Subpopulation Particle Swarm Optimization with a Hybrid Mutation Strategy.

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Swarm Intelligence

Background:

• Real-world optimization problems are increasingly complex, demanding advanced algorithms.
• Standard particle swarm optimization (PSO) struggles with rapid and accurate global optimum discovery.
• Existing PSO variants often face limitations in efficiency and precision.

Purpose of the Study:

• To introduce a novel hybrid particle swarm optimization algorithm (SHMPSO).
• To address the limitations of conventional PSO in finding global optimal solutions.
• To enhance convergence speed and solution accuracy in complex optimization tasks.

Main Methods:

• Developed SHMPSO, integrating seed swarm optimization with PSO.
• Employed a subpopulation coevolution strategy for enhanced information sharing.
• Utilized an elastic candidate-based approach and mean dimension learning for faster convergence.

Main Results:

• SHMPSO demonstrated superior convergence speed compared to benchmark PSO variants.
• The algorithm exhibited robust performance across 21 benchmark functions.
• High-precision solutions were consistently achieved by SHMPSO.

Conclusions:

• SHMPSO effectively overcomes the limitations of standard PSO.
• The hybrid strategy offers significant improvements in convergence and accuracy.
• SHMPSO is a promising tool for tackling complex large-scale optimization problems.