摘要
提出了一种用于船舶噪声分类的局域自适应子波神经网络分类方法。首先利用傅里叶变换对三类船舶噪声进行预处理，然后利用网络局域化构造局域自适应子波神经网络分类器。通过对实际的三类处理后的船舶噪声谱进行自动特征提取并分类，分类结果令人满意，证明了该方法的优越性和工程应用前景。
In this paper, an efficient engineering classification of ship noises based on a local adaptive wavelet neural network is presented.First, the fourier transform preprocessing for three types of noises radiated from ships is necessary, then a local adaptive wavelet neural network classifier is designed by using local network theory. The classifier is used to extract automatically feature from actual ship signals and classify them after preprocessing.The classified results are encouraging, and the method is proved to be superior and efficient in the engineering application in the future.
出处
《系统工程与电子技术》
EI
CSCD
1998年第6期21-25,共5页
Systems Engineering and Electronics
基金
国家"863"基金
关键词
船舶
噪声场
特征选择
分类
神经网络
Adaptive wavelet neural netwok,Feature extraction,Classification.
