基于广义Pareto分布的桥梁车致荷载效应极值估计

摘要
图/表
参考文献
相关文章 (1)

全文: PDF (1100 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要 研究目的: 为准确估计桥梁的车致荷载效应极值和解决GPD模型阈值选择困难的问题,本文充分考虑过桥车辆的随机性特征,基于广义Pareto分布的极值预测理论,提出K-S检验自动选择阈值的方法和桥梁车致荷载效应极值估计的方法。同时,以一座简支T梁桥为工程应用算例,对随机车流荷载作用的桥梁车致荷载效应极值进行估计。
研究结论: (1)K-S检验自动选择阈值有效解决了广义Pareto分布阈值选择困难的问题,该方法计算效率高,有效避免了GPD模型的截尾现象;(2)GPD模型对桥梁车致荷载效应超阈值样本的拟合效果很好,其用于桥梁的车致荷载效应极值估计有效可靠;(3)桥梁的车致荷载效应极值随着重现期的增加而增大;在自由车流荷载作用下,多梁式简支梁桥行车道位置主梁的荷载效应极值大于其他位置主梁的荷载效应极值,其中边梁的承载力首超失效概率最大;为了保证桥梁在未来服役期的运营安全,类似桥梁在建造时可以提高行车道位置主梁的安全储备,运营时可以采取适当的交通管制措施;(4)提出的桥梁车致荷载效应极值估计方法可以在桥梁的安全评估、寿命预测和修复加固中应用,且在实际工程中应用方便。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	陈水生
	赵辉
	李锦华
	涂勇
	王屹昊

关键词 ：极值估计, 广义Pareto分布, K-S检验法, 阈值选择, 随机车流荷载

Abstract：Research purposes: In order to accurately estimate the extreme value of vehicle-induced load effect of bridge and solve the problem of difficult threshold selection of GPD model, based on the extreme value prediction theory of generalized Pareto distribution, the method of K-S test automatic selection threshold and the method of extreme value estimation of bridge vehicle-induced load effect were proposed in this paper. Then, taking a simply supported T-beam bridge as an engineering application example, the extreme value of vehicle-induced load effect of bridge under random traffic flow load was estimated.
Research conclusions: (1) K-S test automatically selects the threshold, which effectively solves the problem of difficult threshold selection of generalized Pareto distribution. This method has high computational efficiency and effectively avoids the truncation of GPD model. (2) The GPD model has a good fitting effect on the superthreshold samples of vehicle-induced load effect of the bridge, and it is effective and reliable for the extreme value estimation of vehicle-induced load effect. (3) The extreme value of vehicle-induced load effect of bridge increases with the increase of return period. Under the free traffic load, the extreme value of the load effect of the main beam at the carriageway of the multi beam simply supported beam bridge is greater than that of the main beam at other locations, and the first overload failure probability of side beam is the largest. In order to ensure the operation safety of the bridge in the future service period, similar bridges can improve the safety reserve of the main beam in the carriageway during construction, and take appropriate traffic control measures according to local conditions during operation.(4)The proposed method can be used in bridge safety assessment, life prediction and repair and reinforcement, and is convenient to be applied in practical engineering.

Key words： extreme value estimation generalized Pareto distribution K-S test method threshold selection random traffic load

收稿日期: 2021-01-07

PACS:

U441.3

基金资助:国家自然科学基金(12062006,11962006);江西省教育厅科技项目(GJJ204615)

作者简介: ^**陈水生,1968年出生,男,教授,博士生导师。

引用本文:

陈水生, 赵辉, 李锦华, 涂勇, 王屹昊. 基于广义Pareto分布的桥梁车致荷载效应极值估计[J]. 铁道工程学报, 2022, 39(3): 69-74.
CHEN Shuisheng, ZHAO Hui, LI Jinhua, TU Yong, WANG Yihao. Extreme Value Estimation of Vehicle-induced Load Effect of Bridge Based on Generalized Pareto Distribution. Journal of Railway Engineering Society, 2022, 39(3): 69-74.

链接本文:

https://tdgcxb.crec.cn/CN/ 或 https://tdgcxb.crec.cn/CN/Y2022/V39/I3/69

[1]	Nowak A S. Live Load Model for Highway Bridges[J].Structural Safety,1993(1-2):53-66.
[2]	Caprani C C,OBrien E J.The Use of Predictive Likelihood to Estimate the Distribution of Extreme Bridge Traffic Load Effect[J].Structural Safety, 2010(2):138-144.
[3]	鲁乃唯,刘扬,肖新辉. 实测车流作用下大跨桥梁荷载效应极值外推法[J].交通运输工程学报,2018(5):47-55.Lu Naiwei,Liu Yang,Xiao Xinhui. Extrapolating Method of Extreme Load Effects on Long-span Bridge under Actual Traffic Loads[J].Journal of Traffic and Transportation Engineering, 2018(5):47-55.
[4]	高欣,王磊.在役桥梁车辆荷载效应极值概率模型建模方法[J].哈尔滨工程大学学报,2013(8):995-999.Gao Xin,Wang Lei.Modeling Method for Extreme Ttraffic-load Effect Probabilistic Model of an Existing Bridge[J].Journal of Harbin Engineering University,2013(8):995-999.
[5]	Hosking J R M. L-Moments: Analysis and Estimation of Distributions Using Linear Combinations of Order Statistics[J].Journal of the Royal Statistical Society Series B -Statistical Methodology,1990(1):105-124.
[6]	袁伟璋,黄海云,张俊平,等.基于实际运营车辆荷载效应的既有桥梁可靠度研究[J].振动与冲击,2019(6):239-244.Yuan Weizhang,Huang Haiyun,Zhang Junping,etc.Study on the Reliability of an Existing Bridge Based on the Actual Operating Vehicle Load Effect[J].Journal of Vibration and Shock,2019(6):239-244.
[7]	Caers J,Beirlant J,Maes M A.Statistics for Modeling Heavy Tailed Distribution in Geology:Part I.Methodology[J].Mathematical Geology,1999,31:391-410.