摘要
采用气相色谱-离子迁移谱(GC-IMS)对4种风味咖啡的挥发性有机物(VOCs)进行快速分析。根据GC-IMS谱图中色谱保留时间和离子迁移谱漂移时间对其进行二维分离,采用随机森林方法对不同风味咖啡进行分类识别并筛选重要风味特征峰,用支持向量机验证特征筛选方法的有效性。结果表明,GC-IMS可有效区分咖啡风味,随机森林模型分类准确率高达94%,并筛选出与咖啡风味相关的6个主要特征峰。
Based on the GC retention time and the IMS drift time of GC-IMS spectrum,the VOCs of four kinds of different coffee were separated two-dimensionally.Random forest method was used to classify coffee with different flavors by characteristic matrix extracted from spectrum,and the primary flavor characteristic peaks were screened.Support vector machine was used to verify the performance of random forest model.The results showed that GC-IMS could effectively distinguish coffee flavor,the accuracy rate of random forest classification model was as high as 94%,and 6 dominating flavor characteristic peaks were screened out.
作者
郭秀丽
高晓光
贾建
何秀丽
Guo Xiuli;Gao Xiaoguang;Jia Jian;He Xiuli(State Key Laboratory of Transducer Technology,Aerospace Information Research Institute,Chinese Academy of Sciences,Beijing 100190,China;School of Electronic,Electrical and Communication Engineering,University of Chinese Academy of Sciences,Beijing 100049,China)
出处
《分析仪器》
CAS
2021年第6期16-21,共6页
Analytical Instrumentation
基金
国家自然科学基金资助项目(No.61871364)。
关键词
气相色谱-离子迁移谱
咖啡风味
挥发性有机物
随机森林
支持向量机
Gas chromatography-ion mobility spectroscopy
Coffee flavor
Volatile organic compounds
Random forest
Support vector machine
