主成分分析法在估算石脑油裂解参数中的应用

The application of principal component analysis (PCA) on computing of the parameters in naphtha pyrolysis

在利用Kumar分子反应动力学模型模拟石脑油裂解过程中,一次反应选择性系数是最重要的参数,它受裂解石脑油性质、操作条件等多种因素的影响。本文提出利用主成分分析法将多种影响因素转化为几个不相关的综合因素-主成分,同时保留尽可能多的原始信息。实例验证显示,利用主成分估算一次反应选择性系数更加方便、准确,有利于提高石脑油裂解模拟的精度。

Based on Kumar molecular kinetic model used in Naphtha Pyrolysis process, the effect of Naphtha properties and operating conditions on computing initial selectivities has been investigated. In this article, these factors are transformed to a few less-related indices with Principal Component Analysis （PCA） , which are called principal components. The case study show, it is more ease and more accurate to estimate initial selectivitics in this application, also making the accuracy of Naphtha Pyrolysis simulation higher.