Author(s): Andrea Menapace; Ariele Zanfei

Linked Author(s): Andrea Menapace

Keywords: Big Data; Imputation; K-Nearest Neighbours; Water Demand; Water Distribution Networks

Abstract: Advancing the management of water distribution networks is considered a current priority for the sustainable management of the urban water cycle. However, this means increasing the complexity of the operational and planning decision-making tools, which are forced to handle an ever-increasing amount of data from the network of smart sensors with which modern systems are equipped. Thus, the goal of the proposed work is the need for proper preprocessing tools and, in particular, the critical role of a filling strategy. In particular, univariate and multivariate approaches are investigated along with different inputs and various algorithms. A real test case is used to carry out the proposed analysis, which consists of the water demand of four districts of the Italian city of Trento. The results highlight the importance of adopting a proper imputation strategy for processing the stream of water demand data to guarantee a reliable feed in the integrated management systems of smart water grids.

DOI: https://doi.org/10.3850/iahr-hic2483430201-383

Year: 2024