CASIA OpenIR  > 毕业生  > 硕士学位论文
基于红外光谱的白酒质量分析算法研究
其他题名Research on Liquor Quality Analysis Algorithms Based on Infrared Spectrocopy
潘磊
学位类型工程硕士
导师彭思龙
2013-07-26
学位授予单位中国科学院大学
学位授予地点中国科学院自动化研究所
学位专业计算机技术
关键词红外光谱 酱香型白酒 理化分析 感官质量分析 偏最小二乘 Infrared Spectrum Maotai-flavor Liquor Physicochemical Analysis Sensory Quality Analysis Partial Least Squares (Pls)
摘要白酒的质量评价是我国白酒行业中一项非常重要的生产环节,是白酒勾兑和调味的基础,是白酒质量安全检验和判断产品质量优劣的重要依据,它通常通过理化分析和感官质量分析来实现。目前,白酒的理化分析主要采用色谱分析来检测其中的微量香味物质成分,但检测出所有组分的绝对含量在色谱分离技术上还未能实现,同时色谱分析必须确保在样品制备过程中待测组分不发生任何损失,且不能从整体上对白酒进行分析。白酒的感官质量分析则主要依靠人的感觉器官作为检测载体和手段来判别酒的色、香、味、格,然而其品酒人员的技术水平和经验积累等综合素质的差异,以及品酒人员生理和心理因素以及环境的影响,都会导致品评结果的偏差和不稳定。傅里叶变换红外光谱技术具有快速、整体和无损鉴定复杂混合物体系等优点,可以更为有效地检测白酒中主要的香味物质成分。在此背景下,将傅里叶红外光谱技术应用于酱香型白酒的质量评价,分别进行理化分析和感官质量分析的研究,就目前来说是一种新的尝试和有益的探索。 论文的主要内容包括: 1、研究了白酒红外光谱与色谱浓度参考值之间的关系,建立了以红外光谱吸光度为自变量,色谱参考值为因变量的偏最小二乘回归模型,分析了红外光谱用于白酒微量成分分析的可行性。通过实验证明,与常用的色谱分析技术相比,红外光谱技术在白酒香味物质的检测方面表现也同样良好。 2、研究了红外光谱进行白酒感官质量分析的可行性,分别从定量和定性两个角度对基酒感官质量进行了分析。首先,通过整理酱香型白酒的品评要点和品酒师的品评经验知识,确定评价项目,分配量值,设计了一种针对某轮次基酒的感官品评标准,并组织品酒人员实施该基酒的实际品评实验。其次,根据感官质量分析的特点和红外光谱的性质,分别对感官品评标准的各品评项进行偏最小二乘回归分析,探讨了红外光谱进行定量分析的可行性。最后,将感官质量品评数据先进行分类处理,再选择线性判别分析和支持向量机两种方法对基酒的红外光谱和感官质量聚类标签进行识别分析。实验结果显示,该基酒的某些特定品评项可用于定量分析,而感官质量的定性分析结果则与品酒师的品评能力表现相符。
其他摘要Liquor quality evaluation is a very important production link in the Chinese liquor industry, and is the foundation of liquor blending and seasoning, and is an important basis for liquor quality and safety inspection and judging the product quality. It is usually communicates via physicochemical analysis and sensory quality analysis. Currently, the liquor physicochemical analysis mainly adopts gas chromatographic (GC) analysis to detect trace flavoring substances, however, detecting the absolute content of all the components is failed to achieve by GC separation technology, as well as GC analysis must ensure that the logging component without any loss in the process of sample preparation, and cannot be overall analysis of the whole substance in liquor. The sensory quality analysis is largely depend on the sensory organ as a detection carrier and means to determine the wine color, fragrance, taste, personality, but the differences of overall quality of the staff, such as their tasting technology level and the accumulation of experience, as well as the impact of physiological and psychological factors and the environment will lead to inaccurate and instability of the evaluation results. Fourier transform infrared (FTIR) spectroscopy technology because of its fast, undamaged and overall identification of complex system and other advantages, can effectively detect main aroma components in liquor. In this context, the study about the quality evaluation of Maotai-flavor liquor by FTIR spectroscopy, focusing on physicochemical analysis and sensory quality analysis, is a new attempt and useful exploration. The main contents include: Firstly, we study the relationship between the liquor infrared spectroscopy and chromatography concentration reference value, a partial least squares (PLS) regression model between IR absorbance and GC reference values is established to analyze the feasibility whether IR can be used for liquor trace component analysis. Experimental results demonstrate that, compared with common GS technology, IR used for the detection of liquor flavoring substance also has a good performance. Secondly, we study on the feasibility about IR used for liquor sensory quality analysis, and analysis the base wine sensory from the perspective of quantitative and qualitative analysis, respectively. First of all, through the point of evaluation order of Maotai-flavor liquor and tasting experience and knowledge of staff, we determine the evaluation item, an...
馆藏号XWLW1951
其他标识符2009M8014629011
语种中文
文献类型学位论文
条目标识符http://ir.ia.ac.cn/handle/173211/7699
专题毕业生_硕士学位论文
推荐引用方式
GB/T 7714
潘磊. 基于红外光谱的白酒质量分析算法研究[D]. 中国科学院自动化研究所. 中国科学院大学,2013.
条目包含的文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
CASIA_2009M801462901(1179KB) 暂不开放CC BY-NC-SA请求全文
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[潘磊]的文章
百度学术
百度学术中相似的文章
[潘磊]的文章
必应学术
必应学术中相似的文章
[潘磊]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。