Efficient FCTV provision considering DWT and DWPT-based noise suppression for overcoming the noise-induced voltage loss in PEM fuel cell

Abstract

This approach gives insight to the design and implementation of the noise suppression based on the wavelet transform (WT) for efficient FCTV signal provision. The most important thing in this approach is to show the comparative analyses on noise suppression between two transforms, such as the discrete wavelet transform (DWT) and discrete wavelet packet transform (DWPT). With an identical mother wavelet of Daubechies db3, the multi-resolution analysis (MRA)-based decomposition and reconstruction processes for reducing the noise-induced voltage loss are basically done in the DWT and DWPT. The marked difference between two transforms is the MRA-based decomposition/reconstruction ability of high frequency component related to the sensing of noisy. For reference, this approach considered two noise suppression techniques such as hard- and soft-thresholdings and checked signal-to-noise ratio (SNR) values for clear evaluation of all comparative analyses. From these results, it is capable of suggesting three conclusions. First, the performance on noise suppression of the DWPT is superior to that of the DWT. Second, the noise-induced voltage loss is more suppressed at soft-thresholding technique when compared to hard-thresholding technique irrespective of the DWT and DWPT. Last, the optimal decomposition/reconstruction levels that have the maximum SNR values are respectively determined in hard- and soft-thresholding techniques. Our definite suggestions sufficiently enable us to achieve an optimal solution for efficient FCTV provision. This approach has been extensively verified by experimental results of the FCTV using a single cell.