基于双递归自适应滤波的高精度频率估计算法  

High-precision frequency estimation algorithm based on double recursion adaptive filtering

作　　者：梁国龙 张瑶 付进 张光普 

机构地区：[1]哈尔滨工程大学,水声技术国家级重点实验室,黑龙江哈尔滨150001

出　　处：《哈尔滨工程大学学报》2013年第2期202-206,共5页Journal of Harbin Engineering University

基　　金：国家自然科学基金资助项目(51009042);高等学校博士学科点专项科研基金资助项目(20102304120030)

摘　　要：针对线谱信号的检测与频率估计问题,在分析自适应瞬时频率估计算法的基础上,提出一种具有双递归特性的信号频率提取算法:将自适应滤波器组成递归环路,通过比较不同的自适应滤波器输出信号的瞬时频率序列方差来进行滤波器中心频率以及带宽的更新,从而获得最优的滤波输出.仿真分析表明,与传统方法相比,此方法具有更好的噪声抑制能力以及频率跟踪能力,还可以获得可控的频率估计精度.According to problems existed in the previous research studies about the detection and frequency estima- tion of a line-spectrum signal, this paper proposes a signal frequency extraction algorithm which has double recur- sion features, based on the analysis of adaptive instantaneous frequency estimation algorithm. The high-precision frequency estimation algorithm was based on a double recursion adaptive filter grouped into a circle. By comparing the instantaneous frequency sequence~ variances of different adaptive filters, update the fiher~ center frequency and bandwidth, in order to get the optimal filter output. The simulation analysis confirmed that, in comparison to the traditional methods, this method provides a better noise immunity and frequency tracking capability, as well as con- trols the frequency estimation accuracy.

关 键 词：线谱信号 自适应滤波 瞬时频率估计 频率估计算法 双递归滤波 

分 类 号：TN911[电子电信—通信与信息系统]