强噪声背景下正弦信号频率估计算法研究

发布时间：2019-06-21 00:31

【摘要】：从强噪声中准确提取单一正弦信号的频率是通信系统、信号处理等领域一个非常重要的问题。目前,强噪声背景下正弦信号频率估计已经成功应用于雷达探测、语音信号处理、声纳地震、通信系统中的信号恢复、桥梁振动检测、生物医学检测以及电子通信技术等众多领域,引起越来越多的学者的关注和重视。因此,对正弦信号频率估计研究具有极其重要的理论意义和实际应用价值。频率是正弦信号最重要的参数和最本质的特征,频率估计研究是信号处理领域的一个经典课题。本论文对几类正弦信号频率估计算法进行研究,包括经典的DFT算法、最大似然估计法、PHD算法、MC算法、TSA算法、Rife算法以及傅里叶系数插值迭代算法,分别在算法原理及计算量的基础上比较了算法的性能,通过仿真实验,得出频率估计均方根误差与信噪比之间的关系,并与频率估计均方根误差克拉美-罗下限做了比较。并且在分析总结各种算法的基础上,提出了相应的改进算法。本文的主要研究工作及结论如下:1、提出了基于正弦信号LP性质的分段自相关频率估计新算法。该算法解决了两步自相关频率估计算法(TSA算法)估计性能和计算量不能兼顾的缺点,在计算量较小的情况下逼近TSA2算法的性能,弥补了 TSA1改进后带来的计算量增加的问题。并对分段自相关算法进行了仿真实验,用结果分析其算法效果较好。2、提出了一种基于FFT的正弦信号频率估计新算法。通过分析Rife算法和傅里叶系数迭代算法的性能可知,Rife算法计算简单、精度不高,而傅里叶系数迭代算法需要两次迭代之后才会到达精度要求,每次迭代的计算量大。结合Rife算法和傅里叶系数迭代算法这两种算法的优缺点,本文新提出了一种改进后的高精度频率估计算法。并对此方法进行了仿真实验,结果分析其算法效果较好。
[Abstract]:The frequency of extracting a single sinusoidal signal from strong noise is a very important problem in the fields of communication system, signal processing and so on. At present, sinusoidal signal frequency estimation under the background of strong noise has been successfully applied to radar detection, speech signal processing, Sonar earthquake, signal recovery in communication system, bridge vibration detection, biomedical detection and electronic communication technology, which has attracted more and more scholars' attention and attention. Therefore, the study of sinusoidal signal frequency estimation is of great theoretical significance and practical application value. Frequency is the most important parameter and essential feature of sinusoidal signal. Frequency estimation is a classical topic in the field of signal processing. In this paper, several kinds of sinusoidal signal frequency estimation algorithms are studied, including classical DFT algorithm, maximum likelihood estimation method, PHD algorithm, MC algorithm, TSA algorithm, Rife algorithm and Fourier coefficient interpolation iterative algorithm. The performance of the algorithm is compared on the basis of algorithm principle and computation. Through simulation experiments, the relationship between root mean square error and signal to noise ratio of frequency estimation is obtained. It is compared with the root mean square error of frequency estimation. Based on the analysis and summary of various algorithms, the corresponding improved algorithms are proposed. The main research work and conclusions of this paper are as follows: 1. A new piecewise autocorrelation frequency estimation algorithm based on sinusoidal signal LP property is proposed. The algorithm solves the disadvantage that the performance and computation of two-step autocorrelation frequency estimation algorithm (TSA algorithm) can not be taken into account, and approaches the performance of TSA2 algorithm when the amount of computation is small, which makes up for the problem of increasing the amount of computation caused by the improvement of TSA1. The piecewise autocorrelation algorithm is simulated, and the results show that the algorithm is effective. 2, a new sinusoidal signal frequency estimation algorithm based on FFT is proposed. By analyzing the performance of Rife algorithm and Fourier coefficient iterative algorithm, it can be seen that the calculation of Rife algorithm is simple and the accuracy is not high, while the Fourier coefficient iterative algorithm needs two iterations to meet the accuracy requirements, and the computation of each iteration is large. Based on the advantages and disadvantages of Rife algorithm and Fourier coefficient iterative algorithm, an improved high precision frequency estimation algorithm is proposed in this paper. The simulation experiment of this method is carried out, and the results show that the algorithm is effective.
【学位授予单位】：哈尔滨工程大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.23

【参考文献】