Developed a dynamic mutation rate control strategy using convergence and diversity metrics, which enables the mutation probability to adapt automatically to the current convergence state. The strategy increases the mutation rate when population convergence increases and diversity decreases, thereby promoting exploration and enhancing genetic diversity. When diversity is maintained at a sufficient level, the mutation rate decreases, thereby preventing unnecessary disruption of evolved solutions. Conducted experiments on benchmark functions with different fitness landscapes, including unimodal, multimodal, and deceptive functions, to validate the effectiveness of the proposed approach. Established the Composite Convergence Score (CCS) as a unified measure integrating the most informative metrics into a normalized indicator of convergence dynamics. Correlation analysis and machine learning-based evaluation confirmed that the CCS can reliably identify generations approaching stagnation, and the dynamic mutation rate strategy guided by the CCS significantly improves optimization performance compared to approaches that use a static mutation rate.
Investigated results revealed that CCS-guided mutation control consistently prevents premature convergence, maintains higher population diversity throughout the evolutionary process, accelerates convergence toward global optima, and improves the overall success rate of reaching desired solutions. The approach reduces the need for manual tuning of mutation parameters, as the CCS automatically balances exploration and exploitation. Also, the methodology allows GA frameworks to automatically update evolutionary parameters.
The research demonstrates that the CCS framework can serve as a generalizable tool for adaptive control in evolutionary computation. It provides a foundation for future development of self-adapting genetic algorithms and may be extended to multi-objective, high-dimensional, or real-world optimization problems, including engineering design, logistics, scheduling, and neural architecture search. The study confirms that integrating quantitative convergence monitoring with dynamic parameter adjustment substantially improves GA reliability, robustness, and solution quality, offering significant opportunities for advancing both theoretical and practical aspects of evolutionary optimization.
1. Katoch, S., Chauhan, S. S., & Kumar, V. (2020). A review on genetic algorithm: past, present, and future. Multimedia Tools and Applications, 80(5), 8091–8126. https://doi.org/10.1007/ s11042-020-10139-6
2. Pandey, H. M., Chaudhary, A., & Mehrotra, D. (2014). A comparative review of approaches to prevent premature convergence in GA. Applied Soft Computing, 24, 1047–1077. https://doi.org/10.1016/j.asoc.2014.08.025
3. Friedrich, T., Oliveto, P. S., Sudholt, D., & Witt, C. (2009). Analysis of Diversity-Preserving Mechanisms for Global Exploration. Evolutionary Computation, 17(4), 455–476. ttps://doi.org/10.1162/evco.2009.17.4.17401
4. Vie, A., Kleinnijenhuis, A. M., & Farmer, D. J. (2020). Qualities, challenges and future of genetic algorithms: a literature review. arXiv.org. https://arxiv.org/abs/2011.05277
5. Leung, N. Y., Gao, N. Y., & Xu, N. Z. (1997). Degree of population diversity – a perspective on premature convergence in genetic algorithms and its Markov chain analysis. IEEE Transactions on Neural Networks, 8(5), 1165–1176. https:// doi.org/ 10.1109/72.623217
6. Cheng, J., Pan, Z., Liang, H., Gao, Z., & Gao, J. (2020). Differential evolution algorithm with fitness and diversity ranking-based mutation operator. Swarm and Evolutionary Computation, 61, 100816. https://doi.org/10.1016/j.swevo.2020. 100816
7. Meng, Z., & Yang, C. (2022). Two-stage differential evolution with novel parameter control. Information Sciences, 596, 321–342. https://doi.org/10.1016/j.ins.2022.03.043
8. Kahraman, H. T., Aras, S., & Gedikli, E. (2019). Fitness-distance balance (FDB): A new selection method for meta-heuristic search algorithms. Knowledge-Based Systems, 190, 105169. https://doi.org/10.1016/j.knosys.2019.105169
9. Zhang, G., Hu, Y., Sun, J., & Zhang, W. (2020). An improved genetic algorithm for the flexible job shop scheduling problem with multiple time constraints. Swarm and Evolutionary Computation, 54, 100664. https://doi.org/10.1016/j.swevo. 2020.100664
10. Pretsel, V., & Shuvar, R. (2024). The speciation in genetic algorithms for preserving population diversity and optimization of functions with suboptimal solutions. Electronics and Information Technologies, 28. https://doi.org/10.30970/eli.28.4
11. Hussain, A., & Cheema, S. A. (2020). A new selection operator for genetic algorithms that balances between premature convergence and population diversity. Croatian Operational Research Review, 11(1), 107–119. https://doi.org/10.17535/ crorr. 2020.0009