Exploring global approximators for multiobjective reservoir control

Efficient multi-purpose reservoir control policies are crucial in the face of frequent and severe floods droughts, to balance water allocation across conflicting demands. Evolutionary Multi-Objective Direct Policy Search (EMODPS) is a popular approach design for systems. EMODPS, however, relies on experimental choices within key components framework particularly when coupling multi-objective evolutionary optimization with nonlinear approximation networks. This study explores suite radial basis functions (RBFs) used map system's states actions flexible manner as time-varying, non-linear relationships. We provide systematic assessment different RBF explore their suitability obtain Pareto efficient policies. use Susquehanna river basin case which competing demands hydropower, environment, urban supply, atomic power plant cooling recreation need be met. Our findings suggest that choice have large impact model outcomes search behavior algorithm.