考虑交通量增长的非平稳荷载效应极值研究

Extreme value of non-stationary bridge vehicle load effect considering traffic flow growth

为准确获取桥梁车载效应极值分布,提出了考虑交通量增长引起的非平稳极值模型。基于传统广义帕累托分布(GPD)采用"时间离散"与"概率联合"思想构建改进广义帕累托分布(AGPD)模型,并进行数值验证;运用AGPD模型对随机车流作用下的荷载效应进行极值外推分析。结果表明:交通量增长使荷载极值呈现非平稳性,破坏了独立同分布假定(IID),传统GPD模型不再适用,而AGPD模型可准确地进行非平稳极值预测分析;在车载分布不变前提下,交通量年增长率分别为0%、1.0%、2.0%、3.0%、4.1%、5.0%、6.0%时,重现期为1950年的荷载效应外推值为《公路桥涵设计通用规范》(JTG D60—2015)规范给定值的1.55倍~1.58倍。

Here,to accurately obtain the extreme value distribution of bridge vehicle load effect,a non-stationary extreme value model considering growth of traffic flow was proposed.Based on the traditional generalized Pareto distribution(GPD),the amend generalized Pareto distribution(AGPD)model was constructed by using the idea of“time discretization”and“probability combination”,and the numerical verification was performed.AGPD model was used to do the extremum extrapolation analysis for vehicle load effect under random traffic flow.The results showed that increase in traffic flow makes the vehicle load extremum be non-stationary to destroy the IID assumption;the traditional GPD model is no longer applicable,while AGPD model can accurately predict and analyze the non-stationary extremum;under the premise of the same vehicle load distribution,when the annual growth rates of traffic flow are 0%,1.0%,2.0%,3.0%,4.1%,5.0%and 6.0%,respectively,the extrapolation value of vehicle load effect with return period of 1950 a is 1.55-1.58 times of the given value in the general code for design of highway bridges and culverts(JTG d60—2015).