摘要
针对双稳随机共振在系统参数选取上的盲目性以及在复杂信号处理上的局限性,提出基于快速独立分量分析(FastICA)的自适应双稳随机共振方法。该方法首先对一组信号进行FastICA处理,然后对得到的源信号进行相关分析并进行信号重组,最后把重组后的信号输入自适应随机共振系统中,达到提取微弱信号特征的目的。实验证明,该方法能有效快速的获取混杂在强噪声信号中的微弱信号的特征。
Depending on the blindness of bistable stochastic resonance on system parameters selection and limitations on the complex signal processing,a FastICA adoptive bistable stochastic resonance method was put forward in the paper.First,this method uses the algorithm of FastICA to process a set of signals and get a series of source signals.Then the source signals were restructured based on the correlation analysis.Finally,the restructuring signals were input the system of adoptive stochastic resonance to get the weak signal feature extracting.This method can efficiently obtain the characteristics of weak signal mixed in strong noise signal though an experiment.
出处
《中国农机化学报》
2016年第5期54-57,65,共5页
Journal of Chinese Agricultural Mechanization
基金
山东省自然基金项目(ZR2013FM005)
山东省高等学校科技计划项目(J10LG22)
关键词
快速独立分量分析
随机共振
微弱信号
特征提取
fast independent component analysis
stochastic resonance
weak signal
feature extraction
