经验模态分解对瞬态电化学响应的滤波处理

用经验模态分解(EMD)对恒电量瞬态响应信号进行滤波处理. 通过计算机仿真实验评估EMD的滤波效果. 为了提高EMD滤波的实用性和减少滤波过程的主观性, 仿真实验主要针对混迭了白噪声、高频噪声和市电干扰以及上述混合噪声的恒电量瞬态响应. 在此基础上, 阐述了EMD滤波对多个时间常数恒电量瞬态响应的重要性, 并利用电化学阻抗谱结果验证了EMD滤波后的恒电量频谱. 结果表明, 经过EMD滤波的恒电量频谱和EIS有很好的相关性. 恒电量瞬态响应的EMD滤波, 采用前两项内在模函数(IMF)和IMF4的50 Hz成分置零的固定模式, 可以获得满意的滤波结果.

Filtering of Electrochemical Transients by Empirical Mode Decomposition Method

A filtering method based on empirical mode decomposition (EMD) for coulostatically-induced transients (CITs) was presented. The effect of EMD filtering was evaluated from different simulation systems. To boost the applicable solution for EMD filtering and reduce the subjectivity of filtering, discussions of the simulating noises were mainly focused on white noise, high frequency noise, 50 Hz noise from commercial power and blend of above-mentioned noises. To illustrate the most importance of EMD filtering for the goodness of fit of CIS, one real corrosion system involving multi-time constant was simultaneously investigated by electrochemical impedance spectrum (EIS) and coulostatically-induced spectrum (CIS). The results reported here demonstrate that EMD de-noising is applicable for CITs. The proposed method shows good performance of de-noising and is an alternative de-noising method for the CITs analysis. CIS with EMD filtering are in good agreement with EIS. For practical application of EMD, removal of the first two items of intrinsic mode function (IMF) and 50 Hz element in forth IMF will be generally satisfactory for CIS analysis.