Multimode Decomposition and Wavelet Threshold Denoising of Mold Level Based on Mutual Information Entropy

Abstract

The continuous casting process is a continuous, complex phase transition process. The noise components of the continuous casting process are complex, the model is difficult to establish, and it is difficult to separate the noise and clear signals effectively. Owing to these demerits, a hybrid algorithm combining Variational Mode Decomposition (VMD) and Wavelet Threshold denoising (WTD) is proposed, which involves multiscale resolution and adaptive features. First of all, the original signal is decomposed into several Intrinsic Mode Functions (IMFs) by Empirical Mode Decomposition (EMD), and the model parameter K of the VMD is obtained by analyzing the EMD results. Then, the original signal is decomposed by VMD based on the number of IMFs K, and the Mutual Information Entropy (MIE) between IMFs is calculated to identify the noise dominant component and the information dominant component. Next, the noise dominant component is denoised by WTD. Finally, the denoised noise dominant component and all information dominant components are reconstructed to obtain the denoised signal. In this paper, a comprehensive comparative analysis of EMD, Ensemble Empirical Mode Decomposition (EEMD), Complementary Empirical Mode Decomposition (CEEMD), EMD-WTD, Empirical Wavelet Transform (EWT), WTD, VMD, and VMD-WTD is carried out, and the denoising performance of the various methods is evaluated from four perspectives. The experimental results show that the hybrid algorithm proposed in this paper has a better denoising effect than traditional methods and can effectively separate noise and clear signals. The proposed denoising algorithm is shown to be able to effectively recognize different cast speeds.

Adaptive Signal Processing Algorithms Based on EMD and ITD

“…Based on the EMD algorithm, the researchers proposed a number of denoising algorithms: Ensemble EMD (EEMD) [15], Conventional EMD (CEMD) [16], Iterative EMD (IEMD) [16], Exponential EMD (EXP-EMD) [17] and their various modifications [18]. All algorithms use so called modes in their work, so we could modify them as follows: keeping the structure of the algorithms, we modify them with using ITD (instead of the modes obtained using EMD, we will use the modes obtained using ITD).…”

An Improved VMD-Based Denoising Method for Time Domain Load Signal Combining Wavelet with Singular Spectrum Analysis

“…For example, Liu et al [26] determined the mode number by detrend fluctuation analysis (DFA) through a large number of simulation experiments and finally achieved a good denoising performance. Lei et al [27] proposed a hybrid algorithm combining VMD and WTD and selected the mode number for VMD through the decomposition results of EMD. In addition, the proposed denoising algorithm is shown to be able to effectively recognize different cast speeds.…”

“…However, pure VMD denoising can not only remove the high-frequency noise but also reduce the effective high-frequency information. Some scholars have proposed to use the wavelet threshold method to remove noise components from high-frequency components and reconstruct the signal [27]. However, WTD cannot completely remove the high-frequency noise, and with the change of operation conditions, the low-frequency signal will be adulterated with low-frequency interharmonics [33], so the performance of the abovementioned denoising method is not ideal.…”

Improving the prediction accuracy of river inflow using two data pre-processing techniques coupled with data-driven model

“…However, there is a need for developing new hybrid approaches with efficient decomposition methods that predict the non-linear, high irregular and noise-corrupted data with high precision. Several new data pre-processing approaches have been proposed and VMD is commonly used because of its efficient mathematical sound and more precise multi-scale components separation (Ali, Khan & Rehman, 2018; Wu & Lin, 2019; Lei, Su & Hu, 2019). VMD, as a data decomposition method, has been applied in the field of signal processing and wind speed prediction (Liu, Mi & Li, 2018; Lei, Su & Hu, 2019).…”

“…Several new data pre-processing approaches have been proposed and VMD is commonly used because of its efficient mathematical sound and more precise multi-scale components separation (Ali, Khan & Rehman, 2018; Wu & Lin, 2019; Lei, Su & Hu, 2019). VMD, as a data decomposition method, has been applied in the field of signal processing and wind speed prediction (Liu, Mi & Li, 2018; Lei, Su & Hu, 2019). Rezaie-Balf et al (2019a) proposed a new hybrid model comprised on Variational Mode Decomposition based ELM (VMD-ELM) to forecast short-term water demand.…”