The Strategic Random Search (SRS) – A new global optimizer for calibrating hydrological models

Abstract

This study introduces a novel global optimization algorithm, Strategic Random Search (SRS), tailored for efficient calibration of hydrological models. SRS outperforms 14 other optimization algorithms on 23 classical benchmark functions and 29 CEC-2017 benchmark functions, demonstrating its superiority on more than half of these tests. Additionally, when applied to rainfall-runoff models, SRS consistently, rapidly, and robustly converges to optimal solutions, surpassing five other algorithms. SRS, developed independently of existing intelligent optimization methods, offers versatility with only two adjustable parameters, making it suitable for various problem types. Through rigorous testing and comparisons, SRS emerges as a robust, widely applicable, and stable convergence algorithm.