针对填料变异特征的连续压实检测指标优化研究

Optimization of continuous-compaction detection index for variability of fillers

为解决填料变异性导致的连续压实检测指标(CMV)准确性低的问题,优化CMV在路基压实质量检测中的评价效果,针对填料变异性的特点,以滑动窗口D作为计算参数并提出3种优化指标:平均值指标μ_CMV,反距离加权平均值指标ω_CMV和普通克里金插值指标k_CMV。根据兰张铁路路基施工段A采集的CMV数据与动态变形模量Evd数据,采用半变异性分析探究CMV的空间自相关性以及滑动窗口D的取值范围。通过对比不同滑动窗口下3种优化指标和CMV的变异系数以及与Evd的相关性系数的变化规律,研究各优化指标和CMV的适用性。研究结果表明:滑动窗口D的取值应以CMV的最大空间自相关距离为依据,且D宜小于2倍空间自相关距离,即11.18 m。随滑动窗口D的增大,μ_CMV以及ω_CMV与Evd的相关性不断提高,k_CMV与Evd的相关性较差且基本保持不变,其中μ_CMV与Evd的相关性提高最显著,且当D取9.5 m时,相关性系数达到峰值为0.763。CMV与Evd的相关性系数仅为0.120,远低于μ_CMV的峰值,表明μ_CMV与常规检测指标的对应关系更好。相同滑动窗口下,μ_CMV的变异系数最低,表明平均值指标μ_CMV更适用于表征路基压实质量。通过在施工段B上预设薄弱区域,进一步验证μ_CMV对路基压实质量的检测准确性,同时建立μ_CMV检测系统,可实现μ_CMV实时计算并辅助现场施工。研究结果可为连续检测技术的工程应用提供参考。

The variability of subgrade fillers has a significant effect on CMV,which results in low accuracy.To improve the accuracy of CMV in the quality detection of subgrade compaction,three optimized indexes were proposed with the sliding window D,whichwere the average indexμ_CMV,the inverse distance weighted average indexω_CMV,and the ordinary kriging interpolation index k_CMV.The data of CMV and Evd(dynamic deformation modulus)were collected from Section-A of the Lanzhou-Zhangye Railway.Based on these data,the spatial self-correlation of CMV was analyzed by the semi-variance analysis,and the range of the sliding window was determined.Under different sizes of the sliding window,the correlation analysis between the three optimized indexes,CMV,and Evd was performed.Their coefficients of variation were calculated,by which the applicability of optimized indexes and CMV was investigated in detail.The results show that the maximum spatial selfcorrelation distance of CMV should be taken as the basis for determining the value of D,and D should not exceed twice the distance consequently.The correlation betweenμ_CMV(andω_CMV)and Evd enhances with increasing D,while the correlation between k_CMV and Evd is weak and stable.The strongest correlation is observed betweenμ_CMV and Evd,and the correlation coefficient peaks at 0.763 when D is taken as 9.5 m.The correlation coefficient between CMV and Evd is merely 0.120 and lower than that ofμ_CMV,indicating the correlation betweenμ_CMV and Evd is better.Given that the coefficient of variation ofμ_CMV is minimum under the same sliding window,μ_CMV is a better choice for evaluating the compaction quality,and the case study in Section-B successfully validates the detection accuracy ofμ_CMV.Furthermore,a detection system based onμ_CMV has been established,which can calculateμ_CMV in real-time and support the on-site construction.This study can provide a reference for the engineering application of continuous detection technology.