多维复杂关联因素下的大坝变形动态建模与预测分析
|
摘要
大坝变形监控模型是多维复杂关联性以数学形式表达的集成载体,因而合理量化和集成关联性有利于更加全面而准确地构建数学模型。本文着重从维度、关联、检验、措施及模型五个层面依次对因子相关性、动态因果关系和序列相似性三种关联性进行阐述,提出了一种兼顾相关性和相似性的大坝变形动态监控模型。该模型以环境因子与大坝变形间的因果关系为基础,分别采用耦合相关性诊断方法、动态最大信息系数和标准化动态时间规整算法度量因子相关性、动态因果关系和序列相似性,并通过非线性模型建立、动态输入修正和交叉多输出改进将多维复杂关联性统一于大坝变形动态建模框架中。以西南地区某混凝土坝工程多测点变形监测数据为例,通过多模型性能对比仿真实验对所提方法的有效性和准确性进行了验证评估。结果表明,与常规模型相比,集成多维复杂关联性的大坝变形动态监控模型的预测性能尤佳,以期为大坝变形安全预报提供一种新型建模方法。
The dam deformation monitoring model is an integrated carrier of multidimensional complex relevance expressed in mathematical form. The reasonable quantification and integration of relevance are conducive to a more comprehensive and accurate mathematical model. To this end,this paper focuses on the three relevance of factor correlation,dynamic causality and time series similarity from five aspects of dimension,relevance,test,measure and model,and proposes a dynamic monitoring model for dam deformation that takes into account both correlation and similarity. Based on the causal relationship between environmental factors and dam deformation,the model is developed by using the coupling correlation diagnosis method,dynamic maximal information coefficient and normalized dynamic time warping algorithm separately to measure factor correlation,dynamic causality and time series similarity. The multidimensional complex relevance is unified in the dynamic modeling framework of dam deformation through the establishment of nonlinear model,dynamic input correction and cross multiple output improvement. Taking the multi-point deformation monitoring data of a concrete dam in southwest China as an example,a comparison of the multi-model performance was made by simulation experiment to verify and evaluate the effectiveness and accuracy of the proposed method. The results show that,compared with conventional models,the dynamic monitoring model for dam deformation with integrated multidimensional complex relevance has better prediction performance,which provides an alternative modeling method for dam deformation safety prediction.
The dam deformation monitoring model is an integrated carrier of multidimensional complex relevance expressed in mathematical form. The reasonable quantification and integration of relevance are conducive to a more comprehensive and accurate mathematical model. To this end,this paper focuses on the three relevance of factor correlation,dynamic causality and time series similarity from five aspects of dimension,relevance,test,measure and model,and proposes a dynamic monitoring model for dam deformation that takes into account both correlation and similarity. Based on the causal relationship between environmental factors and dam deformation,the model is developed by using the coupling correlation diagnosis method,dynamic maximal information coefficient and normalized dynamic time warping algorithm separately to measure factor correlation,dynamic causality and time series similarity. The multidimensional complex relevance is unified in the dynamic modeling framework of dam deformation through the establishment of nonlinear model,dynamic input correction and cross multiple output improvement. Taking the multi-point deformation monitoring data of a concrete dam in southwest China as an example,a comparison of the multi-model performance was made by simulation experiment to verify and evaluate the effectiveness and accuracy of the proposed method. The results show that,compared with conventional models,the dynamic monitoring model for dam deformation with integrated multidimensional complex relevance has better prediction performance,which provides an alternative modeling method for dam deformation safety prediction.
引文
[1] JIA J. A technical review of hydro-project development in China[J]. Engineering,2016,2(3):302-312.
[2]顾冲时,苏怀智,刘何稚.大坝服役风险分析与管理研究述评[J].水利学报,2018,49(1):26-35.
[3]吴中如.水工建筑物安全监控理论及其应用[M].北京:高等教育出版社,2003.
[4]李雷.大坝风险评价与风险管理[M].北京:中国水利水电出版社,2006.
[5] LI M,SHEN Y,REN Q,et al. A new distributed time series evolution prediction model for dam deformation based on constituent elements[J]. Advanced Engineering Informatics,2019,39:41-52.
[6]李萌,包腾飞,杨建慧,等.灰色模型改进的大坝变形分形几何监控模型[J].水利水运工程学报,2016(4):104-110.
[7]吴斌平,岳攀,鄢玉玲,等.考虑时间影响的神经网络组合模型对心墙堆石坝变形的预测研究[J].水力发电学报,2016,35(9):78-86.
[8]刘晗,李明超,沈扬,等.混凝土坝长期运行安全的理想点法评价模型[J].水利水电科技进展,2019,39(1):70-75,81.
[9]吴中如,陈波.大坝变形监控模型发展回眸[J].现代测绘,2016,39(5):1-3.
[10]李明超,任秋兵,沈扬.贝叶斯框架下的大坝变形交互式时变预测模型及其验证[J].水利学报,2018,49(11):1328-1338.
[11]李富强.大坝安全监测数据分析方法研究[D].杭州:浙江大学,2012.
[12]魏博文,彭圣军,徐镇凯,等.顾及大坝位移残差序列混沌效应的GA-BP预测模型[J].中国科学(技术科学),2015,45(5):541-546.
[13]梅泽宇.基于人工蜂群算法的大坝安全监测[D].大连:大连理工大学,2013.
[14] BUI K T T,BUI D T,ZOU J,et al. A novel hybrid artificial intelligent approach based on neural fuzzy inference model and particle swarm optimization for horizontal displacement modeling of hydropower dam[J]. Neural Computing and Applications,2018,29(12):1495-1506.
[15]张豪,许四法.基于经验模态分解和遗传支持向量机的多尺度大坝变形预测[J].岩石力学与工程学报,2011,30(S2):3681-3688.
[16] KANG F,LIU J,LI J,et al. Concrete dam deformation prediction model for health monitoring based on extreme learning machine[J]. Structural Control and Health Monitoring,2017,24(10):e1997.
[17] SU H,LI X,YANG B,et al. Wavelet support vector machine-based prediction model of dam deformation[J].Mechanical Systems and Signal Processing,2018,110:412-427.
[18]李麟玮,吴益平,苗发盛,等.基于变分模态分解与GWO-MIC-SVR模型的滑坡位移预测研究[J].岩石力学与工程学报,2018,37(6):1395-1406.
[19] DAI B,GU C,ZHAO E,et al. Statistical model optimized random forest regression model for concrete dam deformation monitoring[J]. Structural Control and Health Monitoring,2018,25(6):e2170.
[20]杨杰,胡德秀,吴中如.大坝安全监控模型因子相关性及不确定性研究[J].水利学报,2004,35(12):99-105.
[21] WEI B,YUAN D,XU Z,et al. Modified hybrid forecast model considering chaotic residual errors for dam deformation[J]. Structural Control and Health Monitoring,2018,25(8):e2188.
[22]刘昊元.心墙堆石坝变形的组合预测模型与应用研究[D].天津:天津大学,2014.
[23]尚珂.基于支持向量机回归的草地地上生物量遥感估测研究[D].昆明:西南林业大学,2015.
[24]柯善亮,邱卫宁,花向红.基于监测点相似性的变形监测预测模型与数据修复[J].勘察科学技术,2012(5):9-11.
[25]郑付刚.基于局部异常系数的混凝土坝变形异常分析方法[J].水电能源科学,2016,34(6):103-105,31.
[26]孙小冉,马福恒,张帅,等.基于因子分析法的大坝变形安全监控模型[J].水电能源科学,2012,30(4):34-37.
[27]何晓群,刘文卿.应用回归分析[M].北京:中国人民大学出版社,2015.
[28] CORTES C,VAPNIK V. Support-vector networks[J]. Machine Learning,1995,20(3):273-297.
[29]杜树新,吴铁军.回归型加权支持向量机方法及其应用[J].浙江大学学报(工学版),2004,38(3):302-306.
[30]李泽峰,陈洋波,李雪.柳江流域多时间尺度的降雨径流模拟分析[J].中国农村水利水电,2017(12):64-69.
[31] ZIVOT E,WANG J. Modeling financial time series with S-Plus?[M]. Springer Science&Business Media,2006.
[32] RESHEF D N,RESHEF Y A,FINUCANE H K,et al. Detecting novel associations in large data sets[J]. Science,2011,334(6062):1518-1524.
[33]周志华.机器学习[M].北京:清华大学出版社,2016.
[34] MOREL M,ACHARD C,KULPA R,et al. Time-series averaging using comstrained dynamic time warping with tolerance[J]. Pattem Recognition,2018,74:77-89.
[35] ARORA N,ALLENBY G M,GINTER J L. A hierarchical Bayes model of primary and secondary demand[J].Marketing Science,1998,17(1):29-44.
[36] HESKES T. Empirical Bayes for learning to learn[C]//Proceedings of the Seventeenth International Conference on Machine Learning. San Francisco,CA:Morgan Kaufmann Publishers Inc.,2000.
[37] XU S,AN X,QIAO X,et al. Multi-output least-squares support vector regression machines[J]. Pattern Recognition Letters,2013,34(9):1078-1084.
[38]李双平,张斌.基于小波与谱分析的大坝变形预报模型[J].岩土工程学报,2015,37(2):374-378.
[39] LI M,MIAO L,SHI J. Analyzing heating equipment′s operations based on measured data[J]. Energy and Buildings,2014,82:47-56.
[40] LIANG G,HU Y,LI Q. Safety monitoring of high arch dams in initial operation period using vector error correction model[J]. Rock Mechanics and Rock Engineering,2018,51(8):2469-2481.
[2]顾冲时,苏怀智,刘何稚.大坝服役风险分析与管理研究述评[J].水利学报,2018,49(1):26-35.
[3]吴中如.水工建筑物安全监控理论及其应用[M].北京:高等教育出版社,2003.
[4]李雷.大坝风险评价与风险管理[M].北京:中国水利水电出版社,2006.
[5] LI M,SHEN Y,REN Q,et al. A new distributed time series evolution prediction model for dam deformation based on constituent elements[J]. Advanced Engineering Informatics,2019,39:41-52.
[6]李萌,包腾飞,杨建慧,等.灰色模型改进的大坝变形分形几何监控模型[J].水利水运工程学报,2016(4):104-110.
[7]吴斌平,岳攀,鄢玉玲,等.考虑时间影响的神经网络组合模型对心墙堆石坝变形的预测研究[J].水力发电学报,2016,35(9):78-86.
[8]刘晗,李明超,沈扬,等.混凝土坝长期运行安全的理想点法评价模型[J].水利水电科技进展,2019,39(1):70-75,81.
[9]吴中如,陈波.大坝变形监控模型发展回眸[J].现代测绘,2016,39(5):1-3.
[10]李明超,任秋兵,沈扬.贝叶斯框架下的大坝变形交互式时变预测模型及其验证[J].水利学报,2018,49(11):1328-1338.
[11]李富强.大坝安全监测数据分析方法研究[D].杭州:浙江大学,2012.
[12]魏博文,彭圣军,徐镇凯,等.顾及大坝位移残差序列混沌效应的GA-BP预测模型[J].中国科学(技术科学),2015,45(5):541-546.
[13]梅泽宇.基于人工蜂群算法的大坝安全监测[D].大连:大连理工大学,2013.
[14] BUI K T T,BUI D T,ZOU J,et al. A novel hybrid artificial intelligent approach based on neural fuzzy inference model and particle swarm optimization for horizontal displacement modeling of hydropower dam[J]. Neural Computing and Applications,2018,29(12):1495-1506.
[15]张豪,许四法.基于经验模态分解和遗传支持向量机的多尺度大坝变形预测[J].岩石力学与工程学报,2011,30(S2):3681-3688.
[16] KANG F,LIU J,LI J,et al. Concrete dam deformation prediction model for health monitoring based on extreme learning machine[J]. Structural Control and Health Monitoring,2017,24(10):e1997.
[17] SU H,LI X,YANG B,et al. Wavelet support vector machine-based prediction model of dam deformation[J].Mechanical Systems and Signal Processing,2018,110:412-427.
[18]李麟玮,吴益平,苗发盛,等.基于变分模态分解与GWO-MIC-SVR模型的滑坡位移预测研究[J].岩石力学与工程学报,2018,37(6):1395-1406.
[19] DAI B,GU C,ZHAO E,et al. Statistical model optimized random forest regression model for concrete dam deformation monitoring[J]. Structural Control and Health Monitoring,2018,25(6):e2170.
[20]杨杰,胡德秀,吴中如.大坝安全监控模型因子相关性及不确定性研究[J].水利学报,2004,35(12):99-105.
[21] WEI B,YUAN D,XU Z,et al. Modified hybrid forecast model considering chaotic residual errors for dam deformation[J]. Structural Control and Health Monitoring,2018,25(8):e2188.
[22]刘昊元.心墙堆石坝变形的组合预测模型与应用研究[D].天津:天津大学,2014.
[23]尚珂.基于支持向量机回归的草地地上生物量遥感估测研究[D].昆明:西南林业大学,2015.
[24]柯善亮,邱卫宁,花向红.基于监测点相似性的变形监测预测模型与数据修复[J].勘察科学技术,2012(5):9-11.
[25]郑付刚.基于局部异常系数的混凝土坝变形异常分析方法[J].水电能源科学,2016,34(6):103-105,31.
[26]孙小冉,马福恒,张帅,等.基于因子分析法的大坝变形安全监控模型[J].水电能源科学,2012,30(4):34-37.
[27]何晓群,刘文卿.应用回归分析[M].北京:中国人民大学出版社,2015.
[28] CORTES C,VAPNIK V. Support-vector networks[J]. Machine Learning,1995,20(3):273-297.
[29]杜树新,吴铁军.回归型加权支持向量机方法及其应用[J].浙江大学学报(工学版),2004,38(3):302-306.
[30]李泽峰,陈洋波,李雪.柳江流域多时间尺度的降雨径流模拟分析[J].中国农村水利水电,2017(12):64-69.
[31] ZIVOT E,WANG J. Modeling financial time series with S-Plus?[M]. Springer Science&Business Media,2006.
[32] RESHEF D N,RESHEF Y A,FINUCANE H K,et al. Detecting novel associations in large data sets[J]. Science,2011,334(6062):1518-1524.
[33]周志华.机器学习[M].北京:清华大学出版社,2016.
[34] MOREL M,ACHARD C,KULPA R,et al. Time-series averaging using comstrained dynamic time warping with tolerance[J]. Pattem Recognition,2018,74:77-89.
[35] ARORA N,ALLENBY G M,GINTER J L. A hierarchical Bayes model of primary and secondary demand[J].Marketing Science,1998,17(1):29-44.
[36] HESKES T. Empirical Bayes for learning to learn[C]//Proceedings of the Seventeenth International Conference on Machine Learning. San Francisco,CA:Morgan Kaufmann Publishers Inc.,2000.
[37] XU S,AN X,QIAO X,et al. Multi-output least-squares support vector regression machines[J]. Pattern Recognition Letters,2013,34(9):1078-1084.
[38]李双平,张斌.基于小波与谱分析的大坝变形预报模型[J].岩土工程学报,2015,37(2):374-378.
[39] LI M,MIAO L,SHI J. Analyzing heating equipment′s operations based on measured data[J]. Energy and Buildings,2014,82:47-56.
[40] LIANG G,HU Y,LI Q. Safety monitoring of high arch dams in initial operation period using vector error correction model[J]. Rock Mechanics and Rock Engineering,2018,51(8):2469-2481.