摘要
文章通过对酱油挥发性风味物质成分进行检测分析,筛选出78组酱油共有挥发性成分,并获取与酱油发酵方式和生产地区相关的主要特征成分,利用神经网络进行特征成分的分类,鉴别酱油发酵方式及生产地区。利用遗传算法对神经网络分类模型进行优化,提高分类模型的鉴别性能。鉴别分析结果表明,使用原始数据进行酱油发酵方式分类和酱油产地分类的识别准确率可达到80%,使用优化后参数进行酱油发酵方式分类和酱油产地分类的识别准确率可达到100%。
In this paper,through the detection and analysis of the volatile flavor components of soy sauce,78 groups of soy sauce are screened out,and the main characteristic components related to soy sauce fermentation methods and production areas are obtained,and the neural network is used to classify the characteristic components to identify soy sauce fermentation methods and production areas.Genetic algorithm is used to optimize the neural network classification model,improve the identification performance of classification model.The discriminant analysis results show that the identification accuracy of soy sauce fermentation method classification and soy sauce production area classification using the original data could reach 80%,and the identification accuracy of soy sauce fermentation method classification and soy sauce production area classification using the optimized parameters could reach 100%.
作者
蔡炳育
杨帆
CAI Bing-yu;YANG Fan(Suzhou Industrial Park Institute of Services Outsourcing,Suzhou 215123,China;Guizhou University of Finance and Economics,Guiyang 550025,China)
出处
《中国调味品》
CAS
北大核心
2022年第1期180-182,共3页
China Condiment
基金
首批江苏省省级职业教育教师教学创新团队“优秀教学团队”(JSJXCX23)。
关键词
酱油分类模型
神经网络
遗传算法
挥发性物质
soy sauce classification model
neural network
genetic algorithm
volatile substances
