Akademik Veri Yönetim Sistemi
6th International Congress on Engineering and Life Science (ICELIS), Girne, Kıbrıs (Kktc), 2 - 04 Eylül 2025, ss.614-625, (Tam Metin Bildiri)
Metaheuristic algorithms are widely used to solve geophysical inversion problems that are nonlinear and non-unique in nature. Unlike conventional optimization methods, which usually require good starting models and can get stuck in local minima, metaheuristic techniques can effectively search large solution spaces to find global optimum without relying on initial assumptions. However, different metaheuristics may have limitations in terms of convergence speed, or exploration and exploitation abilities. To address these issues, hybrid metaheuristic algorithms have been proposed to combine the advantages of individual metaheuristics while minimizing their drawbacks. In this study, we present the implementation and testing of a hybrid optimization algorithm that combines Differential Evolution (DE) and Particle Swarm Optimization (PSO). The hybrid approach (DE/PSO) operates in a teamwork structure, allowing both algorithms to run independently while sharing information to enhance their search processes. The algorithm was developed in the R programming environment using open-source packages to ensure reproducibility and accessibility for researchers. As a case study, the hybrid algorithm is tested by inverting self-potential (SP) data. The algorithm was first tested on synthetic SP data over a simple geometric model. The method was then applied to a well-studied benchmark field dataset acquired over a Malachite mine in Colorado (USA) to evaluate its real-world performance. The results show that the DE/PSO hybrid algorithm obtains accurate parameter estimates with minimal misfit values while ensuring fast convergence. This research demonstrates that hybrid metaheuristic methods offer efficient and reliable solutions for geophysical data inversion and have the potential for wider application in geophysical exploration.