基于非对称广义高斯模型的局部优化检测中的应用  

Application to Locally Optimum Detection Based on Asymmetric Generalized Gaussian distribution

作　　者：李志明[1] 汪太月[2] 李宏伟[1] 

机构地区：[1]中国地质大学数理学院,湖北武汉430074 [2]黄石理工学院数理学院,湖北黄石435002

出　　处：《海南大学学报（自然科学版）》2009年第1期73-77,共5页Natural Science Journal of Hainan University

基　　金：国家自然科学基金项目(60672049)

摘　　要：为了在非高斯的情况下优化信号的检测,笔者致力于提供一种通用的噪声概率密度函数的现实模型.这个模型仅仅依赖于几个容易且能快速估计的参数,还能够适应于诸如对称和非对称,以及带有不同锐利程度的不同噪声.为了达到这个目的,一种来源于广义高斯函数的高阶统计量的模型被提出,它依赖于3个参数:表示不同锐利程度的峰度参数,以及描写不同于对称函数且联合提供偏斜程度的左右方差参数.这个模型在局部优化检测的设计中得到很好运用,被水下声学噪声干扰的信号检测证实了这个结果.The work is dedicated to provide modeling of a generic noise probability function （ PDF）, in order to optimize signal detection in non-Gaussian environments. The model only depends on a few parameters which can be estimated easily and quickly, and can be adapted to many kinds of noise such as symmetric or asymmetric or with variable sharpness. At the end, we presented a new high order statistic model which derives from the generalized Gaussian function and depends on three parameters： kurtosis, for representing variable, and left and right variance whose combination provides skewness, to describe deviation from symmetry. The model is applied in the design of a Locally Optimum Detection test. Promising experiment results are presented which derive from the application of the test for detecting signals corrupted underwater acoustic ship traffic radiated noise.

关 键 词：广义高斯函数 高阶统计量 最大相对效率 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]