基于主成分分析的多参数岩体结构面优势分组方法研究

Multi-parameter dominant grouping method of rock mass discontinuity based on principal component analysis

对赋存于岩体内的结构面进行统计分析、分组聚类,是进行岩体工程稳定性分析的基础。考虑到结构面的力学和水力学性质受产状、迹长、张开度、粗糙度和充填状态等诸多因素共同影响,提出一种基于主成分分析的多参数岩体结构面优势分组方法。首先,利用主成分分析法选取多参数岩体结构面优势分组的判据,并计算其参与相似性度量的权重值。进而,基于退火遗传算法搜寻模糊C均值聚类算法的全局最优初始聚类中心。最后,以待分组结构面与聚类中心的距离加权和最小的方式建立目标函数,实现多参数岩体结构面优势分组。对计算机模拟的200个结构面进行多参数优势分组并与其他方法对比,计算结果表明,该方法具有较高的分组精度。将该方法用于重庆市三环高速花阳隧道实测结构面多参数优势分组中,分组结果合理、可靠,进一步验证了该方法具有较好的工程应用价值。

Statistical analysis and clustering of discontinuities existing in rock mass are the basis of rock mass engineering stability analysis.Considering that the mechanical and hydraulic properties of discontinuities are affected by many factors such as occurrence,trace length,opening,roughness,and filling state.A multi-parameter dominant grouping method of rock mass discontinuities based on principal component analysis is proposed.Firstly,the principal component analysis method is used to select the criterion of the dominant grouping and calculate the weight value of its participation in similarity measurement.Secondly,the global optimal initial clustering center of the fuzzy C-means clustering algorithm is searched using the annealing genetic algorithm.Lastly,the objective function is established by minimizing the weighted sum of the distance between the discontinuities to be grouped and the clustering center,achieving the dominant grouping of multi-parameter rock mass discontinuity.The 200 discontinuities simulated by the computer are divided into dominant groups using the multi-parameter method and compared with other methods.The results show that this method has higher grouping accuracy.The method is applied to the multi-parameter dominant grouping of the measured discontinuities of Huayang Tunnel of Chongqing Third Ring Expressway.The grouping results are reasonable and reliable,which further verify that the method has significant engineering application value.