Samenvatting
Pressure and flow estimation in Water Distribution Networks (WDN) allows water management companies to optimize their control operations. For many years, mathematical simulation tools have been the most common approach to reconstructing an estimate of the WDN hydraulics. However, pure physicsbased simulations involve several challenges, e.g. partially observable data, high uncertainty, and extensive manual configuration. Thus, datadriven approaches have gained traction to overcome such limitations. In this work, we combine physicsbased modeling and Graph Neural Networks (GNN), a datadriven approach, to address the pressure estimation problem. First, we propose a new data generation method using a mathematical simulation but not considering temporal patterns and including some control parameters that remain untouched in previous works; this contributes to a more diverse training data. Second, our training strategy relies on random sensor placement making our GNNbased estimation model robust to unexpected sensor location changes. Third, a realistic evaluation protocol considers real temporal patterns and additionally injects the uncertainties intrinsic to realworld scenarios. Finally, a multigraph pretraining strategy allows the model to be reused for pressure estimation in unseen target WDNs. Our GNNbased model estimates the pressure of a largescale WDN in The Netherlands with a MAE of 1.94mH$_2$O and a MAPE of 7%, surpassing the performance of previous studies. Likewise, it outperformed previous approaches on other WDN benchmarks, showing a reduction of absolute error up to approximately 52% in the best cases.
Originele taal2  English 

Uitgever  arXiv 
Aantal pagina's  27 
Status  Submitted  17nov.2023 
Vingerafdruk
Duik in de onderzoeksthema's van 'Graph Neural Networks for Pressure Estimation in Water Distribution Systems'. Samen vormen ze een unieke vingerafdruk.Datasets

Water Distribution Networks (WDNs)  network operation state snapshots
Tello Guerrero, A. (Creator), Truong, H. (Creator), Lazovik, A. (Creator) & Degeler, V. (Creator), ZENODO, 4jan.2024
Dataset