Model Predictive Control with exponential data weighting for flood regulation

In this paper, a new control strategy is suggested to alleviate the problem of occurrence of floods due to uncontrolled discharge of water from dams. Simulations are carried out based on real-time data of the actual flood scenario which occurred in Surat in the year, 1998. This flood occurred on account of uncontrolled outflow from Ukai dam located upstream of Surat city in Gujarat, India. In this paper, Model Predictive Control (MPC) with exponential data weighting is proposed as a control technique to solve the flood regulation problem. The simulations are done for three different values of exponential weighting factors. Finally, the results obtained using MPC with exponential data weighting for one such exponential weighting factor have been compared with the results obtained using simple MPC and manual control. Simulation results indicate that MPC with exponential data weighting have some added advantages over the simple MPC, because in MPC with exponential data weighting, current states are given more weight-age than future states.