基于矩估计的遗传算法杂波幅度模型参数估计

选用了适用于多参数估计的遗传算法对Ku波段实测杂波数据进行K分布以及对数正态分布模型进行参数估计.为减少搜索范围,采用了计算量小的矩估计方法确定参数初始值;以常用的最小均方误差为目标函数.结果显示在遗传代数为50左右适应度函数收敛到稳定值,且拟合曲线与杂波数据直方图很好吻合.验证了遗传算法用于杂波模型参数估计的适用性.

Genetic Algorithm Applied to Parameter Estimation for Clutter Amplitude Model Based on MOM

In this paper, a genetic algorithm (GA) which is suitable for multi-parameter estimation is applied to estimate K and log-normal model parameters of measured Ku-band clutter data. The parameter is initialized by MOM to decrease the range of search; the minimum averaged square error(MSE) is chosen as the objective function. Results show that the fitness function will converge to a stable value when the number of generations is around 50 and the curves have a good fit to the histogram of radar da-ta, which testify the wide applicability of GA for parameter estimation of radar clutter.