论文部分内容阅读
中药气味的传统嗅觉鉴别往往受到个人鉴别经验和嗅觉能力的限制,客观性和推广性较差。利用电子鼻技术,结合模式识别算法对中药气味进行鉴别分析,可以提高灵敏度与准确度。本研究采用α-FOX3000电子鼻对不同贮藏时间(1年、3年)西洋参进行气味检测,并结合多层感知器网络识别技术建立判别模型;通过十折交叉验证和外部测试集验证对所建模型进行系统性能的评估。另外,采用逐步判别法对传感器阵列进行了优化。结果表明:该模型对不同贮藏时间西洋参具有较高的回判正确率(均为100%)和较好的泛化能力。优化前后的传感器阵列均能实现对不同贮藏时间西洋参的鉴别。为电子鼻在中药研究领域,尤其是中药的“储藏年限”、“有效期”等方面的应用提供实验依据。
The traditional smell identification of traditional Chinese medicine is often limited by personal identification experience and sense of smell, objectivity and promotion of poor. The use of electronic nose technology, combined with pattern recognition algorithm to identify the odor of traditional Chinese medicine, can improve the sensitivity and accuracy. In this study, α-FOX3000 electronic nose was used to detect the odor of American ginseng at different storage times (1 year and 3 years), and discriminant model was established based on multi-layer perceptron network identification technology. Ten fold cross validation and external test Model for system performance evaluation. In addition, the sensor array is optimized using the stepwise discriminant method. The results showed that this model had a higher accuracy (100%) and better generalization ability for American Ginseng with different storage time. Before and after optimization of the sensor array can achieve the different storage time of American ginseng identification. It provides the experimental basis for the application of electronic nose in the field of traditional Chinese medicine research, especially traditional Chinese medicine’s “shelf-life”, “expiration date” and so on.