摘要
介绍了以Takens的嵌入定理为基础的状态空间重构混沌信号处理方法,为了降低特征维数、减小特征成分的相关性、突出特征的差异性,同时引入了K-L(Karhunen-Loeve)变换。利用舰船噪声信号的混沌特性,提出了一种基于状态空间重构与K-L变换相结合的特征提取方法,并用此方法对舰船噪声信号进行特征压缩。计算机仿真结果表明该方法具有较好的分类效果。
In this paper, the method of chaotic signal processing based on embedding theorem proposed by Takens is illuminated, and K-L (Karhunen-Loeve) transform is introduced to reduce the feature dimensions, decrease the correlation of feature components and increase their difference. Based on the technique of state space reconstruction and K-L transform, a novel method of extracting and reducing the features from ship noise is proposed. The experimental results obtained by using real data show that the method is effective.
出处
《系统仿真学报》
CAS
CSCD
2003年第8期1079-1080,共2页
Journal of System Simulation
关键词
状态空间重构
K-L变换
特征提取
舰船噪声
state space reconstruction
K-L transform
feature extraction
ship noise
