Fuzzy-SVM prediction model of subgrade seismic damage based on Bayesian network

doi:10.16265/j.cnki.issn 1003-3033.2021.11.024

Abstract

Abstract: In order to address the problem of strong subjectivity and lack of consideration of nonlinear characteristics for existing subgrade earthquake damage prediction methods, Fuzzy-SVM prediction model based on BN was established by integrating engineering experience with historical seismic damage samples and by improving network parameter solving method. Fuzzy theory was used to solve prior probability of BN parameters, and SVM was used to solve their actual sample potential probability. Then, based on Bayesian principle, prior probability and actual sample potential probability were fused to obtain a prediction model which not only were consistent with engineering experience of earthquake damage, but also reflected nonlinear characteristics of historical samples. The results show that the prediction model features a accuracy rate of 80.95% when being applied to 42 subgrade hidden danger points in Wenchuan earthquake affected area. It has a higher accuracy than traditional machine learning method (represented by SVM) in the case of small samples. Moreover, it can effectively predict earthquake damage level when subgrade attributes are incomplete.

Key words: Bayesian network (BN), subgrade seismic damage, prediction model, fuzzy-support vector machine (SVM), prior knowledge

CLC Number:

X915.5

LIU Yang, ZHANG Jianjing, LUO Hongsen, YU Haiying, XIANG Bo. Fuzzy-SVM prediction model of subgrade seismic damage based on Bayesian network[J]. China Safety Science Journal, 2021, 31(11): 171-178.

References

[1] ADAFER S, BENSAIBI M. Seismic vulnerability classification of roads[J]. Energy Procedia, 2017,139:624-630.
[2] 苏晓成, 周天来, 刘志峰. 拟静力法公路路基结构抗震稳定性研究[J]. 地震工程学报, 2014,36(3):482-488.
SU Xiaocheng, ZHOU Tianlai, LIU Zhifeng. Seismic stability analysis of highway subgrade based on the pseudo-static method[J]. China Earthquake Engineering Journal, 2014,36(3):482-488.
[3] 陈小云, 吴红刚, 杨涛. 高原地区路基地震液化数值模拟研究[J]. 地震工程学报, 2017,39(4):706-712.
CHEN Xiaoyun, WU Honggang, YANG Tao. Numerical simulation of the earthquake liquefaction of a roadbed in a plateau[J]. China Earthquake Engineering Journal, 2017,39(4):706-712.
[4] 苏晓成. 路基结构的动力稳定性研究[D]. 天津:天津大学, 2010.
SU Xiaocheng. Researchs on the roadbed structure dynamic stability[D]. Tianjin:Tianjin University, 2010.
[5] 邱长林, 张庆建. 路基地震峰值加速度响应特性振动台试验研究[J]. 地震工程学报, 2014,36(4):778-783.
QIU Changlin, ZHANG Qingjian. Shaking-table tests on response characteristics of seismic peak acceleration of subgrade[J]. China Earthquake Engineering Journal, 2014,36(4):778-783.
[6] 陈伟志, 蒋关鲁, 刘勇,等. 川藏铁路钻孔桩加固斜坡路基的振动台试验研究[J]. 岩石力学与工程学报, 2020,39(12):2 540-2 556.
CHEN Weizhi, JIANG Guanlu, LIU Yong, et al. Shaking table test on slope subgrades reinforced by bored piles of Sichuan-Tibet railway[J]. Chinese Journal of Rock Mechanics and Engineering, 2020,39(12):2 540-2 556.
[7] HAZUS. Multi-hazard loss estimation methodology, earthquake model, HAZUS MH-MR4, technical manual[Z]. FEMA and NIBS, 2003.
[8] 廖燚. 汶川地震公路路基震害调查分析及易损性研究[D]. 成都:西南交通大学, 2012.
LIAO Yi. Analysis of seismic hazard survey and study on vulnerability of highway subgrade in Wenchuan earthquake[D]. Chengdu: Southwest Jiaotong University, 2012.
[9] 陈一平, 陈欣. 公路路基路面的震害预测[J]. 工程抗震, 1993(3):40-45.
CHEN Yiping, CHEN Xin. Earthquake damage prediction of highway subgrade pavement[J]. Engineering Earthquake Resistance, 1993(3):40-45.
[10] 曾勇军. 铁路公路震害损失预测模型研究[D]. 成都:西南交通大学, 2015.
ZENG Yongjun. Study on prediction model of railway and highway earthquake disaster losses[D]. Chengdu: Southwest Jiaotong University, 2015.
[11] ADAFER S, BENSAIB M. Seismic vulnerability assessment of road networks[C]. Second European Conference on Earthquake Engineering and Seismology, 2014:1-9.
[12] ANDREA A, CAFISO S, CONDORELLI A. Methodological considerations for the evaluation of seismic risk on road network[J]. Pure and Applied Geophysics, 2005, 162(4):767-782.
[13] 刘阳, 张建经, 李孟芳,等. 基于模糊理论与SVM的边坡地震失稳规模贝叶斯网络评估方法[J]. 岩石力学与工程学报, 2019,38(增1):2 807-2 815.
LIU Yang,ZHANG Jianjing,LI Mengfang, et al. Fuzzy theory-and SVM-based Bayesian network assessment method for slope seismic instability scale[J]. Chinese Journal of Rock Mechanics and Engineering, 2019,38(S1):2 807-2 815.
[14] 宋英华, 马连隆, 杨丽娇,等. 基于模糊BN的企业业务中断风险分析:以洪涝灾害情景为例[J]. 中国安全科学学报, 2019,29(6):1-6.
SONG Yinghua, MA Lianlong, YANG Lijiao, et al. Business interruption risk analysis based on fuzzy BN: a case study of flood disaster[J]. China Safety Science Journal, 2019,29(6):1-6.
[15] 王文和, 董传富, 刘林精,等. 基于FTA-BN模型的页岩气井口装置失效概率分析[J]. 中国安全科学学报, 2019,29(4):89-95.
WANG Wenhe, DONG Chuanfu, LIU Linjing, et al. Failure probability analysis of shale gas wellhead devices based on FTA-BN model[J]. China Safety Science Journal, 2019,29(4):89-95.
[16] ZADEH L A. Fuzzy sets[J]. Information and Control, 1965, 8(3):338-353.
[17] CORTES C, VAPNIK V. Support-vector networks[J]. Machine Learning, 1995, 20(3):273-297.
[18] 曾超,王可君,李曙平,等.“5.12”汶川大地震公路路基震害浅析[J]. 世界地震工程, 2009,25(3):162-166.
ZENG Chao, WANG Kejun, LI Shuping, et al. Primary analysis of the seismic damage to roadbeds during Wenchuan earthquake[J]. World Earthquake Engineering, 2009,25(3):162-166.
[19] ANBAZHAGAN P, RINIVAS S, CHANDRAN D. Classification of road damage due to earthquakes[J]. Natural Hazards, 2012, 60(2):425-460.
[20] 陈乐生. 汶川地震公路震害调查. 路基[M]. 北京:人民交通出版社, 2012:63-372.
[21] 张连文, 郭海鹏. 贝叶斯网引论[M]. 北京:科学出版社, 2006:31-74.
[22] 苏永华, 何满潮, 孙晓明. 岩体模糊分类中隶属函数的等效性[J]. 北京科技大学学报, 2007,29(7):670-675.
SU Yonghua, HE Manchao, SUN Xiaoming. Equivalent characteristic of membership function type in rock mass fuzzy classification[J]. Journal of University of Science and Technology Beijing, 2007,29(7):670-675.
[23] PLATT J C. Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods[J]. Advances in Large Margin Classifiers, 1999, 10(4):61-74.
[24] PRICE D, KNERR S, PERSONNAZ L, et al. Pairwise neural network classifiers with probabilistic outputs[C]. International Conference on Neural Information Processing Systems. MIT Press, 1994:1 109-1 116.
[25] BENOV D M. The manhattan project, the first electronic computer and the Monte Carlo method[J]. Monte Carlo Methods and Applications, 2016, 22(1):73-79.
[26] LAURITZEN S L, SPIEGEHALTER D J. Local computations with probabilities on graphical structures and their application to expert systems[J]. Journal of the Royal Statistical Society, 1988, 50(2):157-224.