基于改进的径向基函数神经网络的辐射源算法研究

多传感器数据融合系统中辐射源识别技术占有重要的位置。本文结合改进的径向基网络给出了辐射源算法的实现结构。结合辐射源预分的数据特点对径向基中高斯核函数进行了修改，使得在不对处理样本初始化的条件下仍有很好的预分效果，预分后采用模糊匹配的方法，完成辐射源的识别。

Research on emitter recognition based on improved radial basis function network

Emitter recognition is an important part in multi-sensor information fusion system. An emitter recognition method is put forward based on improved Radial Basis Function Network (RBFN). Traditional realization of RBFN needs data initialization, which increases the complexity of net study. To solve this problem, an improved RBFN is put forward in this paper, which can avoid data initialization process, keep the correct recognition ratio and get well application. With the RBFN to pre-classify the measure, fuzzy matching is also used to complete the recognition.