Abstract
Developing solutions for complex optimization problems is fundamental to progress in many scientific and engineering disciplines. The Flower Fertilization Optimization (FFO) algorithm, a powerful metaheuristic inspired by the reproductive processes of flowering plants, is one such method. Nevertheless, FFO's effectiveness can be hampered by a decline in population diversity during the search process, which increases the risk of the algorithm stagnating in local optima. To address this shortcoming, this work proposes an improved method called Crisscross Flower Fertilization Optimization (CCFFO). It enhances the FFO framework by incorporating a crisscross (CC) operator, a mechanism that facilitates a structured exchange of information between different solutions. By doing so, CCFFO effectively boosts population diversity and improves its capacity to avoid local optima. Rigorous testing on the challenging CEC2017 benchmark suite confirms CCFFO's superiority; it achieved the top overall rank when compared against ten state-of-the-art algorithms. Furthermore, its practical effectiveness is demonstrated on a complex reservoir production optimization problem, where CCFFO secured a higher Net Present Value (NPV) than its competitors. These results highlight CCFFO's potential as a powerful and versatile tool for solving complex, real-world optimization tasks.