Figure 2
The diagram illustrates the process of discrete wavelet transform (DWT) cleaning of a signal. The original signal (S) is split into a clean trend (cA) and noise (cD). The signal is decomposed into multiple levels: S is decomposed into cA1 and cD1, cA1 is further decomposed into cA2 and cD2, and cA2 is decomposed into cA3 and cD3. The data reconstruction process involves combining cA3, cD3, cD2, and cD1 to reconstruct the original signal S.DWT cleaning of the signal. The original signal (S) is split into a clean trend (cA) and a noise (cD). The model uses a rebuilt signal after the noise is removed [13]