A Chaotic Sobol Sequence-based multi-objective evolutionary algorithm for optimal design and expansion of water networks

•A novel CS-MOSADE algorithm is developed and tested for WDN design problems.•CS-MOSADE algorithm converged faster than the MOSADE and NSGA-II algorithms.•Smaller spacing metric indicated better uniformity in the obtained solutions.•Large reductions in GHG emissions were achieved over the WDN servic...

Full description

Saved in:
Bibliographic Details
Published in:Sustainable cities and society Vol. 87; p. 104215
Main Authors: Sirsant, Swati, Hamouda, Mohamed A., Shaaban, Mostafa F., Al Bardan, Mayyada Salem
Format: Journal Article
Language:English
Published: Elsevier Ltd 01.12.2022
Subjects:
Expansion
GHG emissions
Multi-objective optimization
Optimal pump scheduling
Reliability assessment
Water distribution networks
DE
NSGA-II
GAFO
LP
CERI
GHG
EPS
NFE
HV
MODE-CSFLA
MC
SAES-RBF
EA
Expansion
Water distribution networks
WDN
EF
SAMOHS
CNHV
CS-MOSADE
MOPSO
LCC
PSO
RSM
Clustered-NA-ACO
Optimal pump scheduling
EADC
NP
GHG emissions
Reliability assessment
NLP
QRS
RNSGA-II
MOGA
GA
FADE
NHV
MODE-SaE
RHV
CF
ESS
FORM
PA-DSS
IGD
MOSADE-DP
NYT
WOA
SADE
jDE
ACO
MO-ASMOCH
PVC-O
HA
Multi-objective optimization
MCS
DESAP
ISSN:2210-6707, 2210-6715
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•A novel CS-MOSADE algorithm is developed and tested for WDN design problems.•CS-MOSADE algorithm converged faster than the MOSADE and NSGA-II algorithms.•Smaller spacing metric indicated better uniformity in the obtained solutions.•Large reductions in GHG emissions were achieved over the WDN service life. The design of a water distribution network (WDN) is an optimization problem that is computationally challenging with conflicting objectives. This study offers an enhanced Chaotic Sobol Sequence-based Multi-Objective Self-Adaptive Differential Evolution (CS-MOSADE) algorithm for multi-objective WDN design. The CS-MOSADE algorithm was tested on two benchmark WDNs, and a real WDN. Optimization results indicate that the CS-MOSADE algorithm converged two to three times faster than the MOSADE and NSGA-IIalgorithms and led to better output in terms of even distribution of solutions and convergence towards the true Pareto-optimal front. Smaller spacing metric indicated better uniformity in the obtained solutions; and larger hyper-area and coverage function values depicted better convergence towards the true Pareto-optimal front for the CS-MOSADE algorithm compared to the other algorithms. The CS-MOSADE algorithm was then applied to solve a WDN expansion problem for optimal pump scheduling and minimization of Life Cycle Cost, maximization of reliability and minimization of Green House Gas (GHG) emissions. A significant reduction in GHG emissions of 2.17 x 106 kg was achieved at an additional cost of $0.55 x 107 when optimal pump scheduling was incorporated in the model of the real WDN over service life of 50 years.
ISSN:2210-6707
2210-6715
DOI:10.1016/j.scs.2022.104215