基于ε-SVR与PSO-DE的岩层弹塑性参数反演及应用
|
摘要
针对岩土力学参数正算反演方法存在的瓶颈问题,研究了将支持向量回归机(ε-SVR)与粒子群-差分演化杂交智能优化算法(BBDE)相结合的参数反演方法,采用ε-SVR的预测功能代替耗时的数值仿真计算,将其嵌入到改进了控制参数取值方法的BBDE算法中建立了相应的反演程序,应用该程序对一个实际工程进行了弹塑性力学参数反演,并对反演程序的可行性和有效性进行了研究.结果表明:利用ε-SVR具有的BP神经网络所不可比拟的泛化推广能力,在保证反演精度的同时提高了反演效率;引入BBDE算法使得在减少算法控制参数的同时提高了解的全局收敛性和收敛速度;将反演所得参数输入数值仿真模型获得的测点计算增量位移与实测增量位移比较吻合,相对误差不超过10%.
To solve the bottleneck problems in forward calculation-inverse analysis method for geotechnical mechanic parameter identification,an approach of combining support vector regression machine(hereinafter refer to ε-SVR) with particles swarm-differential evolution hybrid optimization algorithm(hereinafter refer to BBDE) was investigated,and a program was developed in which ε-SVR,functioning as predicting incremental displacement in measuring point(in such way,time-consuming finite difference numerical calculations could be avoided),was imbedded into BBDE main routine as a subroutine.Meanwhile,the method of determining reasonable value of control parameter in BBDE was improved in order to promoting the performance of algorithm.The feasibility and effectiveness of inverse analysis program were testified by applying it to identify elasto-plastic mechanic parameters in an engineering site.The results show that generalization ability of ε-SVR is far beyond that of BP neural network,using ε-SVR as a substitute for numerical calculations can improve the efficiency of inverse analysis significantly and ensure solving precision.BBDE performs very well in finding global best solution and speeding up convergence but has less control parameters than PSO and DE algorithm.Finite difference numerical calculation was executed again after inverse analysis,it shows that the calculated incremental displacements approximately agree with those of field measurement in all measuring points,and the fractional errors are less than 10%.
To solve the bottleneck problems in forward calculation-inverse analysis method for geotechnical mechanic parameter identification,an approach of combining support vector regression machine(hereinafter refer to ε-SVR) with particles swarm-differential evolution hybrid optimization algorithm(hereinafter refer to BBDE) was investigated,and a program was developed in which ε-SVR,functioning as predicting incremental displacement in measuring point(in such way,time-consuming finite difference numerical calculations could be avoided),was imbedded into BBDE main routine as a subroutine.Meanwhile,the method of determining reasonable value of control parameter in BBDE was improved in order to promoting the performance of algorithm.The feasibility and effectiveness of inverse analysis program were testified by applying it to identify elasto-plastic mechanic parameters in an engineering site.The results show that generalization ability of ε-SVR is far beyond that of BP neural network,using ε-SVR as a substitute for numerical calculations can improve the efficiency of inverse analysis significantly and ensure solving precision.BBDE performs very well in finding global best solution and speeding up convergence but has less control parameters than PSO and DE algorithm.Finite difference numerical calculation was executed again after inverse analysis,it shows that the calculated incremental displacements approximately agree with those of field measurement in all measuring points,and the fractional errors are less than 10%.
引文
[1]邓建辉,李焯芬,葛修润.BP网络和遗传算法在岩石边坡位移反分析中的应用[J].岩石力学与工程学报,2001,20(1):1-5.DENG Jian-hui,LI Chao-fen,GE Xiu-run.Applica-tion of BP network and genetic algorithm to displace-ment back analysis of rock slopes[J].Chinese Journalof Rock Mechanics and Engineering,2001,20(1):1-5.
[2]张社荣,何辉.改进的遗传算法在堆石体参数反演中的应用[J].岩土力学,2005,26(2):182-186.ZHANG She-rong,HE Hui.Application of im-proved genetic algorithm to back analyzing parame-ters of rockfill[J].Rock and Soil Mechanics,2005,26(2):182-186.
[3]徐宝松,苗艳东,李波,等.基于粒子群算法的大坝力学参数反演[J].水力发电,2009,35(3):102-104.XU Bao-song,MIAO Yan-dong,LI Bo,et al.In-verse analysis of mechanical parameters of dam basedon PSO[J].Water Power,2009,35(3):102-104.
[4]田明俊,周晶.基于蚁群算法的土石坝土体参数反演[J].岩石力学与工程学报,2005,24(8):1411-1416.TIAN Ming-jun,ZHOU Jing.Inversing soil mechan-ical parameters of embankment dam using ant colonyalgorithm[J].Chinese Journal of Rock Mechanicsand Engineering,2005,24(8):1411-1416.
[5]燕柳斌,陈艳香,苏国韶.岩体弹性模量反分析的进化差分方法[J].广西大学学报:自然科学版,2008,33(4):346-348.YAN Liu-bin,CHEN Yan-xiang,SU Guo-shao.Arock mass elastic modulus back analysis method usingdifferential evolution algorithm[J].Journal of Guan-gxi University:Natural Science Edition,2008,33(4):346-348.
[6]杜小凯,任青文,郑治,等.基于APSO-BP耦合算法的岩体力学参数反馈研究[J].中国矿业大学学报,2008,37(6):756-762.DU Xiao-kai,REN Qing-wen,ZHENG Zhi,et al.Back-analysis of rock mechanic parameters based oncoupling algorithm of adaptive particle-swarm-optimi-zation and back-propagation neural network[J].Jour-nal of China University of Mining&Technology,2008,37(6):756-762.
[7]王炜,刘悦,李国正,等.我国大陆强震预测的支持向量机方法[J].地震学报,2006,28(1):29-36.WANG Wei,LIU Yue,LI Guo-zheng,et al.Sup-port vector machine method for forecasting futurestrong eearth-quakes in chinese mainland[J].ActaSeismologica Sinica,2006,28(1):29-36.
[8]刘俊萍,畅明琦.基于支持向量机的需水预测研究[J].太原理工大学学报,2008,39(3):299-302.LIU Jun-ping,CHANG Ming-qi.Water demand pre-diction model based on support vector machine[J].Journal of Taiyuan University of Technology,2008,39(3):299-302.
[9]邓乃扬,田英杰.支持向量机:理论、算法与拓展[M].北京:科学出版社,2009:43-111.
[10]MAHAMED G H,OMRAN,ANDRIES P.Engel-brecht,ayed salman.bare bones differential evolu-tion[J].European Journal of Operational Research,2009(196):128-139.
[11]STORN R K,PRICE V.Differential evolution-asimple and efficient heuristic for global optimizationover continuous spaces[J].Journal of Global Opti-mization,1997(11):341-359.
[12]肯尼迪,埃伯哈特,史玉回.群体智能:英文版[M].北京:人民邮电出版社,2009:287-368.
[13]郭一楠,王辉,程健.一类自适应混沌文化算法[J].控制与决策,2009,24(4):514-519.GUO Yi-nan,WANG Hui,CHEN Jian.A novel a-daptive chaotic cultural algorithm[J].Control andDecision,2009,24(4):514-519.
[14]王媛,刘杰.基于敏感性分析的裂隙岩体渗流与应力静态全耦合参数反演[J].岩土力学,2009,30(2):311-317.WANG Yuan,LIU Jie.Paremeter inversion forfully coupled problem of steady fluid flow and stressin fractured rock masses based on sensitivity analy-sis[J].Rock and Soil Mechanics,2009,30(2):311-317.
[15]丁德馨,杨仕教,孙钧.岩体弹塑性模型力学参数对位移的影响度研究[J].岩石力学与工程学报,2003,22(5):697-701.DING De-xin,YANG Shi-jiao,SUN Jun.Studiesof influencen degree of mechnical paremeters on dis-placements in elsto-plastic model for rock mass[J].Chinese Journal of Rock Mechanics and Engineer-ing,2003,22(5):697-701.
[2]张社荣,何辉.改进的遗传算法在堆石体参数反演中的应用[J].岩土力学,2005,26(2):182-186.ZHANG She-rong,HE Hui.Application of im-proved genetic algorithm to back analyzing parame-ters of rockfill[J].Rock and Soil Mechanics,2005,26(2):182-186.
[3]徐宝松,苗艳东,李波,等.基于粒子群算法的大坝力学参数反演[J].水力发电,2009,35(3):102-104.XU Bao-song,MIAO Yan-dong,LI Bo,et al.In-verse analysis of mechanical parameters of dam basedon PSO[J].Water Power,2009,35(3):102-104.
[4]田明俊,周晶.基于蚁群算法的土石坝土体参数反演[J].岩石力学与工程学报,2005,24(8):1411-1416.TIAN Ming-jun,ZHOU Jing.Inversing soil mechan-ical parameters of embankment dam using ant colonyalgorithm[J].Chinese Journal of Rock Mechanicsand Engineering,2005,24(8):1411-1416.
[5]燕柳斌,陈艳香,苏国韶.岩体弹性模量反分析的进化差分方法[J].广西大学学报:自然科学版,2008,33(4):346-348.YAN Liu-bin,CHEN Yan-xiang,SU Guo-shao.Arock mass elastic modulus back analysis method usingdifferential evolution algorithm[J].Journal of Guan-gxi University:Natural Science Edition,2008,33(4):346-348.
[6]杜小凯,任青文,郑治,等.基于APSO-BP耦合算法的岩体力学参数反馈研究[J].中国矿业大学学报,2008,37(6):756-762.DU Xiao-kai,REN Qing-wen,ZHENG Zhi,et al.Back-analysis of rock mechanic parameters based oncoupling algorithm of adaptive particle-swarm-optimi-zation and back-propagation neural network[J].Jour-nal of China University of Mining&Technology,2008,37(6):756-762.
[7]王炜,刘悦,李国正,等.我国大陆强震预测的支持向量机方法[J].地震学报,2006,28(1):29-36.WANG Wei,LIU Yue,LI Guo-zheng,et al.Sup-port vector machine method for forecasting futurestrong eearth-quakes in chinese mainland[J].ActaSeismologica Sinica,2006,28(1):29-36.
[8]刘俊萍,畅明琦.基于支持向量机的需水预测研究[J].太原理工大学学报,2008,39(3):299-302.LIU Jun-ping,CHANG Ming-qi.Water demand pre-diction model based on support vector machine[J].Journal of Taiyuan University of Technology,2008,39(3):299-302.
[9]邓乃扬,田英杰.支持向量机:理论、算法与拓展[M].北京:科学出版社,2009:43-111.
[10]MAHAMED G H,OMRAN,ANDRIES P.Engel-brecht,ayed salman.bare bones differential evolu-tion[J].European Journal of Operational Research,2009(196):128-139.
[11]STORN R K,PRICE V.Differential evolution-asimple and efficient heuristic for global optimizationover continuous spaces[J].Journal of Global Opti-mization,1997(11):341-359.
[12]肯尼迪,埃伯哈特,史玉回.群体智能:英文版[M].北京:人民邮电出版社,2009:287-368.
[13]郭一楠,王辉,程健.一类自适应混沌文化算法[J].控制与决策,2009,24(4):514-519.GUO Yi-nan,WANG Hui,CHEN Jian.A novel a-daptive chaotic cultural algorithm[J].Control andDecision,2009,24(4):514-519.
[14]王媛,刘杰.基于敏感性分析的裂隙岩体渗流与应力静态全耦合参数反演[J].岩土力学,2009,30(2):311-317.WANG Yuan,LIU Jie.Paremeter inversion forfully coupled problem of steady fluid flow and stressin fractured rock masses based on sensitivity analy-sis[J].Rock and Soil Mechanics,2009,30(2):311-317.
[15]丁德馨,杨仕教,孙钧.岩体弹塑性模型力学参数对位移的影响度研究[J].岩石力学与工程学报,2003,22(5):697-701.DING De-xin,YANG Shi-jiao,SUN Jun.Studiesof influencen degree of mechnical paremeters on dis-placements in elsto-plastic model for rock mass[J].Chinese Journal of Rock Mechanics and Engineer-ing,2003,22(5):697-701.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心