基于电子鼻技术的黄山毛峰茶品质检测方法

为了更加客观地评价黄山毛峰茶的品质,提出了一种利用电子鼻技术对黄山毛峰茶品质检测的方法。选择4种不同品质等级的茶叶,首先根据传感器响应选择特征变量,然后以这些特征变量作为BP神经网络的输入,建立茶叶品质等级的3层网络预测模型。实验结果表明,本文建立的模型对于训练样本识别准确率为100%,对测试样本识别准确率为89.3%,表明应用电子鼻技术检测黄山毛峰茶品质具有可行性。

A Quality Detection Method for Huangshanmaofeng Tea Using Electronic Nose Technology

In order to objectively evaluate the quality of Huangshanmaofeng tea,aquality detection method for Huangshanmaofeng tea using electronic nose technology is proposed.The feature variables regarding four different grades are first selected according to the responses of sensors.The obtained feature variables are input into a BP neural network and a predictive model with the three-layer BP neural network is built to assess the quality grade of tea.Experimental results show that the recognition accuracy of the training samples is 100%and that of the test samples is reached by 89.3%,which indicates that the proposed method using electronic nose for Huangshanmaofeng tea quality detection is feasible.