摘要
为了对多机协同对抗多目标的作战态势进行定量分析,根据多机协同对抗多目标的空战特征,基于神经网络和证据理论,以传感器和武器能力为证据建立敌我双方对抗态势分析识别框架。将学习引入证据合成过程,用神经网络学习对基本可信数修正系数进行优化,使证据合成能够体现证据之间的相干性和主次关系。以无人作战飞机编队对抗敌地对空防御系统为例,对所研究的态势分析方法进行了验证。获得的计算结果与通过策略实际含义分析的结果相一致,具有合理性。
Based on the characteristics of multi-aircraft cooperated attack against multiple targets, neural networks and evidence theory, an identification frame for analyzing the situation between our aircraft and enemy targets with the sensor and weapon capability as evidences is established for the quantitative analysis of airfight situation. In the process of evidence synthesis, the correction coefficient of basic probability function is optimized by utilizing neural network learning so that the evidence synthesis can show the relativity and weightness between one evidence and others. Situation analysis method is validated by an example of UCAV swarm carrying out mission of SEAD. It is shown that the result obtained from information fusion is consistent with the result obtained by analyzing actual airtight process.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2008年第2期292-296,共5页
Systems Engineering and Electronics
基金
国防“973”基金(5130802)
航空科学基金(05D01002)资助课题
关键词
多架飞机
协同对抗
多目标
态势分析
神经网络
证据理论
multi-aircraft
coordinated attack
multi- target
situation analysis
neural networks
evidence theory
