【摘要】 为充分掌握川南地区的土体特征参数,基于现场及室内试验资料,分析了川南喜德县境内的土体物理参数及固结参数的分布规律,并利用秩相关系数分析土体固结参数与其物理参数间的相关性,再采用BP神经网络建立了该地区土体参数对固结参数的预测模型。结果表明:该地区不同土体参数的规律性具有一定的差异性,且含水率和饱和度的区域变化相对较大,对土体的力学性质影响较大,尤其对土体的压缩变形影响较大,且粉质黏土压缩模量与其物理参数的相关性明显高于粉土压缩模量与其物理参数的相关性;同时,BP神经网络预测模型的预测精度较高,取得了较好的效果,为该地区土体参数规律性研究提供了一种新的思路。更多还原

【Abstract】 In order to fully grasp the characteristic parameters of the soil in South Sichuan,the soil physical parameters and the distribution of consolidation parameters in Xide County, South Sichuan were analyzed based on the field and indoor test data. And the rank correlation coefficient was used to analyze the correlation between soil consolidation parameters and its physical parameters. Then, the BP neural network was used to establish a prediction model of soil parameters for consolidation parameters. The results show that the regularity of different soil parameters in this area has a certain difference. And the regional changes of water content and saturation are relatively large, which have great influence on the mechanical properties of soil, especially on the compression deformation of soil. Then the correlation between the compressive modulus of silty clay and its physical parameters is significantly higher than that of silt compressive modulus and its physical parameters; at the same time, the prediction accuracy of the BP neural network prediction model is higher and a better effect has been obtained, which provides a new idea for the study of the regularity of soil parameters in this area.更多还原