摘要
通过分析若干聚乙烯样品的拉曼光谱,对聚乙烯密度的检测方法进行了研究.分别采用拉曼光谱内标法(RISM)和偏最小二乘法(PLS)建立聚乙烯密度预测模型,其中RISM方法建模所得检测结果与真实值的相关系数为0.558,平均相对偏差为1.1%,标准偏差(SD)为11.7,而PLS方法建模所得检测结果的相关系数、平均相对偏差和交叉验证误差均方根分别为0.985、0.2%和2.2.研究表明:通过拉曼光谱能够实现聚乙烯密度的快速检测;且基于PLS的建模分析方法检测结果更优.
The detection methods for polyethylene (PE) density were studied by analyzing the Raman spectrum for different polyethylene samples. The prediction models for PE density were established by using Raman internal standard method (RISM) and Partial Least Squares (PLS) method. The correlation coefficient,average relative deviation (ARD) and standard deviation (SD) between actual densities of PE samples and predictive values obtained by RISM,were 0.558,11.1% and 11.7,while the values of correlation coefficient,ARD and root mean square error of cross validation(RMSECV) between actual densities and predictive ones obtained by PLS method were 0.985,0.2% and 2.2. The results indicated that the measurement on PE density by Raman spectrum is feasible,meanwhile,PLS method gives a better result than RISM.
出处
《浙江大学学报(工学版)》
EI
CAS
CSCD
北大核心
2010年第6期1164-1168,1184,共6页
Journal of Zhejiang University:Engineering Science
基金
国家自然科学基金资助项目(20776124
20490205
20736011)
国家"863"高技术研究发展计划资助项目(2007AA030208)
关键词
聚乙烯密度
拉曼光谱
偏最小二乘法
polyethylene density
Raman spectrum
partial least squares (PLS)