基于多项式回归模型的气体传感器漂移矫正方法  

Drift correction method of gas sensor based on polynomial regression model

作　　者：王飞 房汉鸣 曾丽蓓 WANG Fei;ZHANG Hai-tao;FANG Han-ming;ZENG Li-bei(Army Logistical University of PLA,Chongqing 401311,China)

机构地区：[1]陆军勤务学院军事物流系,重庆401311

出　　处：《传感器与微系统》2018年第11期28-30,共3页Transducer and Microsystem Technologies

基　　金：国家自然科学基金资助项目(61871402;61271449;61302175);重庆市首批百名学术学科人才专项项目(渝教人2012-44号);重庆市自然科学基金重点资助项目(CSTC2015JCYJBX0017)

摘　　要：针对气体传感器漂移问题,提出了基于多项式回归模型的气体传感器漂移矫正方法:所提方法在分析了气体传感器漂移原理的基础上,通过实验得到了漂移规律并构建了多项式回归模型;通过对多项式回归模型和偏最小二乘回归模型的评价,证明了前者的优越性;运用所提方法进行气体传感器气体检测的实验验证,结果表明:基于多项式回归模型气体传感器漂移矫正方法对气体识别精度的提高十分有效。To solve the problem of gas sensor drift,a method of gas sensor drift correction based on polynomial regression model is proposed. Based on analysis on drift principle of gas sensor,the drift rule is obtained by experiment and the polynomial regression model is constructed.Experimental results of gas sensor detection using this method show that the drift correction method based on polynomial regression model is effective to improve the precision of gas identification.

关 键 词：气体传感器 漂移矫正 多项式回归模型 识别精度 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]