期刊文献+

迭代异方差估计及其在多传感器数据融合中的应用 被引量:4

Iterative Heteroscedastic Variance Estimation with Its Applications for Multisensor Data Fusion
在线阅读 下载PDF
导出
摘要 本文提出一种适应任意噪声分布的迭代渐近无偏估计异方差的方法,它能对多个不同测量噪声的方差进行估计.不同于传统的异方差估计算法,本文提出的迭代异方差估计,可在不损失估计精度和减少运算量的前提下,对多个不同的测量噪声方差进行捕获和跟踪.在多传感器数据融合中的应用结果表明:本文提出的方法具有估计稳定性好、运算简单和具有较强的鲁棒性等优点,仿真和实验的结果均证明了提出方法的有效性和可行性. In this paper, an iterative asymptotic unbiased heteroscedastic variance estimation method, suitable for the measurement noise with arbitrary possibility distributions, is presented. The method can estimate the variances of several different measurement noises at the same time. Unlike traditional heteroscedastic variances estimation algorithms, the proposed iterative estimation method can capture and track the vary variances of the measurement noises with less computation while its estimation accuracy is kept. Its application results on multisensor data fusion show that the presented algorithm is advantageous to the estimation stability, computation simplicity, and good robustness. Both the simulation and experiment results verify the effectiveness of the proposed method.
出处 《电子学报》 EI CAS CSCD 北大核心 2008年第10期1938-1943,共6页 Acta Electronica Sinica
基金 国家自然科学基金(No.60572023)
关键词 迭代异方差估计 最小均方误差 多传感器数据融合 加权最小二乘 iterative heteroscedastic variances estimation minimum mean square error (MMSE) multisensor data fusion weighted least square
作者简介 赵晋 男,1982年8月生于山西大同,2005年毕业于复旦大学电子工程系,获理学学士学位.现为复旦大学电子工程系博士研究生,主要从事统计信号处理、数据融合和无线通信等方面的研究.E-mail:dr.zhaojin@gman.com 张建秋 (本文通信作者)男,1962年生于湖南,1996年毕业于哈尔滨工业大学,获博士学位,1999年至2002年在英国格林威治大学工作,现为复旦大学电子工程系教授,博士生导师,IEEE高级会员.主要研究方向:信息处理理论及其在新型传感器、仪器和测量中的应用.E-mail:iqzhang01@fudan.edu.cn 高羽 女,1978年生于哈尔滨,2001年毕业于哈尔滨工业大学电气工程系,现为复旦大学博士研究生.主要从事多传感器信号处理及数据融合方面的研究.
  • 相关文献

参考文献8

  • 1J S J Peri. Approaches to Multisensor Data Fusion[J]. J Hopkins Apl Tech D,2001,22(4) :624 - 633.
  • 2G. Saporta. Data Fusion and Data Grafting [J]. Computational Statistics Data Anal. 2002,38 : 465 - 473.
  • 3D L Hall. Mathematical Techniques in Multisensor Data Fusion [M]. Norwood, MAArtech House, 1992.
  • 4Lijun Xu, X M Li, F Dong, Y Wang, L A Xu. Optimum estimarion of the mean flow velocity for the multi-electrode inductance fiowmeter[J].Means Sci Technol, 2001,12 (6) : 1139 - 1146.
  • 5Susan D H,Roger A H,David B D. Estimating Heteroscedastic Variances in Linear Models[J].Journal of the American Statistical Association, 1975,70(6) :380 - 385.
  • 6Lijun Xu, Jian Qiu Zhang, Yong Yah. A Wavelet-Based Multisensor Data Fusion Algorithm[ J]. IEEE Trans Instrum. Meas., 2004,50(6) : 1539 - 1544.
  • 7L Angrisani, M D' Apuzzo, R S Lo Moriello. Unscented TransformA Powerful Tool for Measurement Uncertainty Evaluation[J]. IEEE Trans Instrum. Meas. 2006, 55 ( 3 ) : 737 - 743.
  • 8杨惠娟,张建秋.基于小波去噪和数据融合的多传感器数据重建算法[J].复旦学报(自然科学版),2005,44(1):161-165. 被引量:3

二级参考文献7

  • 1Xu L J, Li X M, Dong F, et al. Optimum estimation of the mean flow velocity for the multi-electrode inductance flowmeter [J]. Measurement Science and Technology, 2001, 12: 1139-1146.
  • 2Gao X Z, Valiviita S, Ovaska S J, et al. Neural networks-based approach to the acquisition of acceleration from noisy velocity signal [EB/OL]. http:∥ieeexplore.ieee.org/ie14/5565/14915/0067686.pdf, 1998-05-18/2004-03-04.
  • 3Brown R G. Introduction to Random Signals and Applied Kalman Filtering with Matlab Exercises and Solutions, 3rd Ed [M]. New York: Wiley, 1996.
  • 4Folik J, Abdelrahman M, Kandasamy P. A confidence-based approach to the self-validation, fusion and reconstruction of quasi-redundant sensor data [J]. IEEE Trans Instrum Mea, 2001, 50(6): 1761-1769.
  • 5杨力华 戴道清 黄文良.信号处理的小波引导[M].北京:机械工业出版社,2002..
  • 6Stein C. Estimation of the mean of a multivariate normal distribution [J]. Annals of Statistics, 1981, 11(9):1135-1151.
  • 7Zhang J Q, Ma J, Yan Y. Assessing blockage of sensing line in a differential-pressure flow sensor by using the wavelet transform of its output [J]. Measurement Science and Technology, 2000, 11(3): 178-184.

共引文献2

同被引文献31

引证文献4

二级引证文献8

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部