Hybrid Slime Mold and Arithmetic Optimization Algorithm with Random Center Learning and Restart Mutation

Area of Science:

• Computer Science
• Artificial Intelligence
• Optimization Algorithms

Background:

• Slime mold algorithm (SMA) excels in global search but struggles with late-stage oscillations.
• Arithmetic optimization algorithm (AOA) offers strong randomness and convergence via multiplication/division operators.
• Existing algorithms face challenges in complex function optimization and finding optimal positions.

Purpose of the Study:

• To develop a hybrid optimization algorithm integrating SMA and AOA.
• To enhance global search efficiency, population diversity, and convergence accuracy.
• To address the limitations of individual algorithms in complex optimization tasks.

Main Methods:

• Integration of SMA and AOA, replacing SMA's convergence stage with AOA's operators.
• Incorporation of a random central solution strategy for improved global search and diversity.
• Implementation of restart and mutation strategies to boost convergence accuracy and late-stage optimization.

Main Results:

• The hybrid algorithm (RCLSMAOA) demonstrated effectiveness in comparative experiments on various test functions.
• Statistical tests (Wilcoxon rank sum, Friedman) confirmed the improved performance.
• The enhanced algorithm showed superior convergence accuracy and optimization ability compared to individual methods.

Conclusions:

• The proposed hybrid slime mold and arithmetic optimization algorithm (RCLSMAOA) is effective for complex optimization problems.
• The integration of random central learning, mutation, and restart strategies significantly improves performance.
• The algorithm shows promise for application in practical engineering problems.