摘要
推导了韦布尔分布参量满足的最大似然(ML)估计方程和其为球不变随机矢量(SIRV)的条件,给出了模拟相关韦布尔分布雷达杂波的原理和流程。在此基础上,基于球不变随机过程(SIRP)法仿真了具有高斯谱的韦布尔分布及其特例——瑞利和指数分布,弥补了零记忆非线性(ZMNL)法不能独立控制边缘概率密度函数(PDF)与相关函数的不足。最后不仅验证了模拟数据的PDF与理论分布,估计的功率谱与理论谱的吻合程度,而且分析了ML估计的分布参数与真实值之间的相对误差。结果表明,基于SIRP法仿真相关韦布尔分布雷达杂波是有效的。
This paper derives the maximum likelihood（ML） estimation equations and spherically invariant random vectors（SIRV） conditions for Weibull distribution parameters.The principles and flow of correlated Weibull distribution radar clutter simulation are given.On this basis,Weibull distribution and its special cases—Rayleigh and exponential distributions with Gaussian spectrum are simulated using spherically invariant random process（SIRP） method,which makes up the deficiency that zero memory nonlinearity（ZMNL） method cann＇t control the marginal probability density function（PDF） and correlation function independently.Finally,the agreement between the PDF of experimental data and the theoretical distribution and the agreement,between the estimated power spectrum and the theoretical power spectrum are compared and verified.The relative errors between the ML estimated distribution parameters and the truth-values are also analyzed.The result proves that the simulation of correlated Weibull distribution radar clutter based on SIRP method is effective.
出处
《雷达科学与技术》
2011年第3期253-258,共6页
Radar Science and Technology
基金
总装重点基金(No.9140A07050908)
