电子鼻结合 PCA-BP 神经网络在烟用香料质量控制中的应用

摘要

In this paper,electronic nose was applied in identifying of different batches of flavors.Principal component analysis (PCA)and BP artificial neural network algorithm are used to study the tobacco flavors.The sense character prediction model based on PCA-BP neural network is established to forecast the different kinds of tobacco flavors.The result shows that,after learning,the PCA-BP model has the ability to predict the character of flavors and fragrances accurately.So it is proved effective to use the PCA-BP neural network model in quality control of tobacco flavors and fragrances .%本文通过电子鼻对于不同批次的烟用料液进行分析，应用主成分分析（PCA）和误差反向传播（BP）人工神经网络分析烟用料液，并建立了基于PCA-BP神经网络的烟用香精的质量控制模型。结果表明，PCA-BP神经网络模型经过学习后，能够准确的实现烟用香精香料的质量控制。