一种改进的海杂波数据统计拟合度分析方法

An Improved Analysis Method of Statistical Goodness-of-fit for Sea Clutter Data

海杂波分布模型对于雷达海面目标检测方法和性能估计具有现实意义,因此合适的统计模型的参数估计和拟合度(GOF)分析非常重要。本文介绍了几种常用的GOF分析方法,如Chisquare、K-S、MSD、MMSD,进而提出了分段均方差(SMSD)方法。这种方法将海面杂波振幅的统计分布模型划分为几个函数区间,更容易观察或得到局部最优估计,也为进一步选择检测方法建立了更好的基础。

Distribution models are of practical significance to the radar target detection method and performance evaluation when facing a sea clutter background,so the great importance of parameter estimation of the suited statistical model and analysis of goodness-of-fit(GOF) is highlighted. Several conventional methods for GOF analysis are introduced,such as Chi-square,K-S,MSD,MMSD,and then the segmented mean square difference(SMSD) is proposed. The statistical distribution model of sea clutter amplitude is divided into several function intervals by using the proposed method. It’s easier for us to observe or get partial optimal estimation by applying GOF test methods to the range of some object function intervals and establish a better foundation for choosing detection method further.