中文摘要：

鉴于有监督的Kohonen神经网络在雷达信号识别方面的不足,将S-Kohonen神经网络和最小风险贝叶斯决策相结合,提出了一种加强的S-Kohonen-Bayes方法对雷达信号进行分类,并利用Adaboost强分类器设计对识别结果进行修正.人工仿真实验结果表明,错误率平均降低了36%,改进方法具有良好的识别能力,使用最小风险贝叶斯决策进行修正是有效和必要的.

英文摘要：

In view of the lack of supervised Kohonen neural network in radar signal recognition area. By combining the S-Kohonen neural network with the minimum risk Bayes decision, proposed a strengthening of the S-Kohonen - Bayes approach radar signal classification, and used Adaboost strong classifier design tocorrect the recognition result. Artificial simulation results showed that the error rate reduced by an average of 36%, the improved method of has a good ability to identify, it is validity and necessity to use the minimum risk Bayes decision correcting.