D2MOPSO: MOPSO based on decomposition and dominance with archiving using crowding distance in objective and solution...

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Evolutionary Computation

Background:

• Multi-objective optimization problems (MOPs) present challenges in balancing competing objectives.
• Decomposition simplifies MOPs by breaking them into smaller aggregation problems.
• Dominance is crucial for managing archives in multi-objective optimization.

Purpose of the Study:

• To improve the D2MOPSO algorithm by refining its archiving technique.
• To enhance diversity and coverage in both objective and solution spaces.
• To evaluate the improved D2MOPSO's performance on standard MOP benchmarks.

Main Methods:

• The study improves the D2MOPSO algorithm, integrating dominance and decomposition.
• A novel archiving technique is introduced to enhance solution quality and distribution.
• Performance is assessed using constrained and unconstrained benchmark test problems.

Main Results:

• The improved D2MOPSO shows superior or competitive performance compared to MOEA/D, OMOPSO, and dMOPSO.
• Statistical tests confirm the algorithm's effectiveness and efficiency.
• The method achieves better diversity and coverage in objective and solution spaces.

Conclusions:

• The enhanced D2MOPSO algorithm is highly competitive and efficient for multi-objective optimization.
• The proposed improvements lead to better performance on a wide range of MOPs.
• The algorithm is applicable to both constrained and unconstrained optimization tasks.