基于PCA的时间分辨油荧光光谱分析及优化

发布时间：2018-04-09 13:50

  本文选题：时间分辨荧光光谱　切入点：油荧光分类　出处：《光学精密工程》2017年04期

【摘要】：激光诱导荧光技术可广泛应用于油污染的监测中,然而普通的油荧光光谱技术只能实现油污染监测的粗分类,无法区分原油与燃料油的荧光特征。本文基于主成分分析方法(PCA)的时间分辨油荧光分类方法,实验测量了20种油样本的时间分辨荧光光谱特征,给出了对应的荧光寿命和时间分辨油荧光光谱的时序特征。在此基础上,利用前三个主成分构成的三维特征矢量空间,通过分析不同采集时刻下油样本矢量间相关距离的变化,对油样本的时间分辨荧光光谱进行聚类分析。为了体现油荧光变化的时序性,引入矢量距离的离散度参量,提出基于PCA进行时间分辨油荧光光谱分析的优化方法。实验结果表明,基于时间分辨油荧光光谱识别可实现原油与燃料油的光谱时序特征区分,具备良好的油荧光分类效果。
[Abstract]:Laser induced fluorescence technology can be widely used in oil pollution monitoring, but ordinary oil fluorescence spectroscopy can only realize the crude classification of oil pollution monitoring, and can not distinguish the fluorescence characteristics between crude oil and fuel oil.Based on the time-resolved oil fluorescence classification method based on principal component analysis (PCA), the time-resolved fluorescence spectrum characteristics of 20 kinds of oil samples were measured experimentally, and the corresponding fluorescence lifetime and time-resolved oil fluorescence spectrum timing characteristics were given.On this basis, the time-resolved fluorescence spectra of oil samples were analyzed by using the three dimensional feature vector space of the first three principal components, and by analyzing the variation of the correlation distance between the oil samples at different sampling times.In order to reflect the timing of oil fluorescence variation, an optimized method of time-resolved oil fluorescence spectrum analysis based on PCA was proposed by introducing the dispersion parameter of vector distance.The experimental results show that the time-resolved oil fluorescence spectrum recognition can realize the spectral time series distinction between crude oil and fuel oil, and has a good oil fluorescence classification effect.
【作者单位】： 中国海洋大学信息科学与工程学院;国家海洋局第一海洋研究所;中国科学院海洋研究所;
【基金】：国家自然科学基金资助项目(No.61505221) 国家海洋局国际海洋合作与履约项目(No.QY0516014) 中央级公益性科研院所基本科研业务费专项资金资助项目(No.0215G20)
【分类号】：X834;X55
，

本文编号：1726696