Comparative analysis of the DWT-based denoising technique selection in noise-riding DCV of the Li-Ion battery pack

The noise-riding discharging/charging voltage (DCV) of the Li-Ion battery pack may result in erroneous equivalent circuit model (ECM)-based state-of-charge (SOC) estimation and state-of-health (SOH) prediction of the Li-Ion battery pack. Namely, the noisy DCV is intimately linked with a low battery management system (BMS) performance. Therefore, additional technique should be absolutely required for noise reduction of the DCV. This work presented an implementation of the discrete wavelet transform (DWT)-based denoising technique that enables us to provide the de-noised DCV. Specifically, this work develops prior investigation one step further by performing a comparative analysis of the DWT-based denoising technique that uses different thresholding methods such as hard- and soft-thresholding. With regard to the denoising technique using hard-thresholding that retains a small number of coefficients, it is inevitable that the results are often smoothed at the expense of loosing information of the DCV. On the other hand, it is possible to acquire an improved noise-reduction of the DCV through the denoising technique using soft-thresholding retaining a large number of coefficients. Finally, selection of the proper thresholding method is a significant task in determining the signal-to-noise ratio (SNR), and analytic results indicate the clear comparison by showing the SNR difference between two denoising techniques considering hard- and soft-thresholding. For reference, threshold value has been calculated by VisuShrink method.