电子学报 ›› 2018, Vol. 46 ›› Issue (9): 2108-2114.DOI: 10.3969/j.issn.0372-2112.2018.09.010

• 学术论文 • 上一篇    下一篇

自适应陷波器噪声统计性能与相关性分析

李明, 涂亚庆, 沈艳林, 沈廷鳌, 毛育文   

  1. 后勤工程学院后勤信息与军事物流工程系, 重庆 401311
  • 收稿日期:2016-09-08 修回日期:2018-01-22 出版日期:2018-09-25 发布日期:2018-09-25
  • 作者简介:李明 男,1985年1月出生于山西太原,现为后勤工程学院讲师,主要研究方向为信号处理,智能检测与仪表.E-mail:limitonly@126.com;涂亚庆 男,1963年6月出生于重庆市,现为后勤工程学院教授,博士生导师,主要研究方向为智能检测与仪表、智能自动化系统.E-mail:yqtcq@sina.com
  • 基金资助:
    国家自然科学基金(No.61601493,No.61501491);重庆市自然科学基金(No.CSTC2015jcyjA00029)

Noise Analysis of Statistical Performance and Correlation for Adaptive Notch Filter

LI Ming, TU Ya-qing, SHEN Yan-lin, SHEN Ting-ao, MAO Yu-wen   

  1. Department of Information and Military Logistics Engineering, Logistical Engineering University, Chongqing, China
  • Received:2016-09-08 Revised:2018-01-22 Online:2018-09-25 Published:2018-09-25
  • Supported by:
     

摘要: 噪声是造成自适应陷波器(Adaptive Notch Filter,ANF)频率估计结果有偏的主要原因之一,在低信噪比条件下,噪声将极大的影响ANF频率估计方法的性能,特别在进行ANF统计性能分析的过程中,ANF传递函数的线性近似方法不再适用,此时将无法对ANF频率估计方法的精度等性能进行准确评估.为此,首先利用相关函数和功率谱分析噪声经不同变换后均值与方差的统计性能;其次,利用噪声分析的统计性能结果计算ANF频率估计方法的性能,为提升ANF频率估计方法的精度奠定理论分析基础;最后,对ANF频率估计方法的性能进行理论值与实际值的比对分析,结果表明噪声分析方法的有效性,并可近似预测ANF频率估计方法的稳态性能.

 

关键词: 正弦信号, 噪声分析, 自适应陷波器, 频率估计

Abstract: Noise are one of the main reasons for biased frequency estimation of ANF (Adaptive Notch Filter).Noise will greatly affect the performance of ANF frequency estimation method under low SNR.When analyzing the ANF statistical performance,some linear approximation methods for ANF transfer function will degrade,the performance of ANF frequency estimation method cannot be evaluated precisely.So the noise statistical performances of mean and variance after different transform are analyzed by using correlation function and power spectrum method.Noise statistical performance results are used for performance analysis of ANF frequency estimation method,meanwhile a theoretical analysis foundation is established to improve the accuracy of ANF frequency estimation method.The performances of ANF frequency estimation method are compared by the theoretical values and the actual values,which show the effectiveness of the noise analysis method,and the steady-state performance of frequency estimation can be estimated approximately.

Key words: sinusoidal signal, noise analysis, adaptive notch filter, frequency estimation

中图分类号: