Reliability importance measures for a health-aware control of drinking water networks

This work focuses on a health-aware Model Predictive Control (MPC) scheme, which aims at enhancing the availability of the system. The objective is to extend the uptime of the system by delaying, as much as possible the system reliability decay. The weights of the MPC cost function are set according to some reliability importance measures. This work describes the main reliability importance measures and studies which of them are best suited for a health-aware MPC strategy applied to a Drinking Water Network. The overall system reliability as well as the reliability importance measures are computed online through a Dynamic Bayesian Network.