Dynamic Modeling of Water System and Predictive Flood Control in Karun and Dez Catchment Basins

Automatic flood control approach can prevent dam overtopping and downstream flooding. Doing so, in addition to controlling the optimal storage capacity, flood threats will turn into opportunities for water resources systems. Model-based Predictive Control (MPC), can take the future predicted rainfall runoff, physical and operational constraints, and safety limitations of the water system into account. In a suitable timeframe before destructive floods, MPC creates essential and optimal capacity in the dam’s reservoir to contain floods. In this research, water system modeling method is studied by a control-oriented approach, and the flood automatic control operation using a model predictive control approach. To this end, modeling of the Karun and Dez river system containing six reservoir dams is considered as the case study to model and simulate under the designed MPC. The state space model will extract to design and apply the predictive controller for a prediction horizon of 10 days and eight hours in MATLAB software. The results will show long prediction horizon gives the controller enough time to create the necessary capacity for flood control according to the system constraints, while a short prediction horizon, similar to other controllers without predicting, does not have an effective performance.