Abstract: The electrocardiogram（ECG） is one of the main physiological signals of the human body, and the heart health can be understood through the analysis of the ECG. As the ECG is a weak low-frequency signal, it is very easy to be interfered by the noise from inside and outside of the human body during the collection process, which will affect the diagnosis effect of heart diseases. Baseline drift, power frequency interference and myoelectric interference are not easy to avoid in the process of ECG acquisition. In this paper, the effect of correlation denoising algorithm for ECG is compared and analyzed. Firstly, the ECG in the simulated ideal state are taken as the original data, and the baseline drift, power frequency interference and myoelectric interference existed in the ECG are simulated. Three common denoising algorithms are selected for each kind of noise interference, and the denoising effect is compared by the evaluation indexes of signal-to-noise ratio, mean square deviation and ECG frequency domain characteristics. On this basis, a method for denoising multi-noise ECG is proposed and the denoising flow and effect are given. The results show that：（1）wavelet transform method, notch filter method and wavelet threshold method respectively have the best denoising effect for baseline drift, power frequency interference and myoelectric interference. （2）When the ECG contains two or more kinds of noise, the denoising order of baseline drift, power frequency interference and myoelectric interference is the most effective.

Key words: electrocardiogram（ECG）, multi-noise, denoising effect

摘要： 心电信号是人体的主要生理信号之一，通过对心电信号的分析可了解心脏的健康状态，由于心电信号属于微弱低频信号，所以在采集过程中极易受到来自人体内部和外部的噪声干扰，影响心脏疾病诊断的效果。基线漂移、工频干扰和肌电干扰是心电信号采集过程中不能忽略的噪声干扰。对心电信号的相关去噪算法的效果进行对比分析。首先将模拟理想状态下的心电信号作为原始数据，同时模拟出心电信号中存在的基线漂移、工频干扰和肌电干扰。每种噪声干扰分别选择三种常用的去噪算法，采用信噪比、均方差和心电信号的频域特征的评估指标进行去噪效果的比较。在此基础上，提出了一种多噪声心电信号的去噪方法并给出去噪流程和效果。研究结果表明：（1）对于基线漂移、工频干扰和肌电干扰分别采用小波变换法、陷波滤波法和小波阈值法的去噪效果最好；（2）当心电信号含两种及两种以上噪声时，按照滤除基线漂移、工频干扰和肌电干扰的去噪顺序滤波效果最好。

关键词: 心电信号（ECG）, 多噪声, 去噪效果