岩土层原位导热系数的多元线性回归分析方法

MULTIPLE LINEAR REGRESSION ANALYSIS METHOD FOR IN-SITU THERMAL CONDUCTIVITIES OF ROCK AND SOIL LAYER

由于取样的扰动和室内试验方法的局限性，用岩土样品导热系数代表岩土层原位有效导热系数存在明显误差。基于“杭州市浅层地温能调查评价项目”中40个热响应试验孔的地层结构和现场热响应试验成果，用多元线性回归分析理论和行业数据统计SPSS软件（statistical product and service solutions）程序，以各岩土层厚度为自变量、原位有效导热系数为回归系数，热响应试验所确定的试验孔岩土的综合导热系数为因变量，确定各岩土层的原位有效导热系数，回归方程和回归系数显著可信，可用于计算统计区域内其他地埋管热交换工程孔岩土的综合导热系数。

Due to the perturbation of sampling and the limitations of laboratory test methods, there is a significant error in using the geotemperature coefficient of geomaterials to represent in-situ effective thermal conductivity of rock and soil layers. Based on the stratum structure and on-site thermal response test （TRT） results of 40 thermal response test holes in the ＂Hangzhou Shallow Geothermal Energy Survey and Evaluation Project＂ , using multivariate linear regression analysis theory and industry statistics SPSS software （Statistical Product and Service Solutions） , the thickness of each rock and soil layer as independent variables, in situ effective thermal conductivity as the regression coefficient, experimental hole geotechnieal comprehensive thermal conductivity obtained by TRT as dependent variable to determine in-situ effective thermal conduetivities of each rock and soil layer. It was shown that the regression equation and the regression coefficients are significant, which can be used to calculate the comprehensive thermal conductivity of rock and soil in other buried pipe heat exchange engineering hole in the statistical area.