光谱学与光谱分析
光譜學與光譜分析
광보학여광보분석
SPECTROSCOPY AND SPECTRAL ANALYSIS
2014年
2期
350-353
,共4页
近红外光谱学%西红柿%总酸%可溶性糖%偏最小二乘法(PLS )
近紅外光譜學%西紅柿%總痠%可溶性糖%偏最小二乘法(PLS )
근홍외광보학%서홍시%총산%가용성당%편최소이승법(PLS )
Near infared spectroscopy%Tomato%Total acid%Soluble sugar%Partial least square method
为了探索一种简捷、快速、高效的西红柿品质检测方法，应用近红外光谱技术与光纤传感技术相结合的新方法，快速测量西红柿果浆样品中营养成分的含量。实验所用的主要仪器为近红外光纤光谱仪，波长范围为900～2500 nm。以164个西红柿样品为标准样品，进行了光谱采集及相应的化学值测定。实验数据采用偏最小二乘法（PLS ）进行回归，建立西红柿果浆中总酸及可溶性糖含量的数学模型，并对回归方法进行统计分析。结果为：西红柿果浆中总酸验证集的决定系数（R2）为0.967，均方根误差（RMSEC）为0.133，预测均方根误差（RMSEP）为0.103；总糖验证集的决定系数（R2）为0.976，均方根误差（RMSEC）为0.463，预测均方根误差（RMSEP）为0.460。均达到了较好的预测结果，表明该方法对定量分析西红柿果浆中多组分含量是可行的。基于该方法快速、简便及可对同一样品多组分含量同时分析的优点，它是一种极具发展前途的传感器，正在逐渐成为国际传感器领域的研究热点。
為瞭探索一種簡捷、快速、高效的西紅柿品質檢測方法，應用近紅外光譜技術與光纖傳感技術相結閤的新方法，快速測量西紅柿果漿樣品中營養成分的含量。實驗所用的主要儀器為近紅外光纖光譜儀，波長範圍為900～2500 nm。以164箇西紅柿樣品為標準樣品，進行瞭光譜採集及相應的化學值測定。實驗數據採用偏最小二乘法（PLS ）進行迴歸，建立西紅柿果漿中總痠及可溶性糖含量的數學模型，併對迴歸方法進行統計分析。結果為：西紅柿果漿中總痠驗證集的決定繫數（R2）為0.967，均方根誤差（RMSEC）為0.133，預測均方根誤差（RMSEP）為0.103；總糖驗證集的決定繫數（R2）為0.976，均方根誤差（RMSEC）為0.463，預測均方根誤差（RMSEP）為0.460。均達到瞭較好的預測結果，錶明該方法對定量分析西紅柿果漿中多組分含量是可行的。基于該方法快速、簡便及可對同一樣品多組分含量同時分析的優點，它是一種極具髮展前途的傳感器，正在逐漸成為國際傳感器領域的研究熱點。
위료탐색일충간첩、쾌속、고효적서홍시품질검측방법，응용근홍외광보기술여광섬전감기술상결합적신방법，쾌속측량서홍시과장양품중영양성분적함량。실험소용적주요의기위근홍외광섬광보의，파장범위위900～2500 nm。이164개서홍시양품위표준양품，진행료광보채집급상응적화학치측정。실험수거채용편최소이승법（PLS ）진행회귀，건립서홍시과장중총산급가용성당함량적수학모형，병대회귀방법진행통계분석。결과위：서홍시과장중총산험증집적결정계수（R2）위0.967，균방근오차（RMSEC）위0.133，예측균방근오차（RMSEP）위0.103；총당험증집적결정계수（R2）위0.976，균방근오차（RMSEC）위0.463，예측균방근오차（RMSEP）위0.460。균체도료교호적예측결과，표명해방법대정량분석서홍시과장중다조분함량시가행적。기우해방법쾌속、간편급가대동일양품다조분함량동시분석적우점，타시일충겁구발전전도적전감기，정재축점성위국제전감기영역적연구열점。
In order to explore a simple ,rapid and efficient tomato quality detection method ,in the present experiment near infra-red spectroscopy and optical fiber sensing technology were applied to quickly measure the nutrition ingredient content in tomato juice samples .The main instrument used in this experiment was near infrared optical fiber spectrometer in a wavelength range from 900 to 2 500 nm ,which measured the absorbance of the tomato juice samples ;A collection of one hundred and sixty-four tomato juice samples were selected as the standard samples ,the spectra and the corresponding chemical value were measured . Partial least squares (PLS) was adopted to establish the mathematical model of the total acid and soluble sugar content in tomato juice samples ,and the regression equation was statistically analysed .The total acid in tomato juice prediction correlation coeffi-cient was 0.967 ,calibration standard deviation (RMSEC) was 0.133 ,standard error of prediction (RMSEP) was 0.103;the soluble sugar prediction correlation coefficient is 0.976 ,calibration standard deviation (RMSEC) was 0.463 ,and the standard error of prediction (RMSEP) was 0.460 .The above data achieved better forecasting results ,which showed that the method of quantitative analysis of tomato fruit multicomponent content was feasible .The method is rapid ,simple and can do multicompo-nent analysis on the same sample simultaneously .It is a promising sensor and gradually becoming a international research focus in sensor field .
