解析全相位短区间正弦波频率估计算法研究  

Analytical All-phase Sinusoidal Frequency Estimation Algorithm in Short Interval

作　　者：闫格[1] 黄翔东[1] 刘开华[1] 

机构地区：[1]天津大学电子信息工程学院,天津300072

出　　处：《信号处理》2012年第11期1558-1564,共7页Journal of Signal Processing

基　　金：国家自然科学基金面上项目"基于稀疏采样的全相位互素谱分析理论及应用";项目编号:61271322;天津市科技支撑计划项目;"智能航空铅封技术研究";项目编号:10ZCKFGX03600

摘　　要：为解决波动周期数小于1情况下的正弦波频率估计问题,本文深入分析了短区间频率估计的误差与两边带的谱间干扰的关系,从而引入解析变换来抑制边带谱干扰;并借助全相位FFT获得了中心样点的相位估计,在基波谱线位置引入剪辑措施简化了相位估计流程,最终借助时移相位差校正法完成观测区间小于一个周期情况的正弦波频率估计。仿真实验表明,在波动周期数CiR处于0.4与1之间时,无噪时本文方法的频率测量误差百分比可控制在2%以内。To solve the problem of estimating sinusoidal frequency in case of CIR（ circles in record） 〈 1, this paper analyzed the relationship between error of frequency estimation in short interval and the interference of two side-spectrums, and therefore introduced analytic transform to inhibit this interference. Moreover, all-phase FFT was utilized to obtain the phase estimation of center sample. Furthermore, pruning technique at fundamental spectral position was also adopted which led to simplify the dataflow of phase estimation. Finally, this paper resorted to the technique of time shift phase difference correcting spectrum to achieve the frequency estimation in case of observation interval less than one period. Simulation shows that, the frequency error percentage of the proposed method can be controlled less than 2% when CiR ∈ [ 0.4,1 ] in noiseless case.

关 键 词：全相位FFT 谱间干扰 短区间 波动周期数 解析变换 

分 类 号：TN911.72[电子电信—通信与信息系统]