Deep fuzzy mapping nonparametric model for real-time demand estimation in water distribution systems: A new perspective

•The deep fuzzy mapping model (DFM) is explored for real-time WDS modeling.•The flaw of machine learning-based methods is solved by the proposed DFM.•The proposed DFM method is a pioneering study for real-time WDS modeling.•The DFM is a novel non-parametric model that has a unique analytical solutio...

Full description

Saved in:
Bibliographic Details
Published in:Water research (Oxford) Vol. 241; p. 120145
Main Authors: Zhang, Qingzhou, Yang, Jingzhi, Zhang, Weiping, Kumar, Mohit, Liu, Jun, Liu, Jingqing, Li, Xiujuan
Format: Journal Article
Language:English
Published: England Elsevier Ltd 01.08.2023
Subjects:
calibration
Deep fuzzy mapping
infrastructure
mathematical theory
Model calibration
Nodal demand
Real time estimation
water
water distribution
Water distribution system
Nodal demand
Deep fuzzy mapping
Water distribution system
Model calibration
Real time estimation
ISSN:0043-1354, 1879-2448, 1879-2448
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•The deep fuzzy mapping model (DFM) is explored for real-time WDS modeling.•The flaw of machine learning-based methods is solved by the proposed DFM.•The proposed DFM method is a pioneering study for real-time WDS modeling.•The DFM is a novel non-parametric model that has a unique analytical solution.•The DFM can be used as an alternative model to replace deep learning methods. Hydraulic modeling has been recognized as a valuable tool for improving the design, operation, and management of water distribution systems (WDSs) as it allows engineers to simulate and analyze behaviors of WDSs in real time and help them make scientific decisions. The informatization of urban infrastructure has motivated the real-time fine-grained control of WDSs, making it one of the hotspots in recent years, thereby putting higher requirements on WDS online calibration in terms of efficiency and accuracy, especially when dealing with large-complex WDSs. To achieve this purpose, this paper proposes a novel approach (i.e., deep fuzzy mapping nonparametric model (DFM)) from a new perspective for developing a real-time WDS model. To our knowledge, this is the first work that considers uncertainties in modeling problems using fuzzy membership functions and establishes the precise inverse mapping from pressure/flow sensors to nodal water consumption for a given WDS based on the proposed DFM framework. Unlike most traditional calibration methods that require time to optimize model parameters, the DFM approach has a unique analytical solution derived through rigorous mathematical theory, thus the DFM is computationally fast as a result of sensibly handling the problems whose solutions typically require iterative numerical algorithms and large computational time. The proposed method is applied to two case studies and the results obtained show that it can produce a real-time estimation of nodal water consumption with higher accuracy, computational efficiency, and robustness relative to traditional calibration methods. [Display omitted]
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:0043-1354
1879-2448
1879-2448
DOI:10.1016/j.watres.2023.120145