论文部分内容阅读
水下目标信号的特征提取和分类是当今水声信号处理领域中存在的难题。随着人工神经网络技术的发展 ,众多的研究人员已致力于将人工神经网络应用于水下目标分类的研究中。本文给出了 FKCN( fuzzy Kohonen clustering network)算法 ,并将 FKCN应用于水下目标的分类问题中。实录海上无源声纳目标信号的分类实验验证了该算法的可行性。实验结果表明 :在大量的训练样本和测试样本下 ,FKCN提高了判别的灵活性 ,增加了判别的可信度 ,使系统的整体识别率提高约 2个百分点 ,较 KCN具有更好的分类效果。
The feature extraction and classification of underwater target signal is a difficult problem in the field of underwater acoustic signal processing. With the development of artificial neural network technology, many researchers have devoted themselves to the application of artificial neural network in underwater target classification. In this paper, a fuzzy Kohonen clustering network (FKCN) algorithm is proposed and FKCN is applied to the classification of underwater targets. Experiments on classification of passive target sonar at sea show that the algorithm is feasible. Experimental results show that under a large number of training samples and test samples, FKCN improves discriminative flexibility and increases the credibility of the discriminant, which improves the overall recognition rate of the system by about 2%, which is better than that of KCN .