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

Area of Science:

• Hydraulic Engineering
• Water Resource Management
• Artificial Intelligence in Civil Engineering

Background:

• Water distribution systems (WDSs) require real-time calibration for efficient design, operation, and management.
• Increasing urban infrastructure informatization necessitates accurate and efficient WDS online calibration, especially for large and complex systems.

Purpose of the Study:

• To propose a novel deep fuzzy mapping (DFM) nonparametric model for real-time WDS modeling.
• To establish a precise inverse mapping from sensor data to nodal water consumption, incorporating uncertainty using fuzzy logic.
• To develop a computationally efficient and accurate WDS calibration method.

Main Methods:

• Developed a deep fuzzy mapping (DFM) nonparametric model framework.
• Utilized fuzzy membership functions to address uncertainties in WDS modeling.
• Derived an analytical solution for precise inverse mapping from sensor data to nodal water consumption.
• Applied the DFM approach to two case studies for validation.

Main Results:

• The DFM model provides real-time estimation of nodal water consumption.
• Achieved higher accuracy and computational efficiency compared to traditional calibration methods.
• Demonstrated robustness in WDS modeling and calibration.

Conclusions:

• The DFM approach offers a significant advancement in real-time WDS calibration.
• The method effectively handles uncertainties and provides accurate, efficient, and robust estimations.
• This novel approach meets the growing demands for real-time fine-grained control of WDSs.