|Table of Contents|

[1] Wang Jixin, Wang Yan, Zhai Xinting, Huang Yajun, et al. Automatic determination method of optimal thresholdbased on the bootstrapping technology [J]. Journal of Southeast University (English Edition), 2018, 34 (2): 208-212. [doi:10.3969/j.issn.1003-7985.2018.02.010]
Copy

Automatic determination method of optimal thresholdbased on the bootstrapping technology()
基于自助采样技术的最优阈值选取方法
Share：

Journal of Southeast University (English Edition)[ISSN:1003-7985/CN:32-1325/N]

Volumn:
34
Issue:
2018 2
Page:
208-212
Research Field:
Mechanical Engineering
Publishing date:
2018-06-20

Info

Title:
Automatic determination method of optimal thresholdbased on the bootstrapping technology
基于自助采样技术的最优阈值选取方法
Author(s):
Wang Jixin1;  Wang Yan1;  Zhai Xinting1;  Huang Yajun2;  Wang Zhenyu2
1 School of Mechanical Science and Engineering, Jilin University, Changchun 130025, China
2 Shantui Construction Machinery Co., Ltd., Jining 272073, China
王继新1;  王岩1;  翟新婷1;  黄亚军2;  王振雨2
1吉林大学机械科学与工程学院, 长春 130025; 2山推工程机械股份有限公司, 济宁 272073
Keywords:
load spectrum;  peak over threshold;  threshold selection;  bootstrapping technology;  mean squared error
载荷谱;  超阈值;  阈值选取;  自助采样技术;  均方误差
PACS:
TH243
DOI:
10.3969/j.issn.1003-7985.2018.02.010
Abstract:
In order to predict the extreme load of the mechanical components during the entire life, an automatic method based on the bootstrapping technology(BT)is proposed to determine the most suitable threshold. Based on all the turning points of the load history and a series of thresholds estimated in advance, the generalized Pareto distribution is established to fit the exceedances. The corresponding distribution parameters are estimated with the maximum likelihood method. Then, BT is employed to calculate the mean squared error(MSE)of each estimated threshold based on the exceedances and the specific distribution parameters. Finally, the threshold with the smallest MSE will be the optimal one. Compared to the kurtosis method and the mean excess function method, the average deviation of the probability density function of exceedances determined by BT reduces by 38.52% and 29.25%, respectively. Moreover, the quantile-quantile plot of the exceedances determined by BT is closer to a straight line. The results suggest the improvement of the modeling flexibility and the determined threshold precision. If the exceedances are insufficient, BT will enlarge their amount by resampling to solve the instability problem of the original distribution parameters.
为了预测机械部件整个生命周期内的极限载荷, 提出了一种基于自助采样技术选取最优阈值的自动方法.该方法首先提取载荷历程的所有极值点, 估计出一系列阈值, 用广义帕累托分布函数拟合超越量, 用极大似然估计方法估计相应的分布参数, 然后用自助采样技术计算每个估计的阈值对应的均方误差, 并用数值最小的均方误差所对应的阈值作为最优阈值.数据验证表明:与峰度法和超额均值函数法相比, 基于自助采样法选取的阈值, 其超越量的概率密度函数平均偏差分别降低了38.52%和29.25%, 且自助采样法所确定的超越量数据的QQ图更接近一条直线, 因此, 该方法提高了建模的灵活性及所选阈值的准确性, 而且当超越量数据不足时, 自助采样技术能够通过自动分析未知母体分布的统计学特性进行重新采样, 解决了因超越量数据不足导致的原始分布参数不稳定的问题.

References:

[1] Gomes M I, Guillou A. Extreme value theory and statistics of univariate extremes: A review[J]. International Statistical Review, 2014, 83(2): 263-292. DOI:10.1111/insr.12058.
[2] Mailhot A, Lachance-Cloutier S, Talbot G, et al. Regional estimates of intense rainfall based on the peak-over-threshold(POT)approach[J]. Journal of Hydrology, 2013, 476:188-199. DOI:10.1016/j.jhydrol.2012.10.036.
[3] Naess A, Haug E. Extreme value statistics of wind speed data by the POT and ACER methods[J]. Journal of Offshore Mechanics and Arctic Engineering, 2010, 132(4): 041604. DOI:10.1115/1.4001419.
[4] Bernardara P, Mazas F, Weiss J, et al. On the two step threshold selection for over-threshold modelling[C]// Proceedings of 33rd Conference on Coastal Engineering. Santander, Spain, 2012(33):42. DOI:10.9753/icce.v33.management.42.
[5] Scarrott C J, Macdonald A. A review of extreme value threshold estimation and uncertainty quantification[J]. REVSTAT–Statistical Journal, 2012, 10(1):33-60.
[6] You Z J. Discussion of “a multi-distribution approach to pot methods for determining extreme wave heights” by Mazas and Hamm, [Coastal Engineering, 58:385-394][J]. Coastal Engineering, 2012, 61:49-52. DOI:10.1016/j.coastaleng.2011.11.004.
[7] Mazas F, Hamm L. Reply to discussion by Z.J. You of “A multi-distribution approach to POT methods for determining extreme wave heights” by Mazas and Hamm[J]. Coastal Engineering, 2012, 65:16-18. DOI:10.1016/j.coastaleng.2012.02.008.
[8] Mazas F, Hamm L. A multi-distribution approach to POT methods for determining extreme wave heights[J]. Coastal Engineering, 2011, 58(5):385-394. DOI:10.1016/j.coastaleng.2010.12.003.
[9] Gomes M I, Figueiredo F, Martins M J, et al. Resampling methodologies and reliable tail estimation[J]. South African Statistical Journal, 2015, 49(1): 1-20.
[10] Li F, Bicknell C. A comparison of extreme wave analysis methods with 1994—2010 offshore Perth dataset[J]. Coastal Engineering, 2012, 69:1-11. DOI:10.1016/j.coastaleng.2012.05.006.
[11] Thompson P, Cai Y, Reeve D, et al. Automated threshold selection methods for extreme wave analysis[J]. Coastal Engineering, 2009, 56(10):1013-1021. DOI:10.1016/j.coastaleng.2009.06.003.
[12] Fukutome S, Liniger M A, Sueveges M. Automatic threshold and run parameter selection: a climatology for extreme hourly precipitation in Switzerland[J]. Theoretical and Applied Climatology, 2015, 120(3/4):403-416. DOI:10.1007/s00704-014-1180-5.
[13] Mooney C Z, Duval R D. A nonparametric approach to statistical inference[J].Technometrics, 1993, 36(4):435-436. DOI: 10.2307/1269981.
[14] Efron B, Tibshirani R J. An introduction to the bootstrap[J]. Chapman and Hall, 1993, 23(2):49-54. DOI:10.1111/1467-9639.00050.
[15] Caers J, Maes M A. Identifying tails, bounds and end-points of random variables[J]. Structural Safety, 1998, 20(1):1-23. DOI:10.1016/s0167-4730(97)00036-2.
[16] Shi D J. Practical methods of extreme value statistics[M]. Tianjin: Tianjin Science and Technology Press, 2006.(in Chinese)

Memo

Memo:
Biography: Wang Jixin(1975—), male, doctor, professor, jxwang@jlu.edu.cn.
Foundation item: The National Science and Technology Pillar Program of China(No.2015BAF07B00).
Citation: Wang Jixin, Wang Yan, Zhai Xinting, et al.Automatic determination method of optimal threshold based on the bootstrapping technology[J].Journal of Southeast University(English Edition), 2018, 34(2):208-212.DOI:10.3969/j.issn.1003-7985.2018.02.010.
Last Update: 2018-06-20