Multiscale dynamic analysis of blast furnace system based on intensive signal processing

Area of Science:

• Materials Science
• Chemical Engineering
• Complex Systems

Background:

• Blast furnace ironmaking is a complex industrial process.
• Understanding its dynamic behavior is crucial for efficiency and control.
• Previous analyses often lacked multiscale perspectives.

Purpose of the Study:

• To analyze the multiscale dynamics of silicon content in hot metal from a blast furnace.
• To identify intrinsic mode functions (IMFs) and their characteristics.
• To investigate the coupling and interdependence of subscale structures within the blast furnace system.

Main Methods:

• Application of Hilbert-Huang transform (HHT) and time delay embedding.
• Decomposition of silicon content time series into ten intrinsic mode functions (IMFs).
• Calculation of fractal dimension, Kolmogorov entropy, and Lyapunov exponent for IMF components.
• Analysis of system coupling using dimension of interaction dynamics and interdependence detection algorithms.

Main Results:

• Clear evidence of multiscale features in the blast furnace ironmaking process.
• Identification of non-integer fractal dimension, positive finite Kolmogorov entropy, and positive finite maximum Lyapunov exponent in certain IMFs.
• IMF(3) identified as the main driver in the IMF(2)-IMF(3) coupling system.
• No single subsystem identified as the driver in the IMF(3)-IMF(4) coupling system.

Conclusions:

• The study provides a multiscale theoretical framework for analyzing blast furnace ironmaking.
• Findings offer insights into the complex interdependencies within the blast furnace system.
• The methods used may lead to advanced tools for modeling and controlling blast furnace operations.