PS-FW: A Hybrid Algorithm Based on Particle Swarm and Fireworks for Global Optimization.

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Heuristics

Background:

• Particle Swarm Optimization (PSO) and Fireworks Algorithm (FWA) are efficient but struggle with high-dimensional problems.
• PSO can get trapped in local optima due to limited global exploration.
• FWA may face convergence issues from inefficient local exploitation.

Purpose of the Study:

• To introduce a novel hybrid optimization algorithm, PS-FW.
• To enhance the performance of PSO and FWA for high-dimensional global optimization.
• To balance exploration and exploitation capabilities in optimization.

Main Methods:

• Embedding modified Fireworks Algorithm (FWA) operators into the Particle Swarm Optimization (PSO) process.
• Implementing an abandonment and supplement mechanism for exploration-exploitation balance.
• Proposing a modified explosion operator and a novel mutation operator for faster convergence and prematurity avoidance.

Main Results:

• The PS-FW algorithm demonstrated superior performance across 22 high-dimensional benchmark functions.
• Comparative analysis showed PS-FW outperforming PSO, FWA, and other advanced algorithms.
• PS-FW exhibited efficiency, robustness, and rapid convergence.

Conclusions:

• The proposed PS-FW algorithm is an effective and robust method for global optimization.
• PS-FW successfully addresses the limitations of PSO and FWA in high-dimensional spaces.
• The hybrid approach offers a significant advancement in optimization techniques.