柿饼干燥中单宁变化规律及BP神经网络预测模型的建立-《中国调味品》

文章信息/Info

Title:: The Change of Tannin in Persimmon Drying and the Establishment of BP Neural Network Prediction Model CHEN Chao-feng, WANG Heng-hu,HUANG zi-zhen,TENG Jan-wen*

摘要:: 研究不同温度、不同初始单宁含量、不同水分含量对柿饼干制过程中可溶性单宁变化规律的影响，并建立BP神经网络预测模型。结果表明，在35-55℃范围内温度越高可溶性单宁脱涩时间越短，且每个温度下均出现返涩现象；初始单宁含量越高脱涩时间越长，但初始单宁含量在低浓度范围内，脱涩时间不受单宁浓度的影响；水分含量影响脱涩速率，水分含量越低，脱涩越困难。通过建立的BP神经模型可知，BP网络结构为4-6-1，BP预测模型的相关系数为0.966，验证集模型的相关系数为0.93，证明BP神经网络可以对干燥过程中的可溶性单宁的含量进行预测。

Abstract:: The effects of different temperatures, different initial tannin content and different moisture content on the change of soluble tannin in the process of dried persimmon were studied, and a BP neural network prediction model was established. The results show that the higher the temperature in the range of 35-55 °C, the shorter the time of removaling the soluble tannin and r eturning astringencyat every temperature; the higher the initial tannin content, the longer the removaling the soluble tannin time, but the initial tannin content is in the low concentration range, the removaling time is not The effect of the concentration of tannin; the moisture content affects the rate of removaling, and the lower the moisture content, the more difficult the removaling. According to the established BP neural model, the BP network structure is 4-6-1, the correlation coefficient of the BP prediction model is 0.966, and the correlation coefficient of the verification set model is 0.93, which proves that the BP neural network can be used for the soluble tannin in the drying process. The content is predicted