区域软土地基震陷评估的条件广义方差极小-盲数耦合分析

Coupling model based on conditional generalized minimum variance and blind number theory for evaluating the regional foundation subsidence in soft soil area

针对区域软土地基震陷评估中存在的问题,选取条件广义方差极小法和盲数理论作为基础模型,提出一种耦合评估思路,并设计和建立基于少量参数的耦合式评估模型。该模型首先采用条件广义方差极小法筛选出最具影响性的因素,然后通过完全二次项回归方法拟合出各关键影响因素与震陷量之间的计算关系;最后根据盲数理论计算震陷量的可能值与可信度,从而实现区域地基土震陷等级评价。为了验证耦合模型的适用性和准确性,对实际样本数据进行建模计算,取得较好的评价结果,为区域地基土震陷评估提供一种新的思路。

Coupling model with a few parameters, which was based on conditional generalized minimum variance method and blind number theory, was established in this paper to evaluate the regional foundation subsidence in soft soil area. Firstly, the most critical influences were filtered by conditional generalized minimum variance method. Secondly, the relati onship between key factors and seismic depression were calculated a ccording quadratic regression model. Finally, possible values and faith degree of seismic depression were obtained by blind number theory to evaluate the regional foundation settlement. The applicability and accuracy of the model were proved by instance analysis and the results showed that effect was good. The method in this paper provided a new way for evaluating the regional foundation settlement.