Real-Time Reservoir Operation for Flood Control Using Artificial Intelligent Techniques

Real-time reservoir operation for flood control is a continuous and instant decision-making process based on relevant operating rules, in addition to the immediate rainfall and hydrological information. To reduce downstream flood peak stage and store floodwaters for future use, reservoir operational for flood control plays a critical role. The aim of this study is to establish a real-time reservoir operational model, the intelligent fuzzy flood control model (IFFCM), for flood control to ensure reservoir safety and store floodwaters for future use. The model includes two major artificial intelligent processes: knowledge acquirement and implementation, and fuzzy inference system. The IFFCM is first built by extracting the knowledge from the optimal operating hydrographs of typhoon events obtained by genetic algorithm (GA) and is then inferred reservoir release by the adaptive network-based fuzzy inference system (ANFIS). The Shihmen Reservoir in northern Taiwan is used as a case study with particular attention to the operations of 26 flood events by the proposed method. The results demonstrate that the proposed model can perform much better than historical reservoir operations and effectively reduce downstream peak flood stage and store floodwaters for future use