A quasi-opposition learning and chaos local search based on walrus optimization for global optimization problems.

Area of Science:

• Computational Intelligence
• Metaheuristic Optimization
• Algorithm Development

Background:

• The Walrus Optimization (WO) algorithm, while promising, suffers from slow convergence and local optima entrapment.
• Existing metaheuristic algorithms require improvements in global search capability and convergence speed.

Purpose of the Study:

• To enhance the Walrus Optimization (WO) algorithm by integrating quasi-oppositional-based learning and chaotic local search.
• To improve global search capabilities, expand search range, and accelerate convergence.
• To address premature convergence and local optima issues in the WO algorithm.

Main Methods:

• Integration of quasi-oppositional-based learning into the WO algorithm.
• Incorporation of a chaotic local search mechanism to accelerate convergence.
• Application and comparison of the proposed QOCWO algorithm on 23 standard benchmark functions.

Main Results:

• The QOCWO algorithm demonstrated superior performance compared to seven other algorithms on standard functions.
• Statistical validation using the Wilcoxon rank-sum test confirmed the significance of QOCWO's improvements.
• QOCWO achieved lower costs in two real-world engineering design problems.

Conclusions:

• The proposed QOCWO algorithm effectively overcomes the limitations of the original WO algorithm.
• QOCWO shows significant potential for solving complex optimization problems in both theoretical and practical domains.