摘要
应用人工神经网络的原理, 用半导体氧化物气敏传感器阵列及单片机实现的呼出气体酒精含量的智能分析方法, 具有灵敏、准确、抗干扰能力强的特点。该文着重介绍了系统的原理, 配气方法, 网络的训练方法及酒精含量的预报结果。
An artificial intelligent technique for the determination of alcohol vapor in expiration, based on the principle of the artificial neural network is reported. The developed system consists mainly of a gas _ sensor array and a one _ chip micro _ computer. The technique has the merits of sensitivity, accuracy and antiinterference. The prediction of unknown content of alcohol vapor is satisfactory based on a good training with calibrated alcohol contents provided by gas chromatography.
出处
《分析测试学报》
CAS
CSCD
1999年第6期39-41,共3页
Journal of Instrumental Analysis
基金
国家自然科学基金!资助项目(29485009)
关键词
传感器
人工神经网络
酒精测定
呼气
驾驶员
Sensor, Artificial neural network, Alcohol, Expiration, Prediction