Dams, wastewater treatment plants, and aquaculture areas often require aeration, making it crucial for hydraulic engineers to predict the physical properties of scour holes occurring on the downstream side of conduits. Conduit systems that provide high aeration performance necessitate multiple optimization studies to simultaneously maximize the scour parameters and the system’s physical properties. In this study, the physical parameters of conduit systems and downstream scour parameters were optimized using Artificial Neural Networks (ANN). The optimum system parameters were identified based on the experimental results through ANN-based optimization. Additionally, the real and optimal effects of the parameters on aeration performance and scour were determined. The results of the study demonstrated that ANN can be effectively used in studies aimed at optimizing the aeration performance and scour values of conduits. This study is the first to apply ANN for the simultaneous optimization of both aeration efficiency (Qa/Qw) and scour characteristics (dmax and δ) in a conduit-based hydraulic system. Previous studies have typically focused on either scour or aeration individually; this study integrates both through a multi-objective ANN model. The findings provide novel insights for the design and retrofitting of energy-dissipating conduit structures in high-head dam applications. Unlike traditional empirical or regression-based approaches, the ANN model captures complex, nonlinear relationships among multiple interacting hydraulic variables. It identified optimal parameter combinations — such as 70% gate opening, 1.30–1.50 m conduit length, and 9 mm air hole diameter — that improve both aeration and scour behavior. These outcomes offer direct applicability for real-world hydraulic system optimization.
Arici et al. (Wed,) studied this question.