盲源分离算法在大地电磁信号去噪中的应用

摘要

鉴于大地电磁信号的特点,大地电磁信号的去噪问题一直是研究的热点问题,而盲源分离技术是进行现代信号处理的有力工具.阐述了大地电磁信号的特点和它的噪声分类,利用盲源分离算法维持信号频率不变性的特点,并结合大地电磁信号的频谱特征,提出利用盲源分离算法来对大地电磁信号进行噪声分析和噪声去除.实验证明,无论是直接应用盲源分离算法来对大地电磁信号进行噪声去除,还是应用盲源分离对噪声进行分析后结合其他方法进行去噪处理,都能有效地去除噪声,减少脉冲干扰地影响,提高信号的整体平滑度.%In view of characteristics of magnetotelluric signal,the denoising problem of magnetotelluric signal has always been a research hotspot,while the blind source separation technology is a powerful tool for modern signal processing.It ex pounds the characteristics of magnetotelluric signal and the classification of its noise,and then by using blind source separation algorithm to maintain the characteristics of the signal frequency invariance.Combining with the feature of magnetotelluric signal spectrum,it proposed a viewpoint that we can use blind source separation algorithm to do noise analysis and noise removal for magnetotelluric signal.Experiments show that either directly using blind source separation algorithm to the noise removal for magnetotelluric signal,or applying blind source separation to do analysis of noise in combination with other methods to deal with the denoising,can effectively remove noise,reduce the influence of pulse interference,and enhance the overall smoothness of the signal.