基于粗集理论的雷达辐射源信号识别

将粗集理论（RST）引入到雷达辐射源信号（RES）识别中，提出一种区间连续属性离散化新方法及相应的特征选择算法，将RST与神经网络（NN）结合，设计粗集神经网络（RNN）分类器．实验结果表明，该方法解决了已有方法难以处理的区间连续属性离散化问题，获得的正确识别率比其他3种方法分别高出7．29％、4．34％和4．00％．RNN的平均训练代数比NN少97．54，RNN的平均识别率比NN高2．84％，这表明RNN具有比NN更好的分类能力和泛化能力，从而证实了该方法的有效性和可行性．

Radar Emitter Signal Recognition Based on Rough Set Theory

Rough set theory (RST) was introduced into radar emitter signal recognition. A novel approach was proposed to discretize interval-valued continuous attributes, and the corresponding feature selection method was presented. Rough set neural network (RNN) classifier was designed by combining RST and neural network (NN). Experimental results show that the proposed approach solves the problem of interval-valued continuous attribute discretization existing methods are unable to deal with, and achieves higher 7.29%, 4.34% and 4.00% recognition rate than that of the other methods. The average training generations of RNN are 97.54 less than that of NN and the average recognition rate of RNN is higher 2.84% than that of NN, which indicates that RNN has stronger capabilities of classification and generalization than NN to be expectantly applied to the practice.