Explicit Evolutionary Framework With Multitasking Feature Fusion for Optimizing Operational Parameters in Aluminum...

Area of Science:

• Materials Science
• Chemical Engineering
• Computational Intelligence

Background:

• Aluminum electrolysis cells (AECs) require operational parameter optimization to reduce energy consumption.
• Dynamic heterogeneity of AECs presents challenges for traditional optimization methods.
• Existing explicit evolutionary multitasking (EMT) algorithms lack theoretical grounding for AECs and often ignore intertask correlations.

Purpose of the Study:

• To propose an explicit evolutionary framework with multitasking feature fusion (EMFF) for optimizing AEC operational parameters.
• To address the limitations of existing EMT algorithms by considering intertask feature information correlations.
• To enhance knowledge transfer and synergistic effects in multitasking optimization for AECs.

Main Methods:

• Development of the explicit evolutionary framework with multitasking feature fusion (EMFF).
• Introduction of a multitasking feature fusion mechanism for effective knowledge transfer.
• Implementation of a transfer individual derivation (TID) strategy for rapid evolution of critical knowledge.

Main Results:

• EMFF demonstrates superior performance in benchmark tests.
• The framework shows exceptional results in real-world AEC parameter optimization.
• Effective knowledge transfer and synergistic effects were achieved through the proposed mechanisms.

Conclusions:

• The proposed EMFF framework effectively optimizes AEC operational parameters, leading to reduced energy consumption.
• EMFF overcomes limitations of existing EMT algorithms by incorporating intertask feature correlations.
• The framework offers a promising approach for collaborative optimization in industrial processes.