首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于移动窗口相关系数光谱法的白酒品牌快速判别技术研究
引用本文:张正勇,沙敏,桂冬冬,叶晓菁,王海燕.基于移动窗口相关系数光谱法的白酒品牌快速判别技术研究[J].光谱学与光谱分析,2017,37(10):3122-3126.
作者姓名:张正勇  沙敏  桂冬冬  叶晓菁  王海燕
作者单位:1. 南京财经大学管理科学与工程学院,江苏 南京 210023
2. 浙江工商大学管理工程与电子商务学院,浙江 杭州 310018
3. 江苏省质量安全工程研究院,江苏 南京 210023
4. 西南交通大学数学学院,四川 成都 611756
基金项目:江苏省高校自然科学研究面上项目,国家自然科学基金项目
摘    要:为满足白酒品牌判别领域快速检测的迫切需求,提出了一种以紫外光谱为输入,移动窗口相关系数法为手段的白酒品牌快速判别新方法。研究以海之蓝酒作为对象品牌,实验结果显示各批次海之蓝酒在实验波长区间存在较高的吻合一致性,基于紫外光谱原始谱图的相关系数值和基于移动窗口光谱计算所得的相关系数值均达到0.99以上。由于其他洋河系列白酒与海之蓝酒均属于洋河酒厂生产,原料来源、生产制备工艺等存在较高的相似性,此时若仅使用紫外光谱原始谱图数据结合相似度算法进行品牌判别,则定量化的判别区分存在一定困难。移动窗口法可有效提高谱图细节差异,突出光谱在特征谱段的形状、强度、变化趋势差异,提高实验谱图在细节结构差异辨认分析能力。经过研究筛选,发现以移动窗口相关系数结合紫外光谱特征谱段(270~295 nm)可提高洋河系列酒间的谱图差异,实现这种有较高相似性白酒品牌的判别区分。此外,市售六种其他品牌白酒及乙醇作为对照组,进一步论证了该方案的可行性。该研究的创新之处在于提出了一种移动窗口相关系数光谱法进行白酒品牌快速判别分析的新思路,该方法同时具备速度快,操作简便等优点,具有较高的实际应用潜在价值,并可为其他食品的质量安全判别分析研究提供借鉴。

关 键 词:移动窗口  相似度  白酒  食品  快速判别  
收稿时间:2016-11-27

Research on the Rapid Discrimination Technology of the Chinese Liquor Brands Based on the Moving Window Correlation Coefficient Spectral Method
ZHANG Zheng-yong,SHA Min,GUI Dong-dong,YE Xiao-jing,WANG Hai-yan.Research on the Rapid Discrimination Technology of the Chinese Liquor Brands Based on the Moving Window Correlation Coefficient Spectral Method[J].Spectroscopy and Spectral Analysis,2017,37(10):3122-3126.
Authors:ZHANG Zheng-yong  SHA Min  GUI Dong-dong  YE Xiao-jing  WANG Hai-yan
Institution:1. School of Management Science and Engineering, Nanjing University of Finance and Economics, Nanjing 210023, China 2. School of Management Engineering and Electronic Commerce, Zhejiang Gongshang University, Hangzhou 310018, China 3. Jiangsu Institute of Quality and Safety Engineering, Nanjing 210023, China 4. School of Mathematics, Southwest Jiaotong University, Chengdu 611756, China
Abstract:A new strategy was proposed in this work to meet the urgent needs of the rapid discrimination technology of Chinese liquor brands.The ultraviolet spectra were used as the inputs,the moving window correlation coefficient was applied as the ana-lytical method,and the Haizhilan spirit was used as the research object.The results showed that the quality of different batches of Haizhilan spirit was consistent in the experimental wavelength interval.The correlation coefficient of ultraviolet spectral origi-nal data and the moving window correlation coefficient spectral were all above 0.99.The processing techniques and raw materials of Haizhilan spirit and the other series of Yanghe spirits belonging to the same company had a higher similarity.It was difficult to quantitatively discriminate their brands only by using the ultraviolet spectral original data with correlation coefficient method. The moving window method could effectively improve the spectral detail differences,highlight the shape,intensity and trend differences of the spectral feature bands,so it could enhance the ability of the experimental spectra to identify and analyze the fine structure.After an overall investigation,the moving window correlation coefficient combined with ultraviolet feature spectra (270~295 nm)improved the differences among Yanghe spirits,and then Haizhilan spirit could be discriminated from the other Yanghe spirits with high similarities.Six brands of Chinese liquor and ethanol as the control groups were used to further demon-strate the feasibility of this approach.The innovation of this work is to propose a novel analytical idea for rapid identification of the Chinese liquor brands based on the moving window correlation coefficient spectral method.This method also has the advanta-ges of rapidity,simplicity,and has a great potential in the quality and safety discrimination of other foods.
Keywords:Moving window  Similarity  Chinese liquor  Food  Rapid discrimination
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《光谱学与光谱分析》浏览原始摘要信息
点击此处可从《光谱学与光谱分析》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号