摘要
指出了非线性动态信号参数辨识的重要性;分析了目前采用的方法的不足;对非正交复Morlet小波满足Mercy条件和再生性的命题作了证明;用复Morlet小波构建出一种核函数,与主分量分析方法相结合,对非线性动态信号进行参数辨识和预测;仿真结果验证了该方法的正确性和有效性,表明该方法具有较好的理论价值和实用价值。
The important value of non-linear dynamical signal parameters identificaton was pointed. The shortcomings of current methods were analyzed, It was provedthat non-orthogonal complex Morlet wavelet could satisfy Mercy Condition and have reproduction character in Hilbert Space. A kind of special kernel function was built, which was named Non-orthogonal Complex Morlet Wavelet Kernel Function, combined with Principal Component Analysis （PCA）, and identified parameters and forcasted future information of non-linear dynamic signal. Contrast experiment results show that this kind of kernel function seems to be the most promising one and has some more applied value in this area.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2006年第9期2550-2554,共5页
Journal of System Simulation
基金
国家自然科学基金(10576027)
博士点基金(20040613013)
关键词
非正交复Morlet小波
主分量分析
核函数方法
非线性动态信号
参数辨识
Non-orthogonal Complex Morlet Wavelet
Principal Component Analysis （PCA）
Kernel Function Method
Nonlinear Dynamical Signal
Parameter Identification
