基于遗传算法的海杂波K分布参数估计

摘要

In the high-resolution radar,composite K distribution model can achieve good results on the experimental data of sea clutter. To describe the amplitude characteristics of the sea clutter by K distribution,the key is to determine the parameters of the distribution. This paper discuss the K distribution model and genetic algorithm,as well as improving algorithm point at the binary coding and fitness function. Finally an improved poly-population genetic algorithm is applied to parameter estimation,with the radar raw data released by the CSIR organization,to compare with the moment estimation and GA method, the result of simulate estimation has passed the mean square deviation test and the curves have a good fit to the histogram of radar data, which testify the wide applicability of improve GA for parameter estimation of radar clutter.%在高分辨率雷达中，复合K分布模型对海杂波的实测数据具有良好的拟合效果。使用K分布来描述海杂波幅度特性时，关键在于其分布参数的估计。本文在研究海杂波K分布模型、遗传算法的基础上，针对遗传算法中二进制编码、适应度函数标定存在的缺陷进行改进，然后将多种群遗传算法应用于参数估计，并利用CSIR组织公布的雷达实测数据进行仿真。仿真结果与统计量估计法以及标准遗传算法进行比较表明，利用改进遗传算法得到的拟合曲线与杂波数据直方图吻合较好，改进遗传算法在海杂波模型参数估计中具有较好性能。