非线性偏最小二乘法在混合炸药体系液相色谱中的应用

Nonlinear Partial Least Squares Method for HPLC Analysis of compositeexplosives

本文着重讨论了非线性偏最小二乘法(简称NLPLS)在混合炸药体系定量分析领域里的应用。以HMT、RDX、Tetryl、TNT等常用炸药为对象,对炸药的高效液相色谱分析中所遇到的复杂体系如谱峰严重重叠体系、待分析物含量相对悬殊体系等进行了研究。结果表明,本方法对混合炸药体系的液相色谱分析数据具有较强的处理能力,对完全重叠色谱的解析置信度取95％,误差小于5％。为解决这类炸药的液相色谱定量分析问题增添了另一条途径。

Partial least square method (NLPLS) was used to characterize the quantitative chromatographic analysis results of composite explosives. The well-known HMX RDX Tetryl TNT was taken as examples to examine the reliability of the method on resolving the complicated problems in high performance liquid chro-matography (HPLC) caused by the large range peaks overlap or great disparity of the component's content. The experimental results indicate that the resolution ability of NLPLS to process the HPLC data is better than the commonly used method and its probability, of resolving the complete overlapped HPLC data is 95% with deviation less then 5%.