论文部分内容阅读
采用主成分分析方法,就粘性土多指标反映其性质的规律进行了研究.研究表明,采用液性指数作为单一指标的传统粘性土物理状态划分方法,在反映亚粘土和亚砂土性质时不尽合理.而采用液性指数IL.结合孔隙比e反映粉质粘土的特性更加合理.同样,孔隙比e比液性指数IL能更好地描述亚粘土的天然特性.采用人工神经网络结合主成分分析,得出应用孔隙比e和液性指数IL两个指标来预测桩侧摩阻力更为精确.同时发现在一定临界影响深度范围内(20-30 m),桩侧摩阻力随深度的增加而增加,且粘性土的稠度愈硬,临界深度愈浅.
Principal component analysis method was used to study the law that the multi-index of cohesive soil reflects its nature.The research shows that the method of dividing the physical state of traditional cohesive soil using liquid index as a single index, when reflecting the properties of sub-clay and sub-sands Make reasonable use of liquid index IL combined with the porosity ratio e reflects the characteristics of silty clay is more reasonable Similarly, the porosity ratio e than the liquid index IL can better describe the natural characteristics of loam by using artificial neural network combined with the main Component analysis, it is found that the prediction of pile-side friction is more accurate by using two indexes of porosity ratio e and liquidity index IL. It is also found that within a certain critical depth of influence (20-30 m), the friction of pile- Increase and increase, and the harder the cohesive soil consistency, the more critical depth.