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

Area of Science:

• Materials Science and Engineering
• Signal Processing
• Manufacturing Processes

Background:

• Continuous casting involves complex phase transitions with challenging noise components.
• Existing models struggle with effective noise and signal separation.
• Traditional denoising methods lack sufficient adaptive features for this process.

Purpose of the Study:

• To develop a robust hybrid algorithm for effective noise reduction in continuous casting signals.
• To improve the separation of noise and clear signals compared to existing methods.
• To enhance the recognition of varying cast speeds using denoised data.

Main Methods:

• A hybrid algorithm combining Variational Mode Decomposition (VMD) and Wavelet Threshold Denoising (WTD) was developed.
• Empirical Mode Decomposition (EMD) was used to determine the VMD parameter K.
• Mutual Information Entropy (MIE) identified noise-dominant and information-dominant components for targeted denoising.

Main Results:

• The proposed hybrid VMD-WTD algorithm demonstrated superior denoising performance over traditional methods.
• The algorithm effectively separated complex noise components from clear signals.
• Experimental results confirmed the algorithm's ability to recognize different cast speeds accurately.

Conclusions:

• The hybrid VMD-WTD algorithm offers a significant advancement in signal processing for continuous casting.
• This method provides a reliable approach for noise reduction in complex industrial processes.
• The enhanced signal clarity facilitates better monitoring and control of casting operations.