摘要
刀具磨损是盾构在极硬岩地层中施工遇到的难题之一,因此对刀具磨损的分析与预测对盾构施工意义重大。通过对大连地铁一号线101标段港湾广场站~中山广场站区间隧道盾构刀具进行磨损监测,利用偏最小二乘回归法,对在不同盾构掘进情况下刀具磨损的多组数据进行分析。根据自变量载荷因子情况,分别选取刀具磨损量为因变量,盾构机的贯入度p、推进速度v、刀盘转速n、盾构推力F、盾构扭矩T、土仓压力Ps、刀具半径L等掘进参数为自变量,选用部分自变量和全部自变量组合用于建模,并建立了偏最小二乘回归模型。筛选样本数据和全部样本数据,分别对建立的刀具磨损模型进行验证,并通过已有数据对刀具磨损进行分析与预测。研究结论表明在极硬岩地层中采用偏最小二乘法可以比较合理地选择自变量,同时建立的数学模型也能较好地对不同盾构掘进情况下的刀具磨损进行计算。
Shield cutters wear is one of the difficult problems in extremely hard rock of shield construction. Analysis and prediction on tool wear is significant for shield construction. By monitoring the abrasion of shield cutters used in shield tunnel section of the Harbour Plaza station to Zhongshan square station of 101 section of Dalian metro line 1, analyzed multiple sets of the cutters wear data in different cases of shield tunneling by using the Partial Least Squares method. According to the independent variable load factor cases, According to the independent variable load factor cases, the abrasion of shield cutters is dependent variable;the penetration of shield machine,advancement rate,cutter-head rotation speed,total thrust,cutter torque,earth pressure,cutting tool radius are independent variables.Some independent variables and all combinations of variables for modeling were selected, and the Partial Least Squares regression model was established by using cutters wear as the dependent variable. Adopting selected sample data and all sample data, respectively, the established model is verified, and through the existing data the cutters wear were analyzed and predicted. The research conclusion showed that the selection of independent variable by Partial Least Squares method in extremely hard rock is quite reasonable, and the established cutters wear model can calculate the wear of shield cutters in different shield tunneling cases.
出处
《土木工程学报》
EI
CSCD
北大核心
2015年第S1期250-255,共6页
China Civil Engineering Journal
基金
国家自然科学基金资助项目(51178027)
关键词
极硬岩地层
盾构刀具
偏最小二乘法
刀具磨损
分析预测
extremely hard rock
shield cutter
Partial Least Squares method
shield cutters wear
analysis and prediction
