Keywords
smart water network; network partitioning; graph theory; divide et conquer; network optimization
Location
Session H2: Water Resources Management and Planning - Modeling and Software for Improving Dcisions and Engaging Stakeholder
Start Date
18-6-2014 9:00 AM
End Date
18-6-2014 10:20 AM
Abstract
The application of Information and Communication Technology in different assets of urban life has contributed to generating the notion of Smart City in which Smart WAter Network (SWAN) can be assumed as subsystems. The possibility of inserting remote control valves and flow meters in a WDS allows the implementation of the paradigm of "divide and conquer", that consists into divide a large water network into k smaller subsystems, in order to simplify and improve the management and protection of SWAN. Traditional approaches for water network partitioning, based on empirical guiding principles (such as the maximum number of properties or total length of pipes in a district) combined with trial-and-error procedures, are not able to divide large water networks without a significant decrease of the hydraulic performance. Only recently in the scientific literature some approaches, essentially based on graph theory and partitioning techniques with Multi Level Recursive Bisection, have been proposed. This paper presents the SWANP (Smart WAter Network Partitioning) software that allows to achieve automatically a water network partitioning based on a novel multilevel algorithm based on bisection, genetic algorithm and original heuristic optimization approach based on energetic objective functions. Simulation results, carried out on real water network, confirms the effectiveness of the software that finds system partitioning in compliance with the level of service for the users. The SWANP software, developed in Python v2.7.6 language, includes some performance indices used to compare different water network partition layouts and provides to the operators a Decision Support System to choice the better solution.
Included in
Civil Engineering Commons, Data Storage Systems Commons, Environmental Engineering Commons, Other Civil and Environmental Engineering Commons
SWANP: software for automatic Smart WAter Network Partitioning
Session H2: Water Resources Management and Planning - Modeling and Software for Improving Dcisions and Engaging Stakeholder
The application of Information and Communication Technology in different assets of urban life has contributed to generating the notion of Smart City in which Smart WAter Network (SWAN) can be assumed as subsystems. The possibility of inserting remote control valves and flow meters in a WDS allows the implementation of the paradigm of "divide and conquer", that consists into divide a large water network into k smaller subsystems, in order to simplify and improve the management and protection of SWAN. Traditional approaches for water network partitioning, based on empirical guiding principles (such as the maximum number of properties or total length of pipes in a district) combined with trial-and-error procedures, are not able to divide large water networks without a significant decrease of the hydraulic performance. Only recently in the scientific literature some approaches, essentially based on graph theory and partitioning techniques with Multi Level Recursive Bisection, have been proposed. This paper presents the SWANP (Smart WAter Network Partitioning) software that allows to achieve automatically a water network partitioning based on a novel multilevel algorithm based on bisection, genetic algorithm and original heuristic optimization approach based on energetic objective functions. Simulation results, carried out on real water network, confirms the effectiveness of the software that finds system partitioning in compliance with the level of service for the users. The SWANP software, developed in Python v2.7.6 language, includes some performance indices used to compare different water network partition layouts and provides to the operators a Decision Support System to choice the better solution.
