An organizational evolutionary algorithm for numerical optimization

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Evolutionary Computation

Background:

• The study introduces the Organizational Evolutionary Algorithm (OEA), a novel approach inspired by social organizational structures.
• OEA models populations as organizations composed of individuals, with evolutionary operators mimicking inter-organizational interactions.
• This algorithm is designed for tackling both unconstrained and constrained optimization problems.

Discussion:

• Experimental validation involved 15 unconstrained, 13 constrained functions, and 4 engineering design problems.
• Performance was rigorously compared against existing optimization approaches.
• The OEA demonstrated competitive results in solution quality and computational cost.

Key Insights:

• The OEA achieves high performance, particularly for constrained optimization problems, by integrating simple constraint-handling techniques.
• Systematic parameter analysis revealed the algorithm's robustness and ease of use.
• The organizational structure provides an effective paradigm for evolutionary computation.

Outlook:

• Further research could explore more complex organizational structures and interaction dynamics within the OEA.
• Investigating the scalability of OEA to larger-scale and higher-dimensional optimization problems is warranted.
• Applications of OEA in diverse scientific and engineering domains could be explored.