论文部分内容阅读
稳定分布可更好地描述实际中所遇到的具有显著脉冲特性的随机噪声。在简要介绍稳定分布统计特性的基础上,描述了稳定分布的谱表示。为了对AR SαS过程进行α谱估值,通过对广义Yule-Walker方程进行分析,借鉴杜宾推算法和Burg算法推导了一种求解广义Yule-Walker方程的新方法。仿真结果表明,该方法能够较为准确地估计出AR模型参数,且能准确地估计出AR SαS过程的α谱。
Stable distribution can better describe random noise encountered in practice with significant pulse characteristics. Based on a brief introduction of statistical properties of stable distribution, a stable distribution of spectral representation is described. In order to estimate α-spectrum of AR SαS process, a new method to solve the generalized Yule-Walker equation is deduced by analyzing the generalized Yule-Walker equation, taking into account the Dobbin algorithm and the Burg algorithm. The simulation results show that this method can accurately estimate the AR model parameters and accurately estimate the α spectrum of AR SαS process.