Multi-objective particle swarm optimization algorithm for task allocation and archived guided mutation strategies.

Area of Science:

• Artificial Intelligence
• Computational Intelligence
• Optimization Algorithms

Background:

• Traditional particle swarm optimization (PSO) algorithms face challenges with search inefficiency and biased selection mechanisms.
• Multi-objective optimization problems require sophisticated algorithms to balance competing objectives effectively.

Purpose of the Study:

• To propose a novel multi-objective particle swarm optimization algorithm (TAMOPSO) that addresses the limitations of traditional methods.
• To enhance evolutionary search efficiency and ensure fairness in the optimization process.

Main Methods:

• TAMOPSO employs a task allocation strategy, dividing subpopulations and assigning tasks based on particle distribution.
• An adaptive Lévy flight strategy dynamically adjusts global and local search based on population convergence or dispersion.
• A particle evolution contribution rate index filters valuable historical solutions, and an improved selection mechanism ensures fairness.

Main Results:

• TAMOPSO demonstrated superior performance compared to ten existing algorithms across 22 standard test problems.
• The algorithm effectively improves evolutionary search efficiency and accelerates convergence speed.
• Experimental results indicate enhanced fairness in the particle selection process.

Conclusions:

• TAMOPSO offers a significant advancement in solving multi-objective optimization problems.
• The proposed task allocation and archive-guided mutation strategies contribute to improved performance and fairness.
• The adaptive Lévy flight strategy enhances the algorithm's ability to balance global exploration and local exploitation.