【摘要】 对复杂混合气体 /气味的定性识别是电子鼻技术应用的重要方面。文章以对不同品牌的葡萄酒样本的定性识别为例 ,尝试利用气体传感器温度动态响应信息并结合Kohonen算法 ,实现对复杂混合气体 /气味的定性识别。仿真验证结果表明 :采用仅由 3个气体传感器构成阵列的电子鼻系统 ,利用上述方法对 5种不同品牌的葡萄酒样本的识别准确率达到 10 0 %。文中同时针对识别对象的特点 ,介绍了一种差值抵消的数据预处理方法。更多还原

【Abstract】 To identify the complicated odor is the important aspect of the application of electronic nose (E-nose). This use basing on the dynamic responses of the temperature of the sensors and Kohonen arithmetic is presented for identifying the complicated odor. Using this method, the simulation result of identifying 5 brands of wine by the E-nose composed of only 3 gas sensors shows that preciseness ratio of recognition reaches to 100%. Aiming to the characteristic of the identified objects, a new method of the pretreatment of data is presented.更多还原