摘要
为了对电子目标进行准确识别,一般采取多传感器融合的D-S证据理论方法。但是由于传统的D-S证据理论中各传感器对识别结果的重要性没有区分,造成识别结果不准确。变精度粗集模型是粗集模型的扩展,它允许一定程度错误分类率的存在,更符合实际数据的情况,将变精度粗集理论属性重要度概念应用到各传感器的重要性上,从而实现加权融合的证据理论。仿真结果表明,该方法对电子目标识别有效,尤其在传感器受到干扰时,识别结果更加可靠。
In order to recognize the electronic targets correctly, D-S evidence theory of multi-sensors amalgamation is usually applied to process these problems.But because the importance of every sensor to recognition result is not distinguished in traditional D-S evidence theory,so the recognition result is not exactly correct.Variable Precision Rough Set Model（VPRSM） is the extension of Rough Set Model（RSM） ,which allows wrong sort rate in some extent, so it is more appropriate for the real data.The attribute importance of variable precision rough set model is applied to set every sensor＇ s weight to recognition result, and then the weighted amalgamation evidence theory is realized.Simulation experiment results show that this new method is effective in electronic target recognition,especially when some sensor is interfered, the recognition result is more reliable.
出处
《无线电工程》
2015年第4期32-35,80,共5页
Radio Engineering
关键词
电子目标识别
变精度粗集
证据理论
属性重要度
基本概率赋值
electronic target recognition
VPRSM
evidence theory
attribute importance
basic probability assignment
