应用Monte Carlo模拟法探讨分层性物质的组合取样精度

Applying Monte Carlo Simulation Technique to the Study of Composite Sampling for Chemical Analysis of Stratified Materials

应用MonteCarlo模拟法研究了分层性物质的组合取样精度，探讨了组合样中组分含量的分布规律、组合取样方差的分布规律、组合取样方差估计值的精度与组合样本数目之间的关系等．考察了组分含量服从正态分布、均匀随机分布及多项分布的分层性总体．结果表明，当样本数目较多时，组合取样误差规律对于不同原始分布的总体是相似的．

Monte Carlo simulation technique was applied to the study of composite sampling for chemical analysis of stratified materials. The distribution of the concentration of analyte in the composite samples, the distribution of the composite sampling variance estimates and the effect of the number of composite samples on the precision of composite sampling variance estimates were investigated. Populations with Gaussian, random and polynomial distributions were studied in this paper. The sampling rules are found to be similar to different parent distributions if the number of samples taken is large.