基于卷积神经网络的酱油风味物质数据监测研究-《中国调味品》

文章信息/Info

Title:: Research on ata Monitoring of Soy Sauce Flavor Based on Convolution Neural Network

摘要:: 本文以酱油品质鉴别过程为研究对象，通过对酱油品质相关特性进行分析，利用味觉传感器系统采集酱油品质数据，获取后3次采集的品质数据平均值进行分析。以卷积神经网络为基础，搭建酱油品质数据的识别方式，对酱油品质数据的相对标准偏差以及标准偏差进行识别，并确定酱油品质等级。实验表明卷积神经网络数据识别，可有效的用于酱油品质鉴别过程。

Abstract:: This paper takes the process of soy sauce quality identification as the research object, through the analysis of the relevant characteristics of soy sauce quality, adopts the taste sensor system to collect the soy sauce quality data, and obtains the average value of the quality data collected three times later for analysis. On the basis of convolution neural network, the identification method of soy sauce quality data is established, the relative standard deviation and standard deviation of soy sauce quality data are identified, and the quality grade of soy sauce is determined. Through the experimental verification, the surface convolution neural network data recognition can be effectively used in the soy sauce quality identification process