| 厚膜SnO_2气体传感器的嗅觉特征提取与处理 |
| |
| 引用本文： | 邵应清,邹小波. 厚膜SnO_2气体传感器的嗅觉特征提取与处理[J]. 传感器与微系统, 2000, 0(6) |
| |
| 作者姓名： | 邵应清 邹小波 |
| |
| 作者单位： | 江苏理工大学汽车学院!江苏镇江2120132(邵应清)，江苏理工大学生物与环境工程学院!江苏镇江212013(邹小波) |
| |
| 摘 要： | 用一组厚膜ＳｎＯ２气体传感器阵列模拟人的嗅觉形成过程，对５种不同体积分数乙醇溶液进行分 析。详细叙述了实验过程，分别从每个气体传感器与气体反应的曲线中提取４个特征，用ＢＰ神经网络对 样本特征值的处理，对不同体积分数乙醇溶液进行识别。神经网络对训练集的回判正确率为１００％，对测 试集测试正确率为９０％。
|
| 关 键 词： | 电子鼻 BP神经网络 气体传感器阵列 |
Processing and picking - up the olfaction character of thick film SnO_2 gas sensors |
| |
| SHAO Ying - qing,ZOU Xiao - bo. Processing and picking - up the olfaction character of thick film SnO_2 gas sensors[J]. Transducer and Microsystem Technology, 2000, 0(6) |
| |
| Authors: | SHAO Ying - qing ZOU Xiao - bo |
| |
| Abstract: | An thick - film SnO2 gas sew array used to imitatethe function of human olfaction system is used to analyze five different concentrations of ethanol solution. A detailed exposition is given to the process of the experiment. Four characters from each curve of the gas sensor's reaction to gas are picked - up. BP(Back - Propagation) neural network recognition are used to identify the samples from the five different ethanol solutions. The recognition probability of the network is 100% to the training set and 90% to the testing set. |
| |
| Keywords: | electronic nose BP neural networks gas sensor array |
| 本文献已被 CNKI 等数据库收录！ |
