基于克里金插值法的常压刀盘刀梁区域泥饼发育程度算法及应用研究

Calculation Method for Mud Cake Development Degree at Cutter Beam Area of Cutterhead Under Atmospheric Pressure Using Kriging Interpolation and Its Application

为解决刀盘结泥饼影响盾构施工进度的问题,以某泥水盾构刀盘为研究对象,展开对刀盘刀梁部位泥饼发育程度的研究,提出一种基于克里金插值法的常压刀盘刀梁区域泥饼发育程度算法,并进行应用案例研究。利用全刀盘布置温度传感器获取刀盘温度数据,选取线性插值扩展刀梁区域温度数据,运用克里金插值法进行全刀盘区域的温度仿真计算,结合空间和时间2个维度分析刀盘在正常掘进、外侧结泥饼及内侧结泥饼3种工况下的判断条件,提出分离相邻测温点正/负差值,以刀盘整体正/负差值之和呈现同方向变化作为泥饼发育初始环评判指标;结合所建温度场,引入从泥饼发育初始环到当前环的温度变化率总和作为泥饼发育程度评判指标。针对某大直径常压刀盘泥水盾构案例进行分析判定,对比采用反距离权重插值法、克里金插值法和扩展样本克里金插值法构建的3种温度场用于泥饼发育程度计算,结果表明扩展样本后的克里金插值法的适用性最强,并与工程实际情况进行对比,准确率达70.87%。

The presence of mud cake on the cutterhead considerably affects shield tunneling operations.Consequently,a comprehensive case study is performed to investigate the development of mud cake on the cutter beam area of a slurry shield cutterhead operating under atmospheric pressure.A calculation methodology using Kriging interpolation is established.First,a temperature sensor is positioned on the cutterhead to collect temperature readings.Then,this data is extrapolated through linear interpolation and analyzed using the Kriging interpolation technique.Finally,the criteria for assessing the degree of mud cake development on the cutterhead during normal excavation,peripheral mud cake,and inner mud cake stages are delineated spatiotemporally.A method involving the identification of positive and negative differences between adjacent temperature readings is introduced to obtain an evaluation index for the initial ring′s mud cake development,represented by the cumulative sum of temperature variations across the cutterhead.Using the established temperature distribution data,the cumulative temperature change rates from the initial ring to the current mud cake-developed ring are employed as an index of mud cake progression.Through a detailed case study,the results of inverse distance weighting,Kriging interpolation,and Kriging interpolation with expanded samples are evaluated and compared,revealing that Kriging interpolation with expanded samples attains the highest accuracy of 70.87%.