型更符合工程实际情况。选取拱坝变形作为研究对象,验证所建模型的适用性。结果表明,根据随机森林的相似性矩阵对大坝各测点的分区情况符合物理和工程实际意义,对各分区子集测点利用随机森林模型建立的预测模型,与支持向量机、BP神经网络模型相比,预测结果精度较高、模型稳定性好,为大坝变形监测提供了新思路。

关键词:拱坝变形;监控模型;监测点分区;随机森林;变形预测

中图分类号:TU196.1文献标志码:A文章编号:

16721683(2016)06011606

Random forest model and application of arch dam′s deformation monitoring and prediction

LUO Hao1,2,GUO Shengyong2,BAO Weimin1

1.College of Water Resources and Hydrology,Hohai University,Nanjing 210098,China;

2.Yalong River Hydropower Company Ltd,Chengdu 610051,China)

Abstract:Dam deformation prediction plays an important role in the safety assessment of dam operation.Traditional models lack forecasting precision and the simulation effect is not stable enough.Besides,if abnormal values of dam deformation exist,traditional machine algorithm model lacks the flexibility of dealing with these abnormal data,which will lead to the deviation of the forecasting results.In order to solve these problems,random forest algorithm was introduced to the field of dam deformation monitoring for the first time.Similarity matrix of random forest was applied to divide dam deformation monitoring points into several parts.Random forests prediction model was established for each part,which will avoid the defects of traditional models such as modeling of single point or using the same model for all deformation monitoring points.Establishing forecasting model for different parts of dam was more in line with engineering practice.Deformation data of one arch dam was analyzed and the feasibility of random forest model was verified.The results showed that partition of dam deformation points based on similarity matrix of random forest conformed to the physical and engineering practical pared with support vector machine and BP neural network model,the prediction model of random forests for each part had the higher prediction precision and stability,which provided a new approach in the area of dam safety monitoring.

Key words:arch dam deformation;monitoring model;partitions of monitoring points;random forests;deformation prediction

国内外普遍将大坝变形监测[12]作为主要的监测项目,大坝受各种复杂因素的影响,变形值是反映其运行状态的最直观的表征。根据大坝变形的原型观测资料建立准确的预测模型,对大坝位移进行预测,能及时发现大坝的异常变化,采取措施防止事故发生。因此,大坝变形预报对大坝运行的安全评估起着至关重要的作用。目前应用较多数学模型主要包括统计模型[23]、确定性模型[45]和混合型模型[56],这些模型在一定程度上可以揭示监测值和影响量之间定性和定量关系,但由于影响大坝位移的因素复杂,传统的方法受变量多重共线性的影响或模型参数的选取不恰当,使得模型精度下降。近年来,一些学者将新兴的机器算法如人工神经网络[78]、遗传算法[9]、蚁群算法[1011]、支持向量机[1213]等算法建立大坝监控模型,[JP2]但这些监控模型的研究和应用尚未达到完善的程度,每种方法都存在一定程度上的优缺点。另外,由于大坝具有整体性,布置在坝体和坝基的各测点之间存在差异性和关联性,目前位移监控模型还是以单测点为主,单测点位移监控模型存在很大的局限性,不能反映大坝整移变化情况;多维多测点模型较单测点位移模型更符合工程实际情况,但由于多测点位移监控模型[14]中待定参数较多,要达到一定的变形分析和预报精度,对原型观测数据要求较高,给建模造成很大难度,在实际工程中的应用并不广泛。随机森林(Random Forest,RF)[15]算法是由Breiman在2001年提出的一种新的机器学习技术,随机森林模型能有效地分析非线性、具有高度共线性和相互影响的数据,不需要提前假定模型的数学形式,该算法在在生物学[1617]、土壤学[1819]、医学[20]等领域已经得到了一定的应用,但在大坝安全监测领域应用几乎没有。此外,相似性矩阵是随机森林算法的重要的分析工具之一,尝试利用随机森林算法的相似性矩阵来表征大坝各位移监测点之间的相似性关系,基于这种相似关系,将大坝测点分区,分别对各区建立随机森林回归预测模型。随机森林算法预测精度高、对于异常值的处理和噪声方面具有很大的优势,不易出现过度拟合的线性,能有效处理复杂变量间的共线性问题,该算法为大坝安全监控提供了一种新思路。

4结论

(1)拱坝不同区域变形性态差异较大,若用统一个模型去建模,忽略了各测点变形之间的差异性和关联性。本文充分利用随机森林模型具有的特性,建立了各测点间的相似度矩阵,将大坝测点按照其变形的相似性进行分类,分别对各位移测点子集进行建模。由分区结果来看,符合大坝运行规律。

(2)对各子集测点利用随机森林模型进行建模,与支持向量机、BP神经网络模型相比,预测精度较高、稳定性好。不用对数据进行预处理,利用OBB误差直接估计泛化误差,不用增加交叉验证的步骤。可作为中长期拱坝变形预测的一种新途径。

参考文献(References):

[1]吴中如.水工建筑物安全监控理论及其应用[M].高等教育出版社,2003.(WU Zhongru.Hydraulic structure safety monitoring theory and its application[M].Higher Education,2003.(in Chinese))

[2]吴中如,顾冲时.大坝原型反分析及其应用[M].江苏科学技术出版社,2000.(WU Zhongru,GU Chongshi.Dam prototype back analysis and its application[M].Jiangsu Science and Technology Press,2000.(in Chinese))

[3]李波,顾冲时,李智录,等.基于偏最小二乘回归和最小二乘支持向量机的大坝渗流监控模型[J].水利学报,2008,39(12):13901394,1400.(LI Bo,GU Chongshi,LI Zhilu,et al.Dam seepage monitoring model based on partial least squares regression and least squares support vector machine(SVM)[J].Journal of Water Conservancy.2008,39(12):13901394,1400.(in Chinese))

[4]Dmitry Vjushin.Scaling analysis of trends using DFA.Physica A Statistical Mechanics and its Applications,2001.

[5]Wu Zhongru.Deterministic models and mixed models of safety monitoring of concrete dams[J].Journal of Hydraulic Engineering ASCE,1989.

[6]刘求龙.基于有限元确定性模型大坝形变分析与研究[D].西安:长安大学,2003.(LIU Qiulong.Dam deformation analysis and research based on the finite element uncertainty model[D].Xi ′an:Chang ′an university,2003.(in Chinese))

[7]王小敏,刘小勇.大坝变形分析与预报的有限元法[J].地理空间信息,2009(5):127130.(WANG Xiaomin,LIU Xiaoyong.Dam deformation analysis and prediction based on finite element method[J].Journal of Geographic Space Information,2009(5):127130.(in Chinese))

[8]李少华,董增川.BP神经网络模型应用于大坝原型观测研究[J].水利与建筑工程学报,2004,2(2):3133.(LI Shaohua,DONG Zengchuan.BP neural network model and its application prototype observation study[J].Journal of Water Conservancy and Architectural Engineering,2004,2(2):3133.(in Chinese))

[9]Cevik A.Discussion On "Correction of soil parameters in calculation of embankment settlement using a BP network backanalysis model"[J].Engineering Geology,2008,100(34):146147.[ZK)]

[10][ZK(#]王志军,刘红彩.遗传神经网络在大坝变形预报因子重要度判定中的应用[J].水电自动化与大坝监测,2008,32(5):6971.(WANG Zhijun,LIU Hongcai.Genetic neural network′s application in the forecasting factor importance decision of dam deformation[J].Automation of Hydropower and Dam Monitoring,2008,32(5):6971.(in Chinese))

[11]李士勇.蚁群优化算法及其应用研究进展[J].计算机测量与控制,2003,11(12):911913.(LI Shiyong.Ant colony optimization algorithm and its application of research progress[J].Computer Measurement and Control,2003,11(12):911913.(in Chinese))

[12]Dervis Karaboga,Bahriye Akay.A comparative study of Artificial Bee Colony algorithm[J].Applied Mathematics and Computation.2009(1).

[13]高永刚,岳建平,石杏喜.支持向量机在变形监测数据处理中的应用[J].水电自动化与大坝监测.2005,29(5):3639.(GAO Yonggang,YUE Jianping,SHI Xingxi.Support vector machine(SVM)in the application of deformation monitoring data processing[J].Journal of Hydropower and Dam Monitoring Automation,2005,29(5):3639.(in Chinese))

[14][JP2]Marcelo Espinoza,Johan A K Suykens,Bart De Moor.Fixedsize Least Squares Support Vector Machines:A Large Scale Application in Electrical Load Forecasting[J].Computational Management Science.2006(2).

[15]赵卿.大坝变形分析多测点统计模型应用研究[D].武汉大学,2010.(ZHAO Qin.Applied research of dam deformation analysis multipoint statistical model[D].Wuhan University,2010.(in Chinese))

[16]Breiman L.Random Forests[J].Machine Learning,2001,45(1):532.(in Chinese))

[17]武晓岩,李康.基因表达数据判别分析的随机森林方法[J].中国卫生统计,2006(6):491494.(WU Xiaoyan,LI Kang.Gene expression data discriminant analysis based on method of random forest[J].China Health Statistics,2006(6):491494.(in Chinese))

[18][ZK(#]Prasad A M,Iverson L R,Liaw A.Newer classification andregression tree techniques:Bagging and random forests for ecologicalprediction[J].Ecosystems,2006,9(2):181199.

[19]Mansout K,Mutanga O,Adam E,et al.Multispectral remote sensing for mapping grassland degradation using the key indicators of grass species and edaphic factors[J].Geocarto International,2016,31(5):477491.

[20]Yan D,de Beurs,K M.Mapping the distributions of C3 and C4 grasses in the mixedgrass prairies of southwest Oklahoma using the Random Forest classification algorithm[J].International Journal of Applied Earth Observation and Geoinformation,2016,47:125138.

[21][JP2]Galaz Z,Mekyska J,Mzourek Z,et al.Prosodic analysis of neutral,stressmodified and rhymed speech in patients with Parkinson′s disease[J].Computer Method and Programs in Biomedicine,2016,127:301317.

[22]Liaw A,Wiener M.Classification and Regression by random Forest[J].R News,2002,2(3):1822.

[23]Qi Yanjun,KleinSeetharaman Judith,Bar Joseph Ziv.Random Forest Similarity for ProteinProtein Interaction Prediction from Multiple Sources[C].Pacific Symposium on Biocomputing 2005,10:531542.

[24]RodriguezGaliano V F,Ghimire B,Rogan J.et al.An assessment of the effectiveness of a random forest classifier for landcover classification[J].Journal of Photogrammetry and Remote Sensing,2012,67:93104.

[25]Cootes T F,Ionita M C,Lindner C,et al.Robust and accurate shape model fitting using random forest regression voting[C].European Conference on Computer Vision.Springer Berlin Heidelberg,2012:278291.