欢迎来到天天文库
浏览记录
ID:52400017
大小:753.84 KB
页数:6页
时间:2020-03-27
《一种湿度传感器温度补偿的融合算法.pdf》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、第25卷第l2期传感技术学报V01.25No.122012年l2月CHINESEJOURNALOFSENSORSANDACⅥORSDec.2012AFusionAlgorithmforHumiditySensorTemperatureCompensationXINGHongyan,,PENGJiwei一,Lg)Wenhua,XUIVei,Xiangjuan(1.JiangsuKeyLaboratoryofMeteorologicalObservationandlnJormationProcessing,NanjingUniversityofInformationSc
2、ienceandTechnology,,vnng210044,China;2.Schoolofelectronicandinformationengineering,NangUniversityofInformationScienceandTechnology,?Cang210044,China;3.AtmosphericObservationTechnologyCenter,ChinaMeteorologicalAdministration,Beijing100081,China;4.NingxiaMeteorologicalObservationTechnolo
3、gySupportCenter,Yingchuan750002,China)Abstract:Accordingtothehumiditysensorsontheautomaticweatherstationinfluencedeasilybytemperatureintheactualapplication,RBFneuralnetworkandleastsquarescombiningfusionalgorithmisproposedtorealizecompensationofthehumiditysensor.Thecharacteristiccurveth
4、atthehumiditysensorisundertheinfluenceofthetemperatureisdividedintotwoparts,i.e.anon—linearpartandalinearpart,andintheadaptivedeterminationofthelinearsegmentsandnon—linearsegments,theleastsquaresmethodisusedtofittingastraightlineequationinlinearsegments,thenRBFneuralnetworkisusedtocomp
5、ensatetheimpactoftemperatureinnon—linearsegments.SimulationresultsshowthatcomparedwithBPneuralnetworkandleastsquarespoly,themethodiseasilyimplement,thespeedoffittingtrainingisfaster,andmakesthetemperaturescompensationforhighaccuracy.Thetemperaturecompensationofthehumiditysensorcanbeeff
6、ectivelyusedtoimprovesensormeasurementaccuracyandreliability.Keywords:humiditysensor;fusionalgorithm;RBFneuralnetwork;leastsquares;temperaturecompensationEEACC:7230;7320Rdoi:10.3969/j.issn.1004—1699.2012.12.018一种湿度传感器温度补偿的融合算法术行鸿彦',彭基伟,吕文华,徐伟,武向娟(1.南京信息工程大学江苏省气象探测与信息处理重点实验室,南京210044;2.
7、南京信息工程大学电子与信息工程学院,南京210044;3.中国气象局气象探测中心,北京100081;4.宁夏大气探测技术保障中心,银川750002)摘要:针对自动气象站上湿度传感器在实际应用过程中易受温度影响的问题,提出采用RBF神经网络与最小二乘相结合的融合算法实现湿度传感器的温度补偿。该方法将湿度传感器在温度影响下的特性曲线分为两个非线性段和一个线性段,并且自适应的确定线性段和非线性段,在线性段利用最小二乘方法拟合出直线方程,在非线性段利用RBF神经网络补偿温度产生的影响。仿真结果表明,这种方法简单易行,与一般的BP神经网络和最小二乘多项式方法相比,具有拟合
此文档下载收益归作者所有