基于数据挖掘技术的滑带土抗剪强度预测

PREDICTION OF SHEAR STRENGTH OF SLIP ZONES USING DATA MINING TECHNOLOGY

利用数据挖掘技术对三峡库区重庆段滑带土抗剪强度特征值进行预测研究。利用CHAID算法与相关性分析进行数据预处理。根据CHAID算法分类结果,将滑坡按照原岩的沉积环境进行分类。在分类的基础上,根据各定量指标与抗剪强度的Pearson相关系数筛选滑带土抗剪强度的影响因素。相关性分析结果表明,原岩为不同沉积环境的滑坡其天然状态下内摩擦角的影响因素不同;天然状态下的黏聚力与各定量因素相关性均不高。将相关性高的指标作为输入变量,建立数据挖掘模型。研究结果表明,挖掘出的知识具有良好的适用性。

The data mining techniques were used to predict shear strength of slip zone soil. In order to define the qualitative and quantitative factors affecting the shear strength,the CHAID（ Chi-squared Automatic Interaction Detector） method and correlation analysis were used to pretreat data. According to the result obtained by CHAID method the landslides are classified by sedimentary environment of slip bed. Based on the classification,the influencing factors were selected in the light of Pearson correlation coefficient. The correlation analysis result shows that the internal friction angle of slip zone is affected by different factors as the slip bed of different sedimentary environment. The cohesive force has no obvious relationship with the factors. Stepwise regression was carried out to establish regression model for internal friction angle. The regression equations are passed statistical test and of good fitting degree and prediction accuracy. The forecast equations for the actual project data have good performances.