Hybrid GA-IPA and neural network-based solution for fractional-order nonlinear predator-prey dynamical systems

Abstract

This paper proposes a novel approach for solving fractional-order nonlinear predator–prey dynamical systems using artificial neural networks (ANNs), with the goal of improving accuracy and computational efficiency in modeling complex realworld ecological interactions. A hybrid computational intelligence framework is developed, integrating ANNs with a Genetic Algorithm–Interior Point rithm (GA–IPA) for parameter optimization. The ANN serves as the modeling core, while the GA–IPA optimization refines network parameters to achieve high precision and stability. Numerical simulations confirm that the proposed ANN–GA–IPA approach effectively captures the dynamics of the fractional-order nonlinear predator–prey system. The method demonstrates strong robustness and high accuracy, outperforming conventional techniques in representing dynamic ecological behaviors. This work introduces, for the first time, the application of a hybrid ANN–GA–IPA strategy to fractional-order dynamical systems, delivering improved computational performance and modeling accuracy, offering a powerful tool for analyzing complex biological interactions and presenting potential extensions to other classes of nonlinear systems.