当前位置:主页 > 社科论文 > 生态环境论文 >

用离散小波变换建立的Fisher判别法对海上溢油的鉴别

发布时间:2018-03-05 07:48

  本文选题:荧光特性 切入点:离散小波变换 出处:《光谱学与光谱分析》2017年11期  论文类型:期刊论文


【摘要】:采用恒波长同步荧光光谱法检测分析8种燃料油、7种中东原油、14种非中东原油的荧光特征,结合离散小波变换以及Fisher判别法建立海上溢油油种鉴别的模型。29种油样风化前后均在波长(280±2),(302±2),(332±2)和(380±2)nm处有典型的荧光峰,但在(380±2)nm处风化油样的荧光强度的离散度过大,该波长不适于油种鉴别。在db7小波基函数下对29种原始油样荧光谱图进行6层分解,提取d3细节系数特征,确定波长(255±2),(280±2),(302±2),(332±2)和(354±2)nm处的小波系数并用于Fisher判别模型建立。29种油样在(280±2)nm处均有极值点,燃料油小波系数位于44.06±5.62之间,原油位于22.47±5.12之间,此波长处的小波系数可区分燃料油与原油。建立的Fisher判别模型不仅能区分燃料油和原油还能进一步区分中东原油,Wilks's lambda分布所对应的P值分别为0和0.02,表明模型是可行的。模型验证结果显示,对风化后的建模油样的鉴别正确率达到96.6%,对非建模23种油样鉴别正确率达到95.7%。由于建模油样风化前后的修正余弦相似度为0.91~0.98,因而以未风化油样建立的油种鉴别模型同样适用于风化后油样的辨别。
[Abstract]:The fluorescence characteristics of 14 kinds of non-Middle Eastern crude oil from 8 kinds of fuel oil and 7 kinds of Middle East crude oil were detected and analyzed by constant wavelength synchronous fluorescence spectrometry. Combined with discrete wavelet transform (DWT) and Fisher discriminant method, the model of oil spill identification at sea. 29 oil samples have typical fluorescence peaks at wavelength of 280 卤2, 302 卤2, 332 卤2) and 380 卤2 nm before and after weathering, but the dispersion of fluorescence intensity of weathered oil samples at 380 卤2 nm is very large. The wavelength is not suitable for oil identification. The fluorescence spectra of 29 original oil samples are decomposed into six layers under db7 wavelet basis function, and the characteristics of d3 detail coefficients are extracted. The wavelet coefficients at the wavelength of 255 卤2, 280 卤2 ~ 2, 302 卤2 ~ 2 and 354 卤2 ~ 2 nm were determined and used to establish a Fisher discriminant model. All 29 oil samples had extreme values at 280 卤2 ~ 2 nm. The wavelet coefficients of fuel oil were between 44.06 卤5.62 and 22.47 卤5.12, and the wavelet coefficients of fuel oil were in the range of 44.06 卤5.62 and 22.47 卤5.12, respectively, and the results showed that the wavelet coefficients of fuel oil were in the range of 44.06 卤5.62 and 22.47 卤5.12, respectively. The wavelet coefficients at this wavelength can distinguish fuel oil from crude oil. The established Fisher discriminant model can not only distinguish fuel oil from crude oil, but also further distinguish Middle East crude oil Wilksworth's lambda distribution corresponding to P values of 0 and 0.02, respectively. Feasible. Model validation results show that, After weathering, the accuracy rate of oil sample identification was 96.6, and that of 23 unmodeled oil samples was 95.70.The modified cosine similarity before and after weathering of the model oil sample was 0.910.98, so the oil identification model was established by using unweathered oil sample. Type A also applies to the identification of oil samples after weathering.
【作者单位】: 大连海事大学环境科学与工程学院;
【基金】:中央高校基本科研业务费专项资金项目(01760516)资助
【分类号】:X55;X834

【相似文献】

相关期刊论文 前6条

1 任福民,郭艳君,周琴芳,项静恬;小波变换及其对厄尔尼诺研究的初步应用[J];数理统计与管理;1998年03期

2 江明顺,张爱军;用小波变换研究赤道西太平洋温度细结构[J];海洋学报(中文版);1998年06期

3 嵇春艳;陈明璐;;基于小波变换与神经网络的海洋平台结构损伤识别与定位研究[J];海洋技术;2009年04期

4 张乐;马玉祥;董国海;;基于小波变换的海浪方向谱估计方法研究[J];水运工程;2013年09期

5 黄玉新;裴玉国;张宁川;;基于小波变换的畸形波生成过程时频特性研究[J];水动力学研究与进展A辑;2009年06期

6 李炎保,郄禄文,祝振宇;小波变换在随机海浪及相关课题中的应用与前景[J];力学进展;2003年04期

相关会议论文 前1条

1 朱艳;鲁辛凯;;基于时频分析的海底沉积物类型分类技术研究[A];第13届中国系统仿真技术及其应用学术年会论文集[C];2011年

相关硕士学位论文 前1条

1 田鑫;基于离散小波变换的海洋工程数据降噪研究[D];大连理工大学;2015年



本文编号:1569360

资料下载
论文发表

本文链接:https://www.wllwen.com/shengtaihuanjingbaohulunwen/1569360.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户6f182***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com