时间序列三维荧光光谱的数据压缩

发布时间：2018-03-19 08:53

本文选题：三维荧光光谱　切入点：聚类分析　出处：《光谱学与光谱分析》2017年04期 　论文类型：期刊论文

【摘要】：时间序列三维荧光光谱数据量大和信息丰富的特点虽然有利于有机物的定性和定量分析,但是大量冗余信息的存在增加了计算的复杂度和计算量。在分析了时间序列三维荧光光谱的时频特征后,分别利用聚类分析和二维小波变换从时间维和光谱维对三维荧光光谱进行了压缩。在聚类分析中探讨了样本距离、类间距离、复合相关系数和R平方统计量等关键因素。相关系数和R平方统计量的结合不仅提高了聚类分析的精确度,而且减少了二维小波在光谱维进行数据压缩的工作量。相关的数值实验表明压缩后的数据保留了原有时间序列三维荧光光谱的重要信息。
[Abstract]:The large amount of time series fluorescence spectrum and the rich information are beneficial to the qualitative and quantitative analysis of organic matter. However, the existence of a large amount of redundant information increases the computational complexity and computational complexity. The three dimensional fluorescence spectra were compressed by cluster analysis and two dimensional wavelet transform from time and spectral dimension respectively. In the cluster analysis, the sample distance and the distance between clusters were discussed. The combination of correlation coefficient and R square statistic not only improves the accuracy of cluster analysis, but also improves the accuracy of clustering analysis. In addition, the workload of data compression in spectral dimension of two-dimensional wavelet is reduced. The relevant numerical experiments show that the compressed data retain the important information of the original time series 3D fluorescence spectrum.
【作者单位】：合肥师范学院数学与统计学院;中国科学院安徽光学精密机械研究所环境光学与技术重点实验室;
【基金】：国家自然科学基金项目(61308063) 安徽省自然科学基金项目(1308085QF111) 国家自然科学基金委员会国际(地区)合作与交流项目(61491240110)资助
【分类号】：O657.3

【相似文献】