论文部分内容阅读
为了克服目前地下水动态分类方法中存在的不能揭示分类指标空间到类型空间的非线性映射关系、方法复杂、计算量大等缺陷,可采用基于非线性变换的主成分投影(PCP)-聚类(C)模型,对地下水动态进行分类.方法首先对分类指标数据进行对数中心化变换,然后应用主成分投影法将变换后的多维指标向量映射到最优一维向量空间,并根据各样本指标在一维向量空间的投影值进行聚类分析,由此得到地下水动态分类结果.地下水动态分类结果表明,建议方法概念清晰,结构简单,计算简便,分类结果可信,是一种有效的地下水动态分类方法.
In order to overcome the shortcomings of the current dynamic groundwater classification methods that can not reveal the nonlinear mapping relationship between the classification index space and the type space, the method is complex and the calculation is large, the principal component projection (PCP) clustering based on nonlinear transformation C) model, and classify the groundwater dynamics.Methods Firstly, logarithmically transformed the classification index data, and then the principal component projection method is used to map the transformed multidimensional index vector to the optimal one-dimensional vector space. Based on each sample index The results of dynamic classification of groundwater indicate that the proposed method is clear in concept, simple in structure, simple in calculation and credible in classification results and is an effective groundwater dynamic Classification.