Mean-Field-Type Model Predictive Control (MFT-MPC): An Application to Water Distribution Networks

In this paper, we study and design a mean-field-type model predictive control (MFT-MPC), which is a class of risk-aware control considering within the cost functional not only the mean but also the variance of both the system states and control inputs. Thus, the MFT-MPC is tightly related to the mean-variance paradigm involving risk minimization. We formally present the relationship between the MFT-MPC approach and the chance-constrained control techniques.We show a large-scale application over a distribution water network in order to illustrate the performance of the proposed stochastic optimizationbased control.