关键词: 振冲碎石桩;钻进比功;聚类算法;评估指标;地层辨识;

Abstract: In the treatment for deep overburden foundations by vibro-replacement stone column, effective stratigraphic identification is crucial for the determination and real-time control of construction process control parameters such as filling amount, control current, vibration retention time and reverse insertion depth.This study analyzed the drilling energy consumption in different strata based on real-time monitoring data collected from the construction process of vibro-replacement stone column.The concept of drilling ratio work was introduced and various clustering algorithms were employed to conduct clustering analysis of different depth strata in three dimensions: depth of penetration, current and drilling ratio work.In the analysis of stratigraphic identification of vibro-replacement stone column performed by different clustering algorithms, the performance of the clustering methods can be evaluated using three clustering accuracy indicators: contour coefficient, DB index and CH index.Finally, the ANGENS clustering algorithm was determined as the optimal stratigraphic identification technique for the drilling procedure of vibro-replacement stone column.The results provide an important reference for the massive data mining and analysis during the construction process of vibro-replacement stone column, and supply an important technical support for the quality control of vibro-replacement stone column.

Key words: vibro-replacement stone column; drilling ratio work; clustering algorithm; evaluation index; stratigraphic identification;