既有空间结构位形推算的空间系统抽样方法

Spatial System Sampling Method for Reckoning Geometric Shape of Existing Spatial Structures

对于既有空间结构整体位形推算问题,根据不同节点位置偏差的相关性,提出了空间系统抽样方法及基于空间系统抽样的统计推断方法.基于现有关于空间结构节点位置偏差的研究,结合空间抽样理论,针对既有空间结构现场检测的抽样方法,提出抽样分布模式选取原则与最小样本容量的计算公式;以节点偏差的相关系数矩阵为基础,引入基于正态总体的非独立样本的假设检验理论,通过统计推断建立既有结构精度先验的半实测半概率模型.对一单层5环K6型网壳结构试验模型,采用多种数值建模方法建立计算模型,并进行整体稳定极限承载力对比分析,验证了基于空间系统抽样方法构建的数值模型更接近实际结构.

For reckoning geometric shape of existing spatial where different node position deviations have correlation, SSS (spatial system sampling) methods and statistical inference methods based on SSS are proposed. In a comprehensive consideration of the research on spatial structure node position deviation and the space sampling theory, the sampling mode and the minimum-sample-size formula are given for on-site detection of existing spatial structures. Based on the node position deviation correlation coefficient matrix and the hypothesis testing theory of non-independent samples under normal population, a priori accuracy semi-probabilistic model of the existing structure is established by statistical inference. For a single-layer 5-ring K6 reticulated shell test model, a variety of numerical modeling methods are used to establish the calculation model for conducting comparative analysis of the overall stability ultimate bearing capacity, which verifies that the numerical model based on the spatial system sampling method is closer to the actual structure.