基于云模型的DSm证据建模及雷达辐射源识别方法

为了提高雷达辐射源特征参数存在互相交叠和多个模式情况的雷达辐射源正确识别率，该文提出一种基于云模型的DSm(Dezert-Smarandache)证据建模及雷达辐射源识别方法。该方法首先将存在互相交叠和多个模式的先验雷达辐射源特征参数进行基于云模型的DSm建模，然后将含有噪声的测量信号特征参数进行基于云模型的DSm隶属度赋值，再通过隶属度与基本信度赋值的关系求得DSm模型的基本信度赋值，最后通过DSmT+PCR5的方法将多传感器测量信号的同特征的基本信度赋值进行融合，再将各特征的融合结果进行DSmT+PCR5融合得到最终的识别结果，如果仅为单传感器测量信号的特征参数，则仅将不同特征参数的基本信度赋值进行DSmT+PCR5得到融合识别结果。最后通过多种情况下的仿真实验，验证了该文方法的优越性。

DSm Evidence Modeling and Radar Emitter Fusion Recognition Method Based on Cloud Model

To improve the correct radar emitter recognition rate in cases that radar emitter characteristic parameters are overlapped with each other and existence of multiple modes, a DSm (Dezert-Smarandache) evidence modeling and radar emitter fusion recognition method based on cloud model is proposed. First, the radar emitter characteristic parameters which are overlapped and have multiple modes are modeled in DSm frame based on cloud model, then the degree of membership of unkonwn radar emitter signal belonging to prior radar types of each characteristic parameter is obtained by this model. Second, the basic belief assignments in DSm frame based on cloud model are obtained by the relationship between degree of membership and basic belief assignments. Thirdly, the basic belief assignments of the same characteristic parameters of multi-source unkown emitter signal are fused by DSmT+PCR5, then the fusion results of each characteristic parameters are fused to get the final recognition results. If there are only single-source unknown signal characteristic parameters, the basic belief assignments of each characteristic parameter are fused by DSmT+PCR5 to get the final recognition results. Finally, through the simulation experiments in multiple conditions, the superiority of the proposed method is testified well.