Optimizing control by robustly feasible model predictive control and application to drinkingwater distribution systems

The paper considers optimizing Model Predictive Control (MPC) for nonlinear plants with output constraints under uncertainties. Although the MPC technology can handle the constraints in the model by solving constraint model based optimization task, satisfying the plant output constraints under the model uncertainty still remains a challenge. The paper proposes Robustly Feasible MPC (RFMPC), which achieves feasibility of the outputs in the controlled plant. The RFMPC which is applied to control quantity in Drinking Water Distribution Systems (DWDS) is illustrated by application to the DWDS example. In the simulation exercise, Genetic Algorithm is selected as the optimization solver and the reduced search space methodology is applied in the implementation under MATLABEPANET environment.