Combine Evolutionary Optimization with Model Predictive Control in Real-time Flood Control of a River System

In order to establish successful flood control strategies to prevent or alleviate severe flood damages, real-time optimization-based control can be a supplementary strategy, besides setting operating rules (regulations) for the hydraulic structures. This research combines evolutionary optimization, by means of a Genetic Algorithm (GA), with the Model Predictive Control (MPC) technique to develop and test a real-time flood control method for the 12 gated weirs in the Belgian case study of the river Demer. The evolution of this method is also the main contribution of this study. The combination of GA with MPC allows coping with the highly nonlinear system behaviour and local minimum problems. The system searches for better control actions by minimizing a cost function while at the same time avoiding violation of the defined constraints. The optimization results testify that the system is able to assist the current regulation strategies that are based on fixed regulation rules (three-position controller).