ENHANCED GIANT TREVALLY OPTIMIZER FOR ENGINEERING DESIGN AND EPIDEMIOLOGICAL MODEL

Metaheuristic algorithms are widely used for solving complex optimization problems, but their performance often depends on the initialization strategy. This study proposes an enhanced Giant Trevally Optimizer (GTO) by introducing quasi-random Sobol sequences in the initialization phase, yielding the Sobol-initialized Giant Trevally Optimizer (SGTO). The algorithm was tested on forty benchmark functions, five engineering design problems, and an epidemiological model case study. Experimental results show that SGTO consistently outperforms the original GTO in terms of achieving optimal solutions, convergence, and its ability to maintain a consistent solution across multiple independent runs. Furthermore, the epidemiological case study demonstrates the adaptability of SGTO for tackling more complex real-world problems. This work is the first to adapt Sobol sequences for the GTO and apply it to an epidemiological model. These findings confirm that quasi-random initialization substantially improves exploration and exploitation, establishing SGTO as a versatile and reliable optimization tool.