分布式检测系统的模糊有序统计融合算法  

A Fuzzy Order Statistics Fusion Rule for Distributed Detection System

作　　者：曲付勇[1] 宋玉珍[1] 孟祥伟[2] 

机构地区：[1]海军航空工程学院研究生管理大队,山东烟台264001 [2]海军航空工程学院电子信息工程系,山东烟台264001

出　　处：《信息与电子工程》2007年第6期424-427,共4页information and electronic engineering

摘　　要：为提高基于二元局部判决的分布式检测系统的检测性能,研究了基于模糊隶属度函数的分布式检测系统融合算法,提出了模糊有序统计融合准则。通过对N个传感器进行分布式检测系统仿真,结果表明:模糊有序统计融合准则在均匀环境中能获得比基于二元判决准则更好的检测性能,也好于同样采用模糊隶属度函数的求和准则的检测性能,且在多目标环境下获得了较强的适应性。A new fusion rule, Fuzzy Order Statistics（FOS） fusion rule, is presented to improve the detection performancc of binary local decision in multiscnsor distributed detection systems. Analytical expressions are given for the false alarm probability. Simulation results based on local sensors in a distributed system indicate that in homogeneous environment, order statistics fusion rule based on fuzzy membership function has better performance than that the binary local decision rule has, and order statistics fusion rule is also superior to the ＂snm＂ fusion rule based on fnzzy membership function. The order statistics fusion rule is robust in multi-targets.

关 键 词：模糊 有序统计 融合准则 多传感器 分布式检测 

分 类 号：TP202[自动化与计算机技术—检测技术与自动化装置]