基于RFNN的复杂机动事件检测方法  

A Method for Detecting Complex Mobile Event Based on RFNN

作　　者：陈伟[1] 陆廷金[2] 荣鹏辉[2] 李青[2] 赵延庆[1] 

机构地区：[1]空军工程大学工程学院,西安710038 [2]徐州空军学院,江苏徐州221000

出　　处：《电光与控制》2011年第3期61-65,共5页Electronics Optics & Control

基　　金：国家博士后基金资助项目(20090451225)

摘　　要：针对态势评估中复杂机动事件检测的精度及实时性问题,提出了基于粗糙集-模糊神经网络(RFNN)的事件检测方法,通过粗糙集理论获取数据样本中的最简规则集,然后根据这些规则构造模糊神经网络各层的神经元个数及相关参数初始值,最后用BP算法迭代求出网络的各种参数。仿真结果证明RFNN用于复杂机动事件检测的有效性,同时可以发现其在网络结构和收敛性方面的优势。Considering the requirement to accuracy and real-time performance in complex mobile event detection,a method based on Rough-Fuzzy Neural Network（RFNN） was proposed for event detection.The minimal rules were acquired from data samples by rough set theory.Then,these rules were used for constructing neural cell numbers and relative parameters in fuzzy neural network.Finally,the parameters of network were calculated out by using BP algorithm.Simulation showed the effectiveness of RFNN based complex mobile event detection,and proved that RFNN has superiorities at the aspects of network structure and convergence performance.

关 键 词：态势评估 复杂机动事件 事件检测 RFNN 

分 类 号：V271.4[航空宇航科学与技术—飞行器设计] TP108[自动化与计算机技术—控制理论与控制工程]