Detection of key performance indicators of ethanol diesel by the infrared spectroscopy method
-
摘要:乙醇柴油是柴油替代品的一种，它的使用越来越广泛，乙醇柴油品质由许多指标决定，采用传统方法检测这些指标不仅价格昂贵而且耗时长。近红外光谱技术是一种廉价、快速实时在线检测乙醇柴油品质的有效方法。本文采用近红外光谱技术结合最小二乘支持向量机检测了乙醇柴油的密度、粘度和乙醇含量，比较了线性和非线性校正技术（主成分回归、偏最小二乘回归和最小二乘支持向量机）对乙醇柴油品质的分析效果，同时也比较了不同预处理方法对预测模型能力的影响。实验结果表明，最小二乘支持向量机优于主成分回归和偏最小二乘回归模型，其对乙醇柴油密度、粘度、乙醇含量建模效果最优，相关系数分别是0.995 8、0.995 7和0.995 3；预测均方根误差分别为0.000 68、0.011 3和0.571 4。Abstract:Ethanol diesel is one of the alternatives for the diesel, which is used more and more widely. Many indicators reflect the quality of diesel ethanol fuel. It is not only expensive but also time consuming to detect these indicators with traditional method. Near infrared spectroscopy method is an inexpensive, fast and real time online test for the quality of ethanol diesel oil. In this paper, the density, viscosity and the quality of the ethanol content of diesel are detected, and the analysis effect on the quality of ethanol diesel are compared by linear and non-linear calibration technology, including principal component regression, partial least squares regression and least squares support vector machines (LSSVM), and the effects of different pretreatment methods on the prediction model capabilities are also compared. Experimental results show that LSSVM is better than the principal component regression and partial least squares regression model, with the optimal modeling effect on the density of ethanol-diesel, viscosity, alcohol content. The correlation coefficients are 0.995 8, 0.995 7 and 0.995 3, and the root mean square error of prediction are 0.000 68, 0.011 3 and 0.571 4, respectively.
-
Key words:
- NIRS/
- ethanol diesel/
- density/
- viscosity/
- ethanol content
-
图 2乙醇柴油密度、粘度和乙醇含量在PCR, PLS, LSSVM建模下的RMSEP
Figure 2.RMSEP of ethanol diesel oil density, viscosity and ethanol content under PCR, PLS and LSSVM modeling RMSEP
表 132种浓度乙醇柴油
Table 1.Ethanol diesel with 32 kinds of concentration (%,v/v)
序号 浓度/% 序号 浓度/% 序号 浓度/% 序号 浓度/% 1 0.6 9 5.6 17 10.6 25 15.6 2 1.2 10 6.2 18 11.2 26 16.2 3 1.8 11 6.8 19 11.8 27 16.8 4 2.4 12 7.4 20 12.4 28 17.4 5 3.0 13 8.0 21 13.0 29 18.0 6 3.6 14 8.6 22 13.6 30 18.6 7 4.2 15 9.2 23 14.2 31 19.2 8 5.0 16 10.0 24 15.0 32 20.0 
下载:
导出CSV
表 2PCR方法对乙醇柴油密度、粘度、乙醇含量的性能预测
Table 2.Prediction results of principal component regression(PCR) approach for ethanol diesel properties including:density, viscosity, ethanol content
乙醇柴油性质 最优预处理方法 主成分PC R2 RMSEP 密度/(g·cm-3) MSC-SG 16 0.909 4 0.001 7 粘度/(mPa·s) MSC-SG 16 0.902 4 0.038 6 乙醇含量(体积比)(%) MSC-SG 14 0.883 0 1.968 7 下载:
导出CSV
表 3PLS方法对乙醇柴油密度、粘度、乙醇含量的性能预测
Table 3.Prediction results of partial least squares (PLS) regression approach for ethanol diesel properties including:density, viscosity, ethanol content
乙醇柴油性质 最优预处理方法 主因子数 R2 RMSEP 密度/(g·cm-3) SG 8 0.955 3 0.001 5 粘度/(mPa·s) SG 8 0.951 5 0.027 2 乙醇含量(体积比)(%) SG 8 0.951 0 1.273 7 下载:
导出CSV
表 4LSSVM对乙醇柴油密度、粘度、乙醇含量的性能预测
Table 4.Prediction results of least squares support vector machine (LSSVM) regression approach for ethanol diesel properties including:density, viscosity, ethanol content
乙醇柴油性质 最优预处理方法 [gamma, sig2] R2 RMSEP 密度/(g·cm-3) SGD1-SG 131 969, 2 047 0.995 8 0.000 68 粘度/(mPa·s) SNV-SG 111 607, 5 690 0.995 7 0.011 3 乙醇含量(体积比)(%) SG 28 707, 5 963 0.995 3 0.571 4 下载:
导出CSV
-
[1] BAPTISTA P, FELIZARDO P, MENEZES J C,et al.. Multivariate near infrared spectroscopy models for predicting the methyl esters content in biodiesel[J].Anal. Chim. Acta, 2008, 607(2):153-159.doi:10.1016/j.aca.2007.11.044 [2] 张明祥, 闵顺耕, 李宁, 等.乙醇混合燃料近红外定量分析[J].分析测试学报, 2003, 6:15-18.doi:10.3969/j.issn.1004-4957.2003.01.005ZHANG M X, MIN S G, LI N,et al.. Ethanol blends near-infrared quantitative analysis[J].Instrumental Analysis, 2003, 6:15-18.(in Chinese)doi:10.3969/j.issn.1004-4957.2003.01.005 [3] 蔡剑华, 王先春, 胡惟文.基于形态小波的烟草尼古丁含量近红外光谱检测[J].农业工程学报, 2012(15):281-286.http://www.cnki.com.cn/Article/CJFDTOTAL-NYGU201215048.htmCAI J H, W X CH, HU W W. Near Infrared Spectroscopy based on morphological wavelet nicotine content of tobacco[J].J. Agricultural Engineering, 2012(15):281-286.(in Chinese)http://www.cnki.com.cn/Article/CJFDTOTAL-NYGU201215048.htm [4] 孙通, 应义斌, 刘魁武, 等.梨可溶性固形物含量的在线近红外光谱检测[J].光谱学与光谱分析, 2008(11):2536-2539.doi:10.3964/j.issn.1000-0593(2008)11-2536-04SUN T, YING Y B, LIU K W,et al.. Pear-line Near Infrared Spectroscopy of soluble solids[J].Spectroscopy and Spectral Analysis, 2008(11):2536-2539.(in Chinese)doi:10.3964/j.issn.1000-0593(2008)11-2536-04 [5] 孙旭东, 郝勇, 高荣杰, 等.脐橙糖度近红外光谱在线检测数学模型优化研究[J].光谱学与光谱分析, 2011, 31(5):1230-1235.http://www.cnki.com.cn/Article/CJFDTOTAL-GUAN201105022.htmSUN X D, HAO Y, GAO R J. Orange sugar content line Near Infrared spectroscopy detection mathematical model optimization[J].Spectroscopy and Spectral Analysis, 2011, 31(5):1230-1235.(in Chinese)http://www.cnki.com.cn/Article/CJFDTOTAL-GUAN201105022.htm [6] 史永刚, 粟斌, 李华峰, 等.拉曼光谱及其在石油产品分析中的应用[J].现代仪器, 2010, 16(9106):9-13.http://www.cnki.com.cn/Article/CJFDTOTAL-XDYI201006004.htmSHI Y G, SU B, LI H F,et al.. Merlin Raman spectroscopy and its application in petroleum product analysis[J].Modern Instruments, 2010, 16(9106):9-13.(in Chinese)http://www.cnki.com.cn/Article/CJFDTOTAL-XDYI201006004.htm [7] 李津蓉, 俞柯.一种基于拉曼光谱技术的甲醇汽油定量分析方法[J].科技创新与应用, 2016(15305):59.http://www.cnki.com.cn/Article/CJFDTOTAL-CXYY201605046.htmLI J R, YU K. Methanol gasoline quantitative analysis method based on Raman spectroscopy[J].Science and Technology Innovation and Application, 2016(15305):59.(in Chinese)http://www.cnki.com.cn/Article/CJFDTOTAL-CXYY201605046.htm [8] 欧阳爱国, 刘军.基于近红外与拉曼光谱的甲醇汽油定量分析[J].西南师范大学学报(自然科学版), 2012, 37(18609):98-101.http://www.cnki.com.cn/Article/CJFDTOTAL-XNZK201209019.htmOUYANG A G, LIU J. Methanol-based gasoline Quantitative Near Infrared and Raman Spectra of[J].Southwest China Normal University(Natural Science), 2012, 37(18609):98-101.(in Chinese)http://www.cnki.com.cn/Article/CJFDTOTAL-XNZK201209019.htm [9] 淡图南, 戴连奎.近红外光谱的甲醇汽油定量分析[J].计算机与应用化学, 2011, 3:329-332.doi:10.3969/j.issn.1001-4160.2011.03.019DAN T N, DAI L K. Quantitative methanol gasoline near infrared spectroscopy[J].Computer and Applied Chemistry, 2011, 3:329-332.(in Chinese)doi:10.3969/j.issn.1001-4160.2011.03.019 [10] 姚捷, 戴连奎, 林艺玲.基于拉曼特征峰的甲醇汽油甲醇含量测定[J].光散射学报, 2013, 1:59-65.doi:10.3969/j.issn.1004-5929.2013.01.011YAO J, DAI L K, LIN Y L. Determination of methanol content of methanol-based gasoline Raman characteristic peak[J].J. of Light Scattering, 2013, 1:59-65.(in Chinese)doi:10.3969/j.issn.1004-5929.2013.01.011 [11] BALABIN R M, SAFIEVA R Z, LOMAKINA E I.Chemometr Intell Lab, 2007:88:183. [12] 刘燕德, 周延睿, 潘圆媛.基于最小二乘支持向量机的辣椒可溶性固形物和维生素C含量近红外光谱检测[J].光学精密工程, 2014(2):281-288.http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201402006.htmLIU Y D, ZHOU Y R, PAN Y Y. Based on least squares support chili soluble solids and vitamin C content of vector machine near infrared spectroscopy[J].Opt. Precision Eng., 2014(2):281-288.(in Chinese)http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201402006.htm [13] 刘燕德, 周延睿.基于GA-LSSVM的苹果糖度近红外光谱检测[J].西北农林科技大学学报(自然科学版), 2013, 7:229-234.http://www.cnki.com.cn/Article/CJFDTOTAL-XBNY201307038.htmLIU Y D, ZHOU Y R. Near infrared spectroscopy based GA-LSSVM apple brix[J].J. Northwest Sci-Tech University of Agriculture and Forestry(Natural Science Edition), 2013, 7:229-234.(in Chinese)http://www.cnki.com.cn/Article/CJFDTOTAL-XBNY201307038.htm [14] 彭秀辉, 黄常毅, 刘飞, 等.LSSVM模型下的近红外光谱联合区间波长筛选方法[J].光谱学与光谱分析, 2014, 3:668-672.doi:10.3964/j.issn.1000-0593(2014)03-0668-05PENG X H, HUANG C Y, LIU F,et al.. Near infrared spectroscopy synergy interval wavelength selection method using the LSSVM model[J].Spectroscopy and Spectral Analysis, 2014, 3:668-672.(in Chinese)doi:10.3964/j.issn.1000-0593(2014)03-0668-05 [15] 伍铁斌, 朱红求, 孙备, 等.PLS-LSSVM模型在锌净化中的应用[J].计算机工程, 2012, 10:212-214.doi:10.3969/j.issn.1000-3428.2012.13.063WU T B, ZHU H Q, SUN B,et al.. Application of PLS-LSSVM model in zinc purification[J].Computer Engineering, 2012, 10:212-214.(in Chinese)doi:10.3969/j.issn.1000-3428.2012.13.063 [16] 刘燕德, 邓清.基于高光谱成像技术的脐橙叶片的叶绿素含量及其分布测量[J].发光学报, 2015, 8:957-961.http://www.cnki.com.cn/Article/CJFDTOTAL-FGXB201508021.htmLIU Y D, DENG Q. Measurement of chlorophyll distribution in navel orange leaves based on hyper-spectral imaging technique[J].J. Luminescence, 2015, 8:957-961.(in Chinese)http://www.cnki.com.cn/Article/CJFDTOTAL-FGXB201508021.htm [17] 鲍一丹, 陈纳, 何勇, 等.近红外高光谱成像技术快速鉴别国产咖啡豆品种[J].光学精密工程, 2015, 2:349-355.http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201502005.htmBAO Y D, CHEN N, HE Y,et al.. Near infrared hyper-spectral imaging technology for rapid identification of domestic coffee bean varieties[J].Opt. Precision Eng., 2015, 2:349-355.(in Chinese)http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201502005.htm [18] 郭文川, 董金磊.高光谱成像结合人工神经网络无损检测桃的硬度[J].光学精密工程, 2015, 6:1530-1537.http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201506005.htmGUO W CH, DONG J L. High spectral imaging combined with artificial neural network for nondestructive detection of peach hardness[J].Opt. Precision Eng., 2015, 6:1530-1537.(in Chinese)http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201506005.htm [19] YANG H, GRIFFITHS P R, TATE J D. Comparison of partial least squares regression and multi-layer neural networks for quantification of nonlinear systems and application to gas phase Fourier transform infrared spectra[J].Anal Chim Acta, 2003, 489:125-136.doi:10.1016/S0003-2670(03)00726-8 -
下载:
点击查看大图
图(3)/
表(4)
计量
- 文章访问数:1682
- HTML全文浏览量:642
- PDF下载量:616
- 被引次数:0
