This paper deals with uncertainty estimation and knowledge enhancement in water distribution networks (WDNs). A new three steps data assimilation approach is introduced, which in combination with multi-objective optimization, allows selecting effective and affordable monitoring networks. An innovative cascade of Ensemble Kalman Filters is used to assimilate the information deriving from sensors measuring pressure heads, flow in pipes and demands, with the objective of increasing knowledge while preserving at the same time the structural relationships among state variables. Selection of the most appropriate and economically affordable measurement network, is then based on the derivation of a Pareto front using the NSGA-II algorithm in conjunction with the data assimilation approach. The front is obtained by compromising between the overall sensors cost and the uncertainty reduction (or knowledge enhancement), which is expressed as a function of the Total Variance of state variables. The operational use of the proposed data assimilation approach as well as the effectiveness of the chosen observation network is also demonstrated by showing the reduction of uncertainty deriving from successive assimilations of real-time observations.

Enhancing Knowledge in Water Distribution Networks via Data Assimilation

BRAGALLI, CRISTIANA;FORTINI, MATTEO;TODINI, EZIO
2016

Abstract

This paper deals with uncertainty estimation and knowledge enhancement in water distribution networks (WDNs). A new three steps data assimilation approach is introduced, which in combination with multi-objective optimization, allows selecting effective and affordable monitoring networks. An innovative cascade of Ensemble Kalman Filters is used to assimilate the information deriving from sensors measuring pressure heads, flow in pipes and demands, with the objective of increasing knowledge while preserving at the same time the structural relationships among state variables. Selection of the most appropriate and economically affordable measurement network, is then based on the derivation of a Pareto front using the NSGA-II algorithm in conjunction with the data assimilation approach. The front is obtained by compromising between the overall sensors cost and the uncertainty reduction (or knowledge enhancement), which is expressed as a function of the Total Variance of state variables. The operational use of the proposed data assimilation approach as well as the effectiveness of the chosen observation network is also demonstrated by showing the reduction of uncertainty deriving from successive assimilations of real-time observations.
2016
Bragalli, Cristiana; Fortini, Matteo; Todini, Ezio
File in questo prodotto:
Eventuali allegati, non sono esposti

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11585/570293
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 21
  • ???jsp.display-item.citation.isi??? 18
social impact