基于主成分分析去噪的非接触心电测量

发布时间：2018-03-05 15:12

本文选题：耦合电容　切入点：主成分分析　出处：《燕山大学》2012年硕士论文　论文类型：学位论文

【摘要】：长时间不间断的心电(ECG)监测对心脏突发疾病的预防有着重要的意义，非接触测量是一种日常生活中在不影响人正常活动下对人体心电信号进行监测的有效方法，其中利用耦合电容原理是设计非接触心电测量系统的方法之一。由于测量电极与人体不是固定在一起的，所以人微小的动作就会使耦合电容两极板之间的距离或人体皮肤上被测量的位置发生改变，进而引起采集的心电信号发生变化，，导致在信号处理后经常出现滤波效果不理想或心电信号被削弱而失真的情况。因此需要一种能够与非接触心电测量匹配的滤波方法。本文根据主成分分析(Principal component analysis，PCA)方法的基础理论及其应用，从能量的角度出发，提出了一种快速自适应PCA去噪的算法，为实现对人体心电信号的长期监测提供了一种新的滤波方法。主要的研究内容与方法如下：利用耦合电容原理设计了一种非接触测量系统，采集的心电信号首先在硬件电路中进行滤波(主要是50Hz工频干扰)和放大，然后经A/D转换由模拟信号变为数字信号。总结分析了PCA方法的基础理论，以此为依据确定了PCA去噪的算法：首先用动态嵌入(Dynamical embedding，DE)技术得出一个包含采集信号所有信息的嵌入矩阵，然后提取信号的主成分，再根据Beyesian信息准则确定所需的主成分的个数，最后将所选择的主成分进行线性组合。以不同的信号作为噪声，对PCA去噪算法做了仿真实验，并对其幅值特性做了进一步的分析，证明了这种滤波方法的准确性。用非接触测量的方式采集人体的心电信号，并用PCA去噪的算法对其进行处理，通过实验证明了这种滤波方法能够在保留信号主要特征的前提下可以将干扰信号一次性去除，即使输入信号在发生变化时也能保证滤波效果的稳定性，能很好地与非接触测量的方式结合在一起。
[Abstract]:Continuous monitoring of ECG for a long time is of great significance to the prevention of sudden heart disease. Non-contact measurement is an effective method to monitor human ECG signal in daily life without affecting human normal activities. The principle of coupling capacitance is one of the methods to design non-contact ECG measurement system. Since the measuring electrode is not fixed to the human body, So a tiny human action changes the distance between the two poles of the coupling capacitance or the measured position on the human skin, which in turn causes changes in the ECG signals collected. As a result, the filtering effect is not ideal or the ECG signal is weakened and distorted after signal processing, so we need a filtering method that can match the contactless ECG measurement. Based on the basic theory and application of principal component analysis (PCA) method for Principal component Analysis (PCA), a fast adaptive PCA denoising algorithm is proposed from the point of view of energy. This paper provides a new filtering method for long-term monitoring of human ECG signal. The main research contents and methods are as follows:. A non-contact measurement system is designed based on the principle of coupling capacitance. The collected ECG signals are filtered (mainly 50Hz power frequency interference) and amplified in the hardware circuit, and then converted from analog signals to digital signals by A / D conversion. The basic theory of PCA method is summarized and analyzed, and the algorithm of PCA denoising is determined. Firstly, a embedding matrix containing all the information collected from the signal is obtained by using dynamic embedding technique, and then the principal components of the signal are extracted. Then the number of principal components is determined according to the Beyesian information criterion. Finally, the selected principal components are linearly combined. Using different signals as noise, the PCA denoising algorithm is simulated and its amplitude characteristics are further analyzed. The accuracy of this filtering method is proved. The ECG signal of human body is collected by non-contact measurement and processed by PCA denoising algorithm. The experiment proves that this filtering method can remove the interference signal at one time on the premise of retaining the main characteristics of the signal. Even when the input signal changes, it can ensure the stability of the filtering effect, and can be well combined with the non-contact measurement method.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R318.0;TN911.23

【参考文献】