基于混合算法的数据融合技术研究

信息融合系统可以产生比系统中任一单元更有效、更精确的身份判决结果。在归纳了国内外敌我识别的方法后,提出一种将并行的自适应神经网络-模糊推理系统（ANFIS）、模糊理论以及D-S证据理论相结合的算法,并应用于雷达与IFF的数据融合中,实现空中目标的敌我识别。同时,由于要进行的是动目标识别,提出了利用多次测量数据模拟多传感器的方法来实现多次融合,仿真结果表明,该方法提高了身份判决的置信度,融合结果更加合理。

Research of Data Fusion Technique Based on Hybrid Algorith

Information fusion system can make a more accurate and potent identity＇s estimation than any single source in a system.Making a survey of the research on this issue,a hybrid algorithm based on adaptive neural network-fuzzy inference system（ANFIS）,fuzzy theory and D-S evidence theory is proposed.At the same time this algorithm is applied to data fusion of radar and IFF to realize the identification of air-target＇s friend or foe.Due to the recognition of moving target,using multi-measured data to simulate multi-sensor is proposed.The simulation results show that the degree of confidence is increased,furthermore the fusion result is more reasonable.