论文部分内容阅读
通过分析土体自然历史形成条件,从土性参数变异性表现特征着手,对其进行归类分析,初步回答了土性指标变异性的含义;对比分析了土层剖面的两种建模方法,据此提出了基于随机场理论的变异性统计方法,并通过实际工程勘探数据的统计分析,诠释了该方法的优越性所在。结果表明:土性指标变异性并非土体分层后层与层之间土性指标的差异,而是层内土性点与点之间的空间变异;相对于传统随机变量的建模方法,这种空间变异性用随机场理论建模更加科学合理;随机场理论用空间趋势函数和随机波动分量分别表征土性指标的确定性部分和随机扰动部分,通过去趋势化处理,可把握土性参数不确定性的核心;实例分析显示,本文提出的方法可更加精确地了解土性参数的不确定性,为可靠性理念在实际工程中的应用提供有利的途径。
By analyzing the formation conditions of natural history of soil and starting with the characteristics of variability of soil parameters, the paper classifies and analyzes the characteristics of soil parameters, and initially answers the meaning of the variability of soil indexes. By contrasting the two modeling methods of soil profiles, Based on this, a statistical method of variability based on random field theory is put forward and the superiority of this method is illustrated through the statistical analysis of actual engineering exploration data. The results show that the variability of soil index is not the difference of soil index between soil layer and layer, but the spatial variability of soil layer’s point and point. Compared with the traditional method of modeling random variables, This kind of spatial variability is more scientific and reasonable by modeling the random field theory. Random field theory uses the spatial trend function and the stochastic fluctuation component to characterize the deterministic part and the stochastic disturbance part of the soil index respectively. By de-trending treatment, The example analysis shows that the method proposed in this paper can understand the uncertainty of soil parameters more accurately and provide a favorable way for the application of reliability concept in practical engineering.