摘要
基于模态相关的准则,提出了一种新的信号去噪方法。首先对原信号进行经验模态分解(EMD),其次根据各阶固有模态与原信号的相关系数曲线中第一个极小值点的位置作为标准定出噪声主导性模态与信号主导性模态的分界点,最后将固有模态函数(IMF)中信号起主导作用的模态成分进行重构来实现去噪。将该方法应用于仿真信号与实际光辐射强度信号的去噪,结果表明该方法在有效去噪的同时能较好地保持原信号的细节信息,具有较好的推广性。
Based on the criterion of mode correlation, a new approach for signal de-noising based on empirical mode decomposition （EMD） was proposed. The approach can reconstruct de-noised signal by combing the signal dominant intrinsic mode functions （IMFs） derived from EMD. The dividing point between signal dominant modes and noise dominant modes was computed adaptively according to the relevance of IMFs and signal. Simulations were conducted for simulated signals and a real ray radiation signal using this approach. Experimental results indicate that this approach can efficiently remove the noise without affection by parameters and this approach can be extended to general use.
出处
《西华大学学报（自然科学版）》
CAS
2009年第6期20-24,共5页
Journal of Xihua University:Natural Science Edition
基金
国防科技重点实验室基金(9140C2601090901)
关键词
经验模态分解
模态相关
自适应
信号去噪
empirical mode decomposition
mode relevance
self-adapting
signal de-noising
