Optimal pressure management in water distribution networks through district metered area creation based on machine learning

Novarini, Bernardo Rabah Felicetti; Brentan, Bruno Melo; Lima, Gustavo Meirelles; Luvizotto, Edevar

doi:10.1590/2318-0331.241920180165

Cited by 5 publications

(4 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…One important task on using clustering algorithms, such as k-means, is defining the number of clusters. In Novarini et al 29 various mathematical and engineering criteria for DMA design are evaluated. Mathematical criteria evaluate the quality of the clustering process in terms of external and internal measures.…”

Section: Clustering Process Using a Modified K-means Algorithmmentioning

confidence: 99%

District metered area design through multicriteria and multiobjective optimization

Brentan

Carpitella

Izquierdo

et al. 2021

Math Methods in App Sciences

Self Cite

View full text Add to dashboard Cite

The design of district metered areas (DMA) in potable water supply systems is of paramount importance for water utilities to properly manage their systems. Concomitant to their main objective, namely deliver quality water to consumers, the benefits include leakage reduction and prompt reaction in cases of natural or malicious contamination events. Given the structure of a water distribution network (WDN), graph theory is the basis for DMA design, and clustering algorithms can be applied to perform the partitioning. However, such sectorization entails a number of network modifications (installing cut-off valves and metering and control devices) involving costs and operation changes, which have to be carefully studied and optimized. Given the complexity of WDNs, optimization is usually performed using metaheuristic algorithms. In turn, optimization may be single or multiple-objective. In this last case, a large number of solutions, frequently integrating the Pareto front, may be produced. The decision maker has eventually to choose one among them, what may be tough task. Multi-criteria decision methods may be applied to support this last step of the decision-making process. In this paper DMA design is addressed by: i) proposing a modified k-means algorithm for partitioning; ii) using a multiobjective particle swarm optimization to suitably place partitioning devices; iii) using fuzzy analytic hierarchy process (FAHP) to weight the four objective functions considered; and iv) using technique for order of preference by similarity to ideal solution (TOPSIS) to rank the Pareto solutions to support the decision. This joint approach is applied in a case of a well-known WDN of the literature, and the results are discussed.

show abstract

Section: Clustering Process Using a Modified K-means Algorithmmentioning

confidence: 99%

District metered area design through multicriteria and multiobjective optimization

Brentan

Carpitella

Izquierdo

et al. 2021

Math Methods in App Sciences

Self Cite

View full text Add to dashboard Cite

show abstract

“…The models proposed from data mining and machine learning have become popular in the last decade [21]. Some of the developed applications using machine learning are applied to optimal pressure management and district metered area design [22], leak detection in WDSs [23][24][25][26][27], water demand estimation [28][29][30], and detection of cyberattacks, physical attacks, and contamination in WDSs [31,32].…”

Section: Introductionmentioning

confidence: 99%

A Digital Twin of a Water Distribution System by Using Graph Convolutional Networks for Pump Speed-Based State Estimation

Granados

Zanfei

Brentan

et al. 2022

Water

Self Cite

View full text Add to dashboard Cite

Water distribution system monitoring is currently carried out using advanced real-time control technologies to achieve a higher operational efficiency. Data analysis techniques can be implemented for condition estimation, which are crucial tools for managing, developing, and operating water networks using the monitored flow rate and pressure data at some network pipes and nodes. This work proposes a state estimation methodology that enables one to infer the hydraulic state of the operating speed of pumping systems from these pressure and flow measurements. The presented approach suggests using graph convolutional neural network theory linked to hydraulic models for generating a digital twin of the water system. It is validated on two benchmark hydraulic networks: the Patios-Villa del Rosario, Colombia, and the C-Town networks. The results show that the proposed model effectively predicts the state estimation in the two hydraulic networks used. The results of the evaluation metrics indicate low values of mean squared error and mean absolute error and high values of the coefficient of determination, reflecting high predictive ability and that the prediction results adequately represent the real data.

show abstract

“…They also employed multicriterion decision analysis to compare different solutions. Novarini et al (2019) partitioned DMAs using hybrid model and K-means for D-Town network. They considered mathematical and topological criteria.…”

Section: Introductionmentioning

confidence: 99%

Application of Fuzzy Cluster Analysis and Decision Making Algorithms for Choosing the Optimal Water Distribution Network Design

Arunachalam

Raju

Naidu

2021

Preprint

View full text Add to dashboard Cite

Non-dominated Sorting Genetic Algorithm (NSGA-II) is employed to facilitate multi-objective optimization in Water Distribution Network(s) (WDN) framework for a benchmark problem of Hanoi Network and a real-world problem, Pamapur Network, Telangana, India. Maximization of resilience, minimization of cost and minimization of leakages are considered in a multiobjective context which result in generation of Non-dominated WDN Strategies (NWDNS). In order to simplify the decision making process of engineers, Fuzzy Cluster Analysis (FCA) is employed to categorize NWDNS into groups. Thereafter, Dunn’s Cluster Validity Index (DCVI) is used for identification of optimal cluster size. Representative NWDNS i.e. RNWDNS for each sub-cluster is based on the maximum membership of NWDNS in the respective sub-cluster. Ranking of RNWDNS is performed with three decision making algorithms, namely, Preference Ranking Organization METHod for Enrichment of Evaluations-2 (PROMETHEE-2), Multicriterion Q-analysis-2 (MCQA-2) and VlseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR). Additive ranking rule is also applied to facilitate obtained ranks in group decision making environment to arrive at the optimal WDN. It is observed that 1020 NWDNS generated for Hanoi network are optimally classified into 18 clusters based on DCVI, and A13 representing RNWDNS 37 is found preferable. Whereas 272 NWDNS generated for Pamapur network are classified into 9 clusters where S6 is preferred (representing RNWDNS 203).

show abstract

Optimal pressure management in water distribution networks through district metered area creation based on machine learning

Cited by 5 publications

References 24 publications

District metered area design through multicriteria and multiobjective optimization

District metered area design through multicriteria and multiobjective optimization

A Digital Twin of a Water Distribution System by Using Graph Convolutional Networks for Pump Speed-Based State Estimation

Application of Fuzzy Cluster Analysis and Decision Making Algorithms for Choosing the Optimal Water Distribution Network Design

Contact Info

Product

Resources

About