基于主成分聚类分析的采矿方法优选

 采矿方法优选涉及到多指标体系的分类及综合评价问题,利用主成分分析简化了指标结构,将主成分分析与聚类分析相结合,提出了主成分聚类分析法,并基于该方法对来自某矿山的15个试样的采矿方法进行了优选。在此过程中,针对传统主成分分析方法的缺点和应用中可能出现的误区,通过均值化改进了主成分分析的特征提取,通过以主成分得分为新的数据基础做聚类分析改善了综合评价效果;对主成分含义给出了较为明确的解释;对主成分聚类、第一主成分得分、主成分综合得分的排序结果进行了对比分析。研究表明,主成分聚类分析法则既可以对多变量数据进行合理地分类,又能对各类优劣程度做出综合评价,能充分反映矿山的实际情况,终选出的采矿方法在工业试验后成效显著,验证了该决策方法是切实可行的。

Optimization of Mining Methods Based on Principal Composition Cluster Analysis

In view of the fact that the optimization of the mining methods involves the classification and comprehensive evaluation of a multi-indicator system, in this paper, the target structure is simplified with the principal component analysis, combined with clustering analysis, and the principal component cluster analysis method is proposed, and then the mining methods for 15 samples from a mine are optimized based on this method. In the process, to avoid the shortcomings of the traditional principal component analysis and the errors which may occur in applications, the feature extraction of the principal component analysis is improved by means of the equalization and the comprehensive evaluation is improved by making clustering analysis based on the principal component scores. The meaning of the principal component is interpreted clearly. Moreover, the results of the principal component clustering analysis, the first principal component scores and the principal component composite scores are ranked and analyzed. It is shown that the principal component cluster analysis not only can classify the multivariable data reasonably, but also can make a comprehensive assessment of the performance of various types to fully reflect the actual situation of the mine. The final selected mining method sees a remarkable improvement according to industrial tests, which verifies this decision-making method.