本文選題：近紅外光譜 + 食用醋　； 參考：《光譜學(xué)與光譜分析》2014年09期

【摘要】：以四種品牌152組食用醋樣品為研究對(duì)象,采用漫反射與透射兩種近紅外光譜采集模式分別進(jìn)行光譜數(shù)據(jù)采集,并以此建立了食用醋品牌溯源模型,重點(diǎn)考察光譜采集模式、光譜預(yù)處理方法等對(duì)溯源模型精度的影響。結(jié)果表明,選取114組樣品為訓(xùn)練集,原始光譜數(shù)據(jù)經(jīng)過多元散射校正、二階求導(dǎo)預(yù)處理后,采用偏最小二乘判別分析法(PLS1-DA)建立的食用醋NIRS品牌溯源模型,對(duì)38組測(cè)試集樣品進(jìn)行預(yù)測(cè),透射光譜模型的決定系數(shù)(R2)、校準(zhǔn)均方根誤差(root-mean-square error of calibration,RMSEC)、預(yù)測(cè)均方根誤差(root-mean-square error of prediction,RMSEP)分別為0.92,0.113,0.127,正確識(shí)別率為76.32%;漫反射光譜模型R2,RMSEC,RMSEP分別為0.97,0.102,0.119,正確識(shí)別率為86.84%。由此說明,近紅外光譜結(jié)合PLS1-DA可以用來建立食用醋品牌溯源模型,且漫反射光譜模型預(yù)測(cè)效果更好。
[Abstract]:Taking 152 groups of edible vinegar samples of four brands as the research object, the spectral data were collected by diffuse reflection and transmission near infrared spectrum, and the tracing model of edible vinegar brand was established, and the spectral acquisition model was investigated. The effect of spectral pretreatment methods on the accuracy of traceability model. The results showed that 114 groups of samples were selected as training set, the original spectral data was corrected by multiple scattering, the second order derivative pretreatment was used to establish the NIRS brand traceability model of edible vinegar by partial least squares discriminant analysis (PLS1-DA). The determination coefficient of the transmission spectrum model is R2N, the root-mean-square error of calibration RMSECs, the root-mean-square error of prediction RMSEPs are 0.92n0.1130.127and the correct recognition rate is 76.322.The diffuse reflectance spectral model R2RMSECEPs are 0.970.102n0.119, and the correct RMSEPs are 86.84kgs. The results show that the RMSEPs of the R2GMS model are 0.970.102a 0.19, and the RMSEPs of the RMSECs are 0.970.102a 0.1190.The root-mean-square error of prediction RMSEPs are 0.921030.127.The RMSEP of the diffuse reflectance spectrum model is 0.970.102n0.119. It is concluded that the near infrared spectroscopy combined with PLS1-DA can be used to establish the traceability model of edible vinegar brand, and the diffuse reflectance spectrum model is more effective.
【作者單位】： 光明乳業(yè)股份有限公司乳業(yè)生物技術(shù)國(guó)家重點(diǎn)實(shí)驗(yàn)室;上海理工大學(xué)醫(yī)療器械與食品學(xué)院;上海海事大學(xué)信息工程學(xué)院;上海市食品藥品檢驗(yàn)所;
【基金】：國(guó)家自然科學(xué)基金項(xiàng)目(31101348)資助
【分類號(hào)】：O657.33

【相似文獻(xiàn)】

相關(guān)期刊論文 前2條

1 方瑞斌,王建,邱玲,王亮,李蕾;碳素基體固相微萃取氣相色譜、氣質(zhì)聯(lián)用分析食用醋中的有機(jī)揮發(fā)物[J];化學(xué)研究與應(yīng)用;1999年05期

2 楊平石;朱凱歌;高靜;;離子選擇電極法測(cè)定食醋中的微量鉛[J];化學(xué)傳感器;1992年03期

，

本文編號(hào)：1815572