改进的光纤光栅温度传感网络加权融合算法

英文篇名：Improved FBG Temperature Sensing Network Weighted Fusion Algorithm
中文刊名：南开大学学报(自然科学版)
英文刊名：Acta Scientiarum Naturalium Universitatis Nankaiensis
作者：袁其平 ; 马磊明 ; 宋殿友 ; 童峥嵘 ; 杨秀峰 ; 曹晔
英文作者：Yuan Qiping1 ; Ma Leiming2 ; Song Dianyou1 ; Tong Zhengrong2 ; Yang Xiufeng2 ; Cao Ye2(1.School of Electronics Information Engineering ; Tianjin University of Technology ; Tianjin 300384 ; China ; 2.School of Computer and Communication Engineering ; Tianjin University of Technology ; Tianjin 300384 ; China)
中文关键词：光栅网络 ; 数据融合 ; 加权算法 ; 异常值处理
英文关键词：grating network ; data fusion ; weighted algorithm ; outlier disposal
出版日期：2012-12-20
机构：天津理工大学电子信息工程学院;天津理工大学计算机与通信工程学院;
年：2012
期：06
出版单位：南开大学学报(自然科学版)

摘要

针对光栅传感网络数据加权融合中出现异常值会导致融合结果偏差较大的问题,在分析光栅传感系统中的噪声和干扰后,提出一种改进的自适应加权融合算法.该算法适用于观测数据不满足正态分布的情况,先对原始数据进行截尾加权融合,再以截尾融合结果代替异常值进行加权最小均方差融合,既充分利用了传感数据,也降低了融合算法的复杂度.在光纤光栅温度传感网络实验中使用该算法,其输出方差为0.083 9,小于加权最小均方误差算法方差0.383 9和截尾加权算法方差0.254 2,证实了该算法可行性.
To deal with outliers caused by noise and disturbance during weighted fusion of Fiber Bragg Grating(FBG) temperature sensing network resulting in augmenting error,an improved adaptable weighted fusion algorithm is presented after analysing nosie and interference in FBG sensing system.This algorithm is suitable for sensing data processing which don't obey normal distribution.Firstly,Trimmed Mean Weighted(TMW) algorithm is used to process multisensor data fusion,and then replace outliers with the fusion result of TMW algorithm,finally Weighted Minimum Mean Square Error(WMMSE) algorithm is used to process twice-fusion,here data are fully used and complexity of algorithm are obviously reduced.New algorithm is applied in fiber Bragg grating temperature sensing,output's variance 0.083 9 less than WMMSE algorithm's variance 0.383 9 and TMW algorithm's variance 0.254 2,correctness and accuracy are testified.

引文

1李俊,吴瑾,高俊启.一种监测钢筋腐蚀的光纤光栅传感器的研究[J].光谱学与光谱分析,2010,30(1):283-286.
    2 Dong J Z,Huang D F,Yao H.Advances in multi-sensor data fusion:algorithms and applications[J].Sensors,2009,9(10):7 771-7 784.
    3张卫华,童峥嵘,苗银萍,等.数据融合在光纤光栅传感网络中的应用[J].南开大学学报:自然科学版,2009,42(5):62-65.
    4 Khaleghi B,Razavi S N.Multisensor data fusion:Antecedents and directions[C/OL]//3rd international conferenceon Signals,Circuits and Systems(SCS),Yadis Resort Hotal,Medenine,Tunisia,Nov 6-8,2009.[2011-03-11].http://ieeexplore.ieee.org/stamp.jsp?tp=&arnumber=5412296.
    5 Gao S S,Zhong Y M,Li W.Random weighting method for multisensor data fusion[J].Sensors,2011,11(9):1 955-1 961.
    6 Su K L,Jau Y M,Jeng J T.Modeling of nonlinear aggregation for information fusion systems with outliers based onthe choquet Integral[J].Sensors,2011,11(3):2 426-2 446.
    7Kay S M.统计信号处理基础[M].罗鹏飞,张文明,刘忠,等,译.北京:电子工业出版社,2003.
    8金建华,阙沛文.多传感器测量数据的切尾加权融合算法[J].仪器仪表学报,2002(S1):303-304.
    9乔学光,丁锋,贾振安,等.基于光源滤波的高精度光纤光栅地震检波解调系统[J].光学学报,2010,8:2 219-2 223.