PARTICLE SWARM OPTIMIZATION-BASED DETERMINATION OF HYDRAULIC JUMP LOCATION IN SLUICE GATE FLOWS

The hydraulic jump is a critical phenomenon in open channel hydraulics, and understanding its behavior is essential for the design and safety of hydraulic structures. In this study, 96 experiments were conducted using five different gate openings to model the location of hydraulic jumps in an open channel. The Particle Swarm Optimization (PSO) algorithm, a metaheuristic optimization technique, was employed to develop both linear and nonlinear predictive models. Experimental data from gate openings (e) of 2.5 cm, 3.5 cm, 4 cm, and 5 cm were used to train the models, while data from a e=6 cm gate opening were used for testing. The results demonstrated that the PSO algorithm effectively modeled the hydraulic jump location, yielding high accuracy and consistency with experimental observations. Model performance was evaluated using the Coefficient of Determination (R²), Nash-Sutcliffe Efficiency (NSE), and Mean Squared Error (MSE). The linear model outperformed the nonlinear model, achieving NSE = 0.954, R² = 0.983, and MSE = 0.022. Furthermore, the upstream total head (H) and gate opening (e) were identified as the most influential parameters affecting the hydraulic jump location.