《毕业论文非参数谱估计算法与实现》.doc

非参数谱估计的算法与实现 内容摘要： 功率谱估计是随机信号分析中的一个重要内容，主要研究信号在频域中的各种特征，目的是根据有限数据在频域内提取被噪声的有用信号。以多元天线阵结合现代数字信号处理为基础的新型测向技术。USIC算法实现波达方向估计的实现。 关键词：周期图法 平均周期图法 BT谱估计 Capon法 Music算法 波达方向Algorithms and Implementations of Non-Parameter Spectral Estimations Abstract： Power spectrum estimate is an important part of the random signal analysis, which primarily studies the random signal characteristics in the frequency domain and aims at recovering the useful signal in the frequency domain contaminated with noise based on the limited observed data. The design focuses on the algorithm and implementation of the non-parameter spectral estimations. First, the most general spectral estimation – periodogram method is discussed. As the periodogram method is a biased estimate with the large variance and low resolution, the improved methods based on the periodgram method need to be studied. The main improvements include changing the shape of windows, averaging and smoothing the data, etc. As a result, several other non-parameter spectral estimation algorithms based on the periodgram method (such as the BT method, the averaging periodgram method, including the Bartlett method and Welch method) and their realizations by the software of MATLAB are given. Moreover, the direction of arrival (DOA) estimation is to apply the power spectrum estimation algorithms into the space domain, i.e., DOA is the new direction finding technique based on the multiple sensor array combined with the modern digital signal processing. Accordingly, the realizations of DOA estimation algorithms based on the method of Capon and MUSIC are discussed and compared in this design. Key words: Periodgram Averaging Periodgram BT Spectrum Capon method Music Algorithm Direction of arrival (DOA)目 录 1绪 论 1 1.1 引言 1 1.1.1 非参数谱估计 1 1.1.2 空间谱估计 1 1.2 本文的主要工作 2 2 几种常用的非参数谱估计算法与实现 2 2.1 引言 2 2.2 周期图法 3 2.2.1 周期图谱估计 3 2.2.2 周期图法的性质 3 2.2.3 周期图法在MATLAB中的仿真实现 4 2.3 BT法 5 2.3.1 BT法谱估计 5 2.3.2 BT谱估计的性质 6 2.3.3 BT谱估计