摘要
提出一种基于Bootstrap重抽样技术的岩土参数标准值确定方法,将置信概率为95%对应的重抽样均值作为标准值,并通过一组实测土体的抗剪强度参数阐述标准值的确定方法和过程。对两个实例进行研究,分别探讨Bootstrap重抽样次数(即Bootstrap样本集大小)和原始数据的变异性对标准值精度和收敛性的影响。结果表明,基于Bootstrap方法确定的岩土参数标准值具有较好的收敛性和精度,当重抽样次数不小于100时,Bootstrap重抽样次数和原始样本的变异性对标准值的收敛性和精度不会产生显著影响。该方法能够有效地处理岩土参数的不确定性问题。
Scientific and rational determination of the representative values of geotechnical param- eters constitutes a premise for the design of a safe and economic project. In this paper, an ap- proach based on a Bootstrap resampling technique is proposed to determine the standard values of geotechnical parameters that have a confidence probability of 95% over the Bootstrap sample mean. First, the Bootstrap method is briefly introduced and a set of shear strength parameters is utilized to elaborate the procedures. Two examples were analyzed to assess the influence of the number of Bootstrap simulations (i.e., the size of bootstrap samples} and the variability of origi- nal data on the convergence and accuracy of standard values. It was shown that the standard val- ues determined using the proposed Bootstrap method have good convergence and accuracy. When the number of Bootstrap simulations was at least 100, the number of bootstrap simulations and the variability of original data had a negligible impact on the convergence and accuracy of standard values. Therefore, this method is capable of assessing the uncertainty in geotechnical parameters.
出处
《地震工程学报》
CSCD
北大核心
2015年第B07期7-11,共5页
China Earthquake Engineering Journal