Developing a hybrid CEEMDAN-PE-HE-SWT method to remove the noise of measured carbon dioxide blast wave

Abstract

Monitoring the carbon dioxide blast wave is an efficient method for determining the power of carbon dioxide blasting. Due to the harsh measurement environment, noise is inevitably mixed into the recorded signals. This work proposed a hybrid denoising strategy based on complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), permutation entropy (PE), Hurst exponent (HE), and stationary wavelet transform (SWT). Firstly, the measured signal is decomposed by CEEMDAN, and then multiple intrinsic mode functions (IMFs) are divided into pure, mixed, and noisy by PE and HE. The noise in the mixed IMFs is removed by SWT. Finally, the blast wave is reconstructed through the pure and denoised mixed IMFs. The denoising performance of the proposed approach was compared with that of the other six approaches. The results indicate that the proposed approach has better performance and could be regarded as an effective denoising method for carbon dioxide blast waves.