Research And Implementation Of ECG Signal Denoising Method

Cardiovascular disease is harmful to human health,and wearable devices can be used to monitor heart health in real time.Fabric electrodes are commonly used as conductive electrodes for obtaining ECG signals in wearable devices because of their softness and long-term wear characteristics.However,the fabric electrode is in contact with the surface of the skin without a medium such as a gel,and thus is more susceptible to noise interference than the wet electrode.The motion artifact is one of the interference noise,which is caused by the relative motion between the electrode and the skin or the deformation of the skin,and the removal of the motion artifact is the current measurement of the ECG signal by the fabric electrode under dynamic conditions.A major challenge.This thesis mainly studies and explores the method of removing motion artifacts.The main work and innovations include:(1)Design and implementation of electrode structure based on removing motion artifacts.The skin potential variation(SPV)is highly correlated with the motion artifacts of the ECG signal,and can be used as a reference signal for adaptive filtering to remove motion artifacts in the ECG signal.In order to obtain the SPV signal,this thesisinnovatively proposes an electrode structure using a transpolar resistor to connect a pair of fabric electrodes.The fabric electrode-skin interface equivalent circuit was established,and the data formula of ECG signal and SPV signal source was derived and analyzed according to the proposed structure.The equivalent circuit simulation was carried out to optimize the resistance of the transpolar resistor and to prepare the chest strap for suppressing motion artifact.Finally,the experimental results of the optimal range of the transpole resistance are verified.(2)Research and implementation of adaptive cancellation algorithm based on SPV signal.In this thesis,the adaptive cancellation algorithm based on SPV signal is used to remove motion artifacts.The improved adaptive algorithm is derived by introducing variable step size and symbol function.By constructing simulated ECG data,combining SPV signal for simulation verification,finding the optimal adaptive filtering parameters,and performing simulation experiments.The results show that the improved adaptive algorithm is better than the basic LMS algorithm in improving the signal-to-noise ratio before and after filtering and the convergence of the algorithm.(3)Experimental circuit design and test verification.In this thesis,the experimental circuit for synchronous acquisition of ECG signals and SPV signals is designed as the experimental platform for the proposed method of removing motion artifacts.The effectiveness test of SPV signals and the performance verification of SPV-based adaptive algorithms are carried out by comparing experiments with acceleration signals.The test results show that the SPV signal is used as the adaptive reference signal,and the motion artifact is better than the acceleration signal.The proposed symbol normalized LMS algorithm is better than the LMS algorithm in performance.In this thesis,the skin potential change signal is used as the reference signal,and the method of combining the symbol normalized LMS algorithm achieves the purpose of suppressing or eliminating the motion artifacts in the ECG signal.