Application of Differential Equations Based Physical Laws in Genetic Algorithms
2026 10th International Artificial Intelligence and Data Processing Symposium (IDAP), Malatya, Türkiye, 5 - 06 Eylül 2026, ss.1-5, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/idap71869.2026.11728215
- Basıldığı Şehir: Malatya
- Basıldığı Ülke: Türkiye
- Sayfa Sayıları: ss.1-5
- İnönü Üniversitesi Adresli: Evet
Özet
This study proposes a novel hybrid algorithmic methodology capable of performing parameter optimization through the hybridization of the Genetic Algorithm (GA), leveraging the balance and empirically proven mathematical formulations of physical laws. Within this scope, a Physical Laws (PL) pool comprising three distinct fundamental formulations Noether’s Theorem (N), the Law of Conservation of Energy (E), and the Law of Universal Gravitation (G) was established. The variables and values within the differential formulations of these laws were derived from the state updates of the GA, which served as the core implementation algorithm. Through this proposed framework, GA parameters are adaptively recalculated at each iteration using the differential equations representing these scientific laws. The proposed methodology was evaluated on 41 standard benchmark functions and compared with the classical GA in terms of solution quality, convergence speed, and generalizability. Consequently, a modular, flexible, and self optimizing hybrid model framework adaptable to various real world problems was achieved. By integrating the mathematical structure of differential physical laws into the metaheuristic optimization process, this study aims to contribute a novel perspective to the literature and provide a paradigm with a significantly expanded search space.