基于主成分分析的岩石质量综合评价模型与应用

主成分分析法能够在保证原始数据信息损失最小的情况下,以少数的综合变量取代原有的多维变量,使数据结构大为简化。试验选取了能间接反映岩石质量的5项主要指标,分别为岩石的干密度、变形模量、饱和吸水率、干抗压强度和纵波传播波速。采用主成分分析法对这些间接指标进行相关的数值处理,得出了评价岩石质量的数学模型。在对16组岩石样品的5项指标进行室内测试后,运用该模型对岩石样品的岩石质量进行了评价,并对该16组岩石样品的岩石质量进行了排序,将评价结果与实际样品进行对比,结果表明,该模型的评价结果与岩石的实际质量有很好的符合性。主成分分析法用于评价岩石质量有较高的可信度,并能够客观、准确、迅速地评价某种岩石的岩石质量。

A model for comprehensively evaluating rock quality base on principal component analysis and its application

A few integrative variables can substitute for original multidimensional variables by principal component analysis with little data information lost; and the data organizations are relatively simple by this method. In order to get the main factors that reflect the quality of rock, the principal component analysis is put into use. Five indirect factors on rock quality such as dry density, distortion module, saturation drink rate, dry resistance strength and longitudinal wave velocity are chosen to test in laboratory. The main parameters have been found after processing those indirect factors by principal component analysis; and the evaluation model for rock quality has been got. The quality of rock is evaluated by the model base on principal component analysis, and the quality sort of sixteen groups of rock samples is given. It is shown that the evaluation results by the model and the actual quality of rock samples are well coherent. The model has good reliability on the evaluation of the quality of rock. It can evaluate the quality of rock quickly, impersonally and truly.