基于小波特征提取的气体传感器温度调制模式研究  

Research on temperature modulation mode of gas sensor based on wavelet feature extraction

作　　者：孙晶 赵寒涛[1,2] SUN Jing;ZHAO Hantao(Institute of Intelligent Manufacturing,Heilongjiang Academy of Sciences,Harbin 150090,China;Institute of Advanced Technology,Heilongjiang Academy of Sciences,Harbin 150020,China)

机构地区：[1]黑龙江省科学院智能制造研究所,黑龙江哈尔滨150090 [2]黑龙江省科学院高技术研究院,黑龙江哈尔滨150020

出　　处：《传感器与微系统》2019年第12期40-42,46,共4页Transducer and Microsystem Technologies

基　　金：黑龙江省院所基本应用技术研究专项项目(ZNBZ2019ZDH01)

摘　　要：为了有效提取传感器对被测气体的响应特征,通过对单个微热板式气体传感器在不同温度调制模式下,对3种可燃性气体的动态响应信号结合主成分分析(PCA)和概率神经网络(PNN)进行分析和测试,利用小波变换对温度调制下传感器的动态响应信号进行小波特征提取。结果表明:提取的小波特征系数能较好地反映被测气体的信息。In order to effectively extract the response characteristics of the sensor to the measured gas,the wavelet transform is used to extract the wavelet response of the dynamic response signal of the sensor under different temperature modulation modes.The dynamic response signals of three flammable gases in different temperature modulation modes are combined with principal component analysis(PCA)and probabilistic neural network(PNN)for analysis and testing of a single micro-hot plate gas sensor.The results show that the extracted wavelet characteristic coefficient can better reflect the information of the measured gas.

关 键 词：温度调制模式 特征提取 主成分分析(PCA) 概率神经网络(PNN) 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]