摘要
针对以傅立叶分析为代表的传统方法在非线性信号分析上的不足,提出了将分形理论应用于非线性信号含噪分析的方法,设计了具体实验的流程,并结合实际心电数据进行了相关的实验。实验结果表明:周期性非线性时间序列与随机噪声在分形指数上有明显差异,当随机噪声对原始信号的影响程度变化时,分形维数也随之变化;分形维数可以有效地描述工程数据被噪声污染的程度,因而具有良好的工程应用前景。
In terms of the disadvantage of traditional methods represented by Fourier analysis applied in analysis of nonlinear signals, a new approach based on fractal theory was provided to analysis random noises in nonlinear signals. Experiment results on real electrocardiograph（ECG） signals demonstrated remarkable gap between the periodical nonlinear time series and random noi- ses; moreover, fractal dimensions varied with different levels of impact from random noise on original time series. These results imply that fractal dimension can effectively describe how the data is affected by random noise and thus usher the prospect in engineering application.
出处
《机电工程》
CAS
2008年第9期4-7,共4页
Journal of Mechanical & Electrical Engineering
基金
国家自然科学基金资助项目(60574079)
关键词
分形
噪声分析
非线性
心电信号(图)
fractal
noise analysis
nonlinear
electrocardiograph （ECG）
