分形方法在海洋溢油的气相色谱鉴别中的应用研究  

Application of fractal in identification of marine oil spill with gas chromatography

作　　者：李洪利[1] 王坚红[1] 程军[1] 刘刚[1] 高鑫鑫 

机构地区：[1]气象灾害省部共建教育部重点实验室,南京信息工程大学海洋科学学院,江苏南京210044 [2]南通海洋工作站,江苏南通226006

出　　处：《海洋科学》2012年第5期93-97,共5页Marine Sciences

基　　金：国家自然科学基金资助项目(40776017);国家908海洋综合调查项目子课题(2654);国家908扩展项目(908-01-BC21);南京信息工程大学科研基金项目(QD45);江苏高校优势学科建设工程资助项目

摘　　要：将分形理论应用于溢油的气相色谱鉴别,建立了一种新的海洋溢油气相色谱鉴别方法。即首先采用小波分解将气相色谱图曲线分解至不同的尺度,计算各尺度曲线分量的分形维数,基于分形维数参量,采用欧几里得距离判断不同样品气相色谱曲线的近似程度,达到对溢油样品进行鉴别的目的。溢油事故实例数据分析表明,分形维数能较好的反映不同色谱曲线的差异,所建立的方法在筛选溢油样品时,能在一定程度上缩小鉴别范围,可作为溢油鉴别的一种有效辅助方法。The fractal was used in gas chromatography （GC） analysis to identify marine oil spill. In this method, the gas chromatographic curves were decomposed firstly into different scales by wavelet decomposition and then the fractal dimensions of each scale were calculated. At last, the Euclidean distance between the different spilled oil samples, which can reflect the similarity between two samples, were obtained on the basis of fractal dimensions. Therefore, the selection of spilled oil samples became feasible using this method. Comparing the analysis results of two spilled oil events, the results of our method can significantly identify the sample of oil spill. So, the fractal di- mensions can show the difference of different chromatographic curves and our method is a supplemental way to reduce the identification range of oil samples, and should be a reliable factor to determine the source of oil spill.

关 键 词：溢油 分形维数 气相色谱 小波分解 

分 类 号：X55[环境科学与工程—环境工程]