毕业论文

打赏
当前位置: 毕业论文 > 食品科学 >

基于近红外光谱的黄桃脆片品质检测

时间:2018-09-20 18:05来源:毕业论文
优化水果脆片分级技术,并对今后水果脆片的品质检测提供理论基础。本文选取600个黄桃脆片样品作为研究对象,结合近红外光谱技术,分别于900-2500 nm和1000-2500 nm两种波段下进行光谱采

摘要:为了优化水果脆片分级技术,并对今后水果脆片的品质检测提供理论基础。本文选取600个黄桃脆片样品作为研究对象,结合近红外光谱技术,分别于900-2500 nm和1000-2500 nm两种波段下进行光谱采集,采用平滑、标准正态变量变换(SNV)、多元散射校正(MSC)、一阶导数和二阶导数五种预处理方法分别对原始光谱进行预处理之后,采用偏最小二乘法(PLS)、最小二乘支持向量机(LS-SVM)和极限学习机(ELM)三种方法进行模型建立,比较得出最佳预处理方法和模型建立方法。研究结果表明,在900-2500 nm的近红外光谱下,采用数据平滑对黄桃脆片样品的颜色(L*、a*、b*)和硬度进行预处理效果最好,RMSECV分别为2.8399、2.3333、3.8568、1.2776,采用MSC对黄桃脆片样品的可溶性固形物进行预处理效果最好,RMSECV为0.3648,预处理之后,采用LS-SVM进行建模效果最好,各项指标Rc和Rp分别为0.8484和0.8212、0.9052和0.8738、0.9380和0.9097、0.4872和0.3737、0.8719和0.5825;在1000-2500 nm的近红外光谱下,采用数据平滑对黄桃脆片样品的颜色(L*、a*、b*)、硬度和可溶性固形物进行预处理效果最好,RMSECV分别为2.9174、2.4662、4.5507、1.2804、0.3541,预处理之后,采用PLS进行建模效果最好,各项指标Rc和Rp分别为0.8483和0.8129、0.9060和0.8557、0.9196和0.8611、0.6472和0.4848、0.5057和0.2924。该研究对黄桃脆片各项品质指标的预测和分析提供技术支持。28404
毕业论文关键词:黄桃脆片;品质评价;近红外光谱;光谱预处理
Detection for Quality of Yellow Peach Chips Based on NIR Spectroscopy
Abstract:In order to optimize the classification of fruit chips and provide a theoretical basis for the quality testing of fruit chips in the future, in this work, 600 yellow peach chips samples were detected by near infrared spectroscopy with the band of 900-2500 nm and 1000-2500 nm, respectively, then use the smooth transformation, standard normal variate transformation (SNV), multiplicative scatter correction (MSC), 1st derivative and 2nd derivative to preprocess the spectra. After preprocessed, partial least squares (PLS), least squares support vector machines (LS-SVM) and extreme learning machine (ELM) were used to build regression models. The research results showed that in the range of 900-2500 nm, the smooth transformation was the best preprocessing method to predict color and firmness of yellow peach chips, with RMSECV of L*, a*, b* and firmness were 2.8399, 2.3333, 3.8568, 2.8399. And the MSC was the best method to predict soluble solids contents with RMSECV of 0.3648. After preprocessed, the LS-SVM were best in all models, for the L*, a*, b*, firmness and soluble solids contents, the Rc and Rp were 0.8484 and 0.8212, 0.8212 and 0.9052, 0.8212 and 0.9097, 0.4872 and 0.3737, 0.4872 and 0.4872, respectively; In the range of 1000-2500 nm, the smooth transformation was the best preprocessing method to predict color, firmness and SSC of yellow peach chips with RMSECV of 2.9174, 2.4662, 4.5507, 1.2804, 0.3541, respectively. After preprocessed, the PLS was the best method, with Rc and Rp were 0.8483 and 0.8483, 0.9060 and 0.8557, 0.9196 and 0.8611, 0.6472 and 0.4848, 0.5057 and 0.2924, respectively. All in all, the study can provide technical support to predict the quality indexes of the yellow peach chips. 源¥自%六:维;论-文'网=www.lwfree.cn
Key words: peach chips; quality evaluation; near-infrared spectroscopy; spectroscopy data pretreatment
目  录
摘要1
关键词1
Abstract1
Key words2
引言2
1 材料与方法2
1.1  仪器和设备2
1.2  试验材料3
1.3  试验流程3
1.3.1  颜色测定3
1.3.2  近红外光谱采集3
1.3.3  硬度测定4
1.3.4  近红外光谱采集4
1.3.5  水分的测定4
1.3.6  可溶性固形物的测定4
1.4  数据处理方法4 基于近红外光谱的黄桃脆片品质检测:http://www.lwfree.cn/shiping/20180920/23231.html
------分隔线----------------------------
推荐内容