一种人工神经网络方法在被动声纳目标识别中的应用研究

A Method of Artificial Neural Network Applied in Passive Sonar Target Recognition

水下目标识别是水声信号处理领域中的一个重要课题．本文针对水下目标辐射噪声类别特征不明显的特点，提出了一种有监督的模糊ＡＲＴ神经网络（ＳＦＡＲＴ）分类器，通过对声纳目标信号的分类试验表明：该网络分类器具有迭代次数少，识别率高的优点，是一种很有实际应用价值的分类器模型．

Underwater target recognition (UTR) is a very important topic of acoustic signal processing.In this paper,a supervised fuzzy ART neural network classifier is presented.The experimental results show that the classifier based on SFART is a very practical and useful classifier model.