Fuzzy optimal control of reservoir-assisted stormwater treatment areas for aquatic ecosystem restoration

Attachment of stormwater treatment areas (STAs) or constructed wetlands to stormwater retention reservoirs can achieve substantial reductions in pollutant loadings if properly operated and maintained. Besides water quality improvement, optimally operated reservoir-assisted STAs provide support for ecosystem remediation, flood control, and supplemental water supply. An adaptive, multiobjective real-time control model is developed for reservoir-assisted STA systems that incorporates fuzzy rule-based operating rules optimized using a genetic algorithm interacting with a simulation model of the system. The model is applied to the North Fork reservoir-assisted STA located in the watershed of the St. Lucie Estuary, Florida. Optimal daily feedback operational policies are developed for managing freshwater discharges to the Estuary for coastal ecosystem restoration, maximizing the natural treatment efficiency and of the STA, and providing supplemental water supply for irrigation. Testing and validation results from application of the fuzzy optimal control model confirm achievement of multiple targets and criteria for the North Fork project, while demonstrating potential for adaptive management of reservoir-assisted STA systems throughout the coastal regions of south Florida.