基于Adaboost的改进BP神经网络地表沉陷预测
|
摘要
BP神经网络可以解决地表沉陷等非线性关系问题,为了更精确地进行地表沉陷变形预测,引入Adaboost算法对BP神经网络进行改进,并运用Matlab R2014a建立基于Adaboost的BP神经网络地表沉陷预测模型。首先通过BP神经网络进行训练、测试,经过多次迭代,将每个BP神经网络作为一个弱预测器加权组合,形成强预测器,并首次对青岛地铁3号线保河区间隧道进行地表下沉值预测。预测结果表明:Adaboost的BP神经网络预测下沉值的平均绝对误差为0.585 3 mm,平均相对误差为5.82%,与BP神经网络预测相比,绝对误差降低了2.594 7 mm,相对误差降低了27.46%,由此表明Adaboost的BP神经网络适用于地表沉陷预测,且预测精度更高。
A BP neural network could solve a non-linear relationship problem of the surface subsidence. In order to more accurately predict the surface subsidence deformation,an Adaboost algorithm was introduced to improve the BP neural network. A Matlab R2014 a was applied to establish a surface subsidence prediction model with BP neural network based on the Adaboost. Firstly,a training and test were conducted with the BP neural network. After several iterations,each BP neural network as a weak predictor would be weighted and combined and then would form a strong predictor. A surface subsidence prediction was conducted in Baohe section tunnel of Qingdao Metro No.3 Line. The prediction results showed that the predicted average absolute error predicted with the BP neural network of the Adaboost was 0.585 3 mm,the average relative error was 5.82 %. In a comparison with the prediction of BP neural network,the absolute error was reduced by 2.594 7 mm and the relative error was reduced by 27.46 %. Therefore the BP neural network with the Adaboost could be suitably applied to the prediction of the surface subsidence prediction and the prediction accuracy would be higher.
A BP neural network could solve a non-linear relationship problem of the surface subsidence. In order to more accurately predict the surface subsidence deformation,an Adaboost algorithm was introduced to improve the BP neural network. A Matlab R2014 a was applied to establish a surface subsidence prediction model with BP neural network based on the Adaboost. Firstly,a training and test were conducted with the BP neural network. After several iterations,each BP neural network as a weak predictor would be weighted and combined and then would form a strong predictor. A surface subsidence prediction was conducted in Baohe section tunnel of Qingdao Metro No.3 Line. The prediction results showed that the predicted average absolute error predicted with the BP neural network of the Adaboost was 0.585 3 mm,the average relative error was 5.82 %. In a comparison with the prediction of BP neural network,the absolute error was reduced by 2.594 7 mm and the relative error was reduced by 27.46 %. Therefore the BP neural network with the Adaboost could be suitably applied to the prediction of the surface subsidence prediction and the prediction accuracy would be higher.
引文
[1]刘彤.王家塔矿区采空区地表沉陷机理及模型预测研究[D].长春:吉林大学,2017.
[2]蒋博林.浅埋暗挖地铁隧道施工安全性和风险分析研究[D].重庆:重庆交通大学,2011.
[3] HOSSEINI S,MOHAMMADNEJAD M,HOSEINI S,et al.Numerical and analytical investigation of ground surface settlement due to subway excavation[J].Geosciences,2012,23(6):470-475.
[4]余学义,张恩强.开采损害学[M].北京:煤炭工业出版社,2010.
[5]王军保,刘新荣,李鹏,等.MMF模型在采空区地表沉降预测中的应用[J].煤炭学报,2012,37(3):411-415.WANG Junbao,LIU Xinrong,LI Peng,et al. Application of MMF model to prediction of surface subsidence in goaf area[J]. Journal of China Coal Society,2012,37(3):411-415.
[6]王宁,吴侃,刘锦,等.基于Boltzmann函数的开采沉陷预测模型[J].煤炭学报,2013,38(8):1352-1356.WANG Ning,WU Kan,LIU Jin,et al. Prediction model of mining subsidence based on Boltzmann function[J].Journal of China Coal Society,2013,38(8):1352-1356.
[7]张兵,崔希民,胡青峰.开采沉陷动态预计的正态分布时间函数模型研究[J].煤炭科学技术,2016,44(4):140-145.ZHANG Bing,CUI Ximin,HU Qingfeng. Research on the normal distribution time function model for mining subsidence dynamic prediction[J]. Coal Science and Technology,2016,44(4):140-145.
[8]高彩云,崔希民,高宁.熵权遗传算法及极限学习机地铁隧道沉降预测[J].测绘科学,2016,41(2):71-75.GAO Caiyun,CUI Ximin,GAO Ning. Entropy weight genetic algorithm and extreme learning machine subwaytunnel settlement prediction[J].Mapping Science,2016,41(2):71-75.
[9]周纯择,阳军生,牟友滔,等.南昌上软下硬地层中盾构施工地表沉降的BP神经网络预测方法[J].防灾减灾工程学报,2015,35(4):556-562.ZHOU Chunze,YANG Junsheng,MU Youtao,et al. A BP neural network prediction method for surface settlement of shield tunneling in the upper soft lower hard rock formation in Nanchang[J].Journal of Disaster Prevention and Mitigation Engineering,2015,35(4):556-562.
[10]乔金丽,范永利.基于改进BP网络的盾构隧道开挖地表沉降预测[J].地下空间与工程学报,2012,8(2):134-139,156.QIAO JinLi,FAN YongLi. Prediction of ground settlement of shield tunnel excavation based on improved BP neural network[J]. Chinese Journal of Underground Space and Engineering,2012,8(2):134-139,156.
[11]高彩云,崔希民,高宁,等.GRNN神经网络在矿区地表变形预测中的应用[J].金属矿山,2014(3):97-100.GAO Caiyun,CUI Ximin,GAO Ning,etal. Application of GRNN neural network to predict ground surface deformation in mining area[J].Metal Mine,2014(3):97-100.
[12]黄丽.BP神经网络算法改进及应用研究[D].重庆:重庆师范大学,2008:8-13.
[13] LI B,ZHANG L,ZHANG W. Surface Subsidence Prediction Model based on BP Neural Network[J]. Journal of Convergence Information Technology,2013,8(6):229-235.
[14] Wenyu L,Zhihui Z.Surface subsidence prediction model of coal mining based on chaos optimization and BPneural network[J].Electronic Journal of Geotechnical Engineering,2015,20(12):4401-4411.
[15]吴俊利,张步涵,王魁.基于Adaboost的BP神经网络改进算法在短期风速预测中的应用[J].电网技术,2012,36(9):221-225.WU Junli,ZHANG Buhan,WANG Kui. Application of BP neural network improved algorithm based on Adaboost in short term wind speed prediction[J]. Power System Technology,2012,36(9):221-225.
[16] WU H,ZOU B J,ZHAO Y Q,et al.An automatic video text detection method based on BP-Adaboost[J].Multimedia Tools&Applications,2016,75(13):1-24.
[17]王改革,郭立红,段红,等.基于Elman-Adaboost强预测器的目标威胁评估模型及算法[J].电子学报,2012,40(5):901-906.WANG Gaige,GUO Lihong,DUAN Hong,et al.A model and algorithm for target threat assessment based on Elman-Adaboost strong predictor[J]. Journal of Electronics,2012,40(5):901-906.
[18]王小川.MATLAB神经网络43个案例分析[M].北京:北京航空航天大学出版社,2013.
[19]李翔,朱全银.基于Adaboost算法和BP神经网络的税收预测[J].计算机应用,2012,32(12):3558-3560.LI Xiang,ZHU Quanyin. Tax revenue forecasting based on Adaboost algorithm and BP neural network[J].Computer Applications,2012,32(12):3558-3560.
[20]袁杰.地铁隧道开挖诱发的地表沉降与变形预测分析[D].北京:北京交通大学,2010:17-19.
[21]段宝福,宋立坤,周鑫明,等.浅埋暗挖地铁隧道地表沉降研究现状[J].现代隧道技术,2017,54(4):25-32.DUAN Baofu,SONG Likun,ZHOU Xinming,et al.Research status of surface subsidence in shallow-buried subway tunnel[J].Modern Tunneling Technology,2017,54(4):25-32.
[22]徐星星.青岛地铁3号线暗挖段地表变形BP神经网络预测研究[D].青岛:中国海洋大学,2012:48-53.
[2]蒋博林.浅埋暗挖地铁隧道施工安全性和风险分析研究[D].重庆:重庆交通大学,2011.
[3] HOSSEINI S,MOHAMMADNEJAD M,HOSEINI S,et al.Numerical and analytical investigation of ground surface settlement due to subway excavation[J].Geosciences,2012,23(6):470-475.
[4]余学义,张恩强.开采损害学[M].北京:煤炭工业出版社,2010.
[5]王军保,刘新荣,李鹏,等.MMF模型在采空区地表沉降预测中的应用[J].煤炭学报,2012,37(3):411-415.WANG Junbao,LIU Xinrong,LI Peng,et al. Application of MMF model to prediction of surface subsidence in goaf area[J]. Journal of China Coal Society,2012,37(3):411-415.
[6]王宁,吴侃,刘锦,等.基于Boltzmann函数的开采沉陷预测模型[J].煤炭学报,2013,38(8):1352-1356.WANG Ning,WU Kan,LIU Jin,et al. Prediction model of mining subsidence based on Boltzmann function[J].Journal of China Coal Society,2013,38(8):1352-1356.
[7]张兵,崔希民,胡青峰.开采沉陷动态预计的正态分布时间函数模型研究[J].煤炭科学技术,2016,44(4):140-145.ZHANG Bing,CUI Ximin,HU Qingfeng. Research on the normal distribution time function model for mining subsidence dynamic prediction[J]. Coal Science and Technology,2016,44(4):140-145.
[8]高彩云,崔希民,高宁.熵权遗传算法及极限学习机地铁隧道沉降预测[J].测绘科学,2016,41(2):71-75.GAO Caiyun,CUI Ximin,GAO Ning. Entropy weight genetic algorithm and extreme learning machine subwaytunnel settlement prediction[J].Mapping Science,2016,41(2):71-75.
[9]周纯择,阳军生,牟友滔,等.南昌上软下硬地层中盾构施工地表沉降的BP神经网络预测方法[J].防灾减灾工程学报,2015,35(4):556-562.ZHOU Chunze,YANG Junsheng,MU Youtao,et al. A BP neural network prediction method for surface settlement of shield tunneling in the upper soft lower hard rock formation in Nanchang[J].Journal of Disaster Prevention and Mitigation Engineering,2015,35(4):556-562.
[10]乔金丽,范永利.基于改进BP网络的盾构隧道开挖地表沉降预测[J].地下空间与工程学报,2012,8(2):134-139,156.QIAO JinLi,FAN YongLi. Prediction of ground settlement of shield tunnel excavation based on improved BP neural network[J]. Chinese Journal of Underground Space and Engineering,2012,8(2):134-139,156.
[11]高彩云,崔希民,高宁,等.GRNN神经网络在矿区地表变形预测中的应用[J].金属矿山,2014(3):97-100.GAO Caiyun,CUI Ximin,GAO Ning,etal. Application of GRNN neural network to predict ground surface deformation in mining area[J].Metal Mine,2014(3):97-100.
[12]黄丽.BP神经网络算法改进及应用研究[D].重庆:重庆师范大学,2008:8-13.
[13] LI B,ZHANG L,ZHANG W. Surface Subsidence Prediction Model based on BP Neural Network[J]. Journal of Convergence Information Technology,2013,8(6):229-235.
[14] Wenyu L,Zhihui Z.Surface subsidence prediction model of coal mining based on chaos optimization and BPneural network[J].Electronic Journal of Geotechnical Engineering,2015,20(12):4401-4411.
[15]吴俊利,张步涵,王魁.基于Adaboost的BP神经网络改进算法在短期风速预测中的应用[J].电网技术,2012,36(9):221-225.WU Junli,ZHANG Buhan,WANG Kui. Application of BP neural network improved algorithm based on Adaboost in short term wind speed prediction[J]. Power System Technology,2012,36(9):221-225.
[16] WU H,ZOU B J,ZHAO Y Q,et al.An automatic video text detection method based on BP-Adaboost[J].Multimedia Tools&Applications,2016,75(13):1-24.
[17]王改革,郭立红,段红,等.基于Elman-Adaboost强预测器的目标威胁评估模型及算法[J].电子学报,2012,40(5):901-906.WANG Gaige,GUO Lihong,DUAN Hong,et al.A model and algorithm for target threat assessment based on Elman-Adaboost strong predictor[J]. Journal of Electronics,2012,40(5):901-906.
[18]王小川.MATLAB神经网络43个案例分析[M].北京:北京航空航天大学出版社,2013.
[19]李翔,朱全银.基于Adaboost算法和BP神经网络的税收预测[J].计算机应用,2012,32(12):3558-3560.LI Xiang,ZHU Quanyin. Tax revenue forecasting based on Adaboost algorithm and BP neural network[J].Computer Applications,2012,32(12):3558-3560.
[20]袁杰.地铁隧道开挖诱发的地表沉降与变形预测分析[D].北京:北京交通大学,2010:17-19.
[21]段宝福,宋立坤,周鑫明,等.浅埋暗挖地铁隧道地表沉降研究现状[J].现代隧道技术,2017,54(4):25-32.DUAN Baofu,SONG Likun,ZHOU Xinming,et al.Research status of surface subsidence in shallow-buried subway tunnel[J].Modern Tunneling Technology,2017,54(4):25-32.
[22]徐星星.青岛地铁3号线暗挖段地表变形BP神经网络预测研究[D].青岛:中国海洋大学,2012:48-53.