A Novel Selection Approach for Genetic Algorithms for Global Optimization of Multimodal Continuous Functions

Area of Science:

• Computational Intelligence
• Optimization Algorithms

Background:

• Genetic algorithms (GAs) are heuristic search techniques crucial for solving constrained optimization problems.
• The performance of GAs heavily relies on their core operators, particularly chromosome selection.
• Balancing population diversity (exploration) and convergence speed (exploitation) is a key challenge in GA design.

Purpose of the Study:

• To introduce a novel stairwise selection (SWS) scheme for genetic algorithms.
• To address the inherent challenges of exploration and exploitation within genetic algorithms.
• To evaluate the effectiveness and robustness of the proposed SWS scheme against existing selection methods.

Main Methods:

• Development of the stairwise selection (SWS) scheme, an improvement over standard genetic algorithm operators.
• Comparative analysis of SWS against multiple established selection schemes using ten benchmark functions across various dimensions.
• Statistical validation including Chi-square goodness of fit test and performance index (PI) calculation.

Main Results:

• The stairwise selection (SWS) scheme effectively manages the exploration-exploitation trade-off in genetic algorithms.
• Empirical results demonstrate that SWS significantly outperforms competing selection schemes in terms of robustness, stability, and effectiveness.
• Statistical analyses and graphical representations confirm the superior performance of SWS.

Conclusions:

• The proposed stairwise selection (SWS) scheme offers a significant advancement in genetic algorithm optimization.
• SWS provides a robust and effective solution for handling complex optimization problems.
• The method's validated performance makes it a valuable contribution to the field of computational intelligence.