water

Article
Improving Multi-Objective Optimization Methods of Water
Distribution Networks
Rahel Amare Kidanu 1, * , Maria Cunha 2 , Elad Salomons 1 and Avi Ostfeld 1

1 Technion—Israel Institute of Technology, Faculty of Civil and Environmental Engineering, Haifa 32000, Israel;
eladsa@[Link] (E.S.); ostfeld@[Link] (A.O.)
2 Department of Civil Engineering, CEMMPRE, University of Coimbra, 3004-531 Coimbra, Portugal;
mccunha@[Link]
* Correspondence: rahel-amare@[Link]

Abstract: Water distribution network design is a complex multi-objective optimization problem and
multi-objective evolutionary algorithms (MOEAs) such as NSGA II have been widely used to solve
this optimization problem. However, as networks get larger, NSGA II struggles to find the diverse
and uniform solutions that are critical in multi-objective optimization. This research proposes an
improved version of NSGA II that uses three new-generation methods to target different regions of
the Pareto front and thus increase the number of solutions in critical regions. These methods include
saving an archive, local search around extreme and uncrowded Pareto front, and local search around
the knee area of the Pareto front. The improved NSGA II is tested on benchmark networks of different
sizes and compared to the best-known Pareto front of the networks determined by MOEAs. The
results show that the proposed algorithm outperforms the original NSGA II in terms of broadening
the Pareto front solution range, increasing solution density, and discovering more non-dominated
solutions. The improved NSGA II can find solutions that cover all parts of the Pareto front using a
single algorithm without increasing computational effort.

Keywords: water distribution network; multi objective optimization; MOEAs; NSGA II; improved
NSGA II

Citation: Kidanu, R.A.; Cunha, M.;

Salomons, E.; Ostfeld, A. Improving
1. Introduction
Multi-Objective Optimization Water supply systems are an essential part of a vast urban infrastructure system
Methods of Water Distribution that underpins most economic activity and is used by urban populations to perform
Networks. Water 2023, 15, 2561. basic domestic activities [1]. A water distribution network is a hydraulic infrastructure
[Link] that is usually the most costly component of urban infrastructure that delivers water to
Academic Editor: Dandy Graeme nodes while maintaining predetermined pressure requirements [2,3]. Designing such
networks involves choosing optimal combinations of values for decision variables such as
Received: 31 May 2023 pipe sizes, tank shapes and sizes, pump types, and valve locations, while also satisfying
Revised: 6 July 2023 several constraints [4]. Optimal design problems for Water Distribution Systems (WDSs)
Accepted: 7 July 2023
typically involve selecting the best pipe size, which is challenging due to the problems’
Published: 12 July 2023
high dimensionality, discreteness, and non-linearity [5]. To solve these problems, advanced
optimization techniques are required due to the complexity of WDSs and their components’
arrangement [6].
Copyright: © 2023 by the authors.
WDS optimization aims to maximize system performance and profitability while
Licensee MDPI, Basel, Switzerland.
minimizing resource consumption and cost, all of which are typically limited [7]. As
This article is an open access article constructing these systems is costly, the optimization objective is to find the least expensive
distributed under the terms and network [8]. The design of water systems is a multi-objective optimization problem (MOPs)
conditions of the Creative Commons that needs to simultaneously optimize a number of objectives in addition to cost such
Attribution (CC BY) license (https:// as operational, life cycle, and maintenance costs, system reliability, and water quality [9].
[Link]/licenses/by/ Optimizing for one objective may result in suboptimal performance for others [8]. To
4.0/). address this complex problem, engineers have used single-objective optimization models

Water 2023, 15, 2561. [Link] [Link]

Water 2023, 15, 2561 2 of 17

in the past, but multi-objective optimization approaches have become increasingly popular
in recent years, providing a way to investigate trade-offs between objectives and make
informed decisions.
Optimizing water distribution networks (WDNs) is crucial for their performance and
reliability. Network resilience, which refers to the network’s ability to recover from internal
or external disturbances, is an important aspect of this optimization [10]. Incorporating
network resilience in the objective function can lead to better results in terms of surplus
power and redundancy [10]. Mechanical or hydraulic failure in WDNs can cause increased
head losses and network failure, which is why it is essential to have excess power available
for internal dissipation. Simply optimizing for cost can leave the network vulnerable
to hydraulic and mechanical failures [11]. Including reliability in optimization can be
computationally intensive, so approximate methods like the resilience index [12] can be
used. The inclusion of network resilience in WDN optimization is critical for maintaining
the network’s performance and reliability in both normal and failure conditions [13].
Various optimization methods are available for water distribution systems and can
be categorized into deterministic and stochastic techniques. Deterministic techniques in-
volve linear programming, non-linear programming, and dynamic programming, while
stochastic techniques include population-based algorithms and single point-based meth-
ods. Metaheuristics, a type of stochastic technique, provides a near-optimal solution in a
single run by using principles from nature and involving random components. Genetic
algorithms, inspired by mechanisms of biological evolution, have been widely applied in
water resources planning and management [14,15]. These algorithms are capable of solving
nonlinear, nonconvex, multimodal, and discrete problems, expanding their capabilities in
handling complex environmental and water resource applications [15].
Classical search and optimization methods may not be effective in dealing with MOPs
due to their inability to find multiple solutions in a single run and the challenges in
handling problems with discrete variables and multiple optimal solutions [9]. Evolutionary
algorithms (EAs), particularly MOEAs, are effective for MOPs as they use a population of
solutions to find multiple Pareto-optimal solutions in a single run [16]. Diversity preserving
mechanisms can be incorporated into MOEAs to find widely different Pareto-optimal
solutions [9]. EAs are less affected by the shape or continuity of the Pareto front, unlike
classical methods [17]. In water distribution system design, Pareto archived evolutionary
strategy (PAES), Strength Pareto evolutionary algorithm (SPEA-2), and non-dominated
sorting genetic algorithm II (NSGA-II) [18] are the most widely used MOEAs that effectively
handle multiple objectives and find a set of solutions that are not dominated by others.
The use of these MOEAs has greatly improved the efficiency and effectiveness of the water
distribution system design process.
The Non-Dominated Sorting Genetic Algorithm (NSGA-II) is a widely used algorithm
for solving MOPs with both continuous and discrete variables [19]. It is considered one
of the most representative algorithms in multi-objective optimization [20] and had been
recognized as the EA state-of-the-art in solving WDS-related optimization problems [21].
NSGA-II is an improved version of the Non-Dominated Sorting Genetic Algorithm (NSGA)
that overcomes some of its limitations, such as the absence of elitism and the need to define
sharing parameters for diversity preservation [22]. The design of NSGA-II incorporates
an elitist strategy to expand the sample space and improve optimization accuracy, and the
crowding distance operator to preserve diversity. NSGA-II is computationally efficient, with
an overall complexity of at most O(MN2 ), where M is the number of objective functions
and N is the population size. The fast non-dominated sorting method used in NSGA-
II reduces computational complexity, making it efficient and powerful in exploring the
decision space of MOPs. NSGA-II has been widely implemented and applied to various
MOPs, demonstrating its effectiveness and reliability.
NSGA-II has been combined with other methods and tools to enhance its effectiveness
in solving optimization problems in WDS. The Robust NSGA-II (RNSGA-II) [23] algo-
rithm was developed based on NSGA II to ensure that the solutions are robust enough
Water 2023, 15, 2561 3 of 17

and able to sustain longer over multiple generations in the optimization process. The
Epsilon-dominance Non-dominated Sorting Genetic Algorithm II (e-NSGA-II) is another
optimization algorithm that is based on the original NSGA II that utilizes the e-dominance
concept, adaptive population sizing, and self-terminating algorithm to achieve well-spread
and well-converged Pareto-optimal solutions. Non-dominated Sorting Genetic Algorithm
III (NSGA-III) [24] is an improved version of NSGA-II using a reference-point-based non-
dominated sorting approach. However, NSGA-II is still the most preferred algorithm
compared to NSGA-III, and variations of NSGA-II, such as RNSGA-II, e-NSGA-II, and a
combination of NSGA-II with other methods, have shown that NSGA-II is still relevant
and capable of handling the task [25].
Wang et al. [26] applied five state-of-the-art MOEAs to twelve design problems col-
lected from the literature, with minimum time invested in parameterization by using the
recommended settings. The study found that the MOEAs were complementary to each
other, and that NSGA-II remained a good choice for two-objective optimization of water
distribution systems (WDSs) and generally outperformed the other MOEAs in terms of
the number of solutions contributed to the best-known Pareto front of each problem. The
spread (both extent and uniformity) of its contribution was also comparable, if not better,
than other MOEAs. Overall, the paper contributes to the best-known approximations to
the true Pareto fronts of a wide range of benchmark problems, and the results are going
to be used in this research for comparison of some of the benchmark networks. However,
NSGA-II and other MOEAs had limitations with intermediate and large-sized problems,
resulting in a limited range of solutions with low diversity. In order to overcome these
limitations and achieve better results for multi-objective optimization problems, this study
proposes an improved version of NSGA-II. In their work, Cunha et al. [27] introduced
MOSA-GR, a novel multi-objective simulated annealing algorithm equipped with innova-
tive searching mechanisms. MOSA-GR demonstrated superior performance by generating
Pareto fronts that surpassed those produced by MOEAs, even when merging their results.
In this study, the new searching mechanisms from MOSA-GR will be applied to NSGA-II to
develop an improved algorithm for better results and contribute to the advancement of the
existing literature on multi-objective optimization algorithms, enabling effective handling
of complex real-world problems with enhanced diversity and density of solutions.

2. Materials and Methods

2.1. Problem Definition of the Multi-Objective Problems
The optimization problems in Water Distribution Systems (WDSs) design usually
involve two objective functions: minimizing the cost and maximizing the network reliability
or robustness. In this context to solve the multi-objective optimization, the objectives
defined in model formulation are minimizing the cost while maximizing resilience. The
decision variables are diameter, which can be adjusted to obtain an optimal solution that
satisfies both objectives. The cost objective function considers the expenditure of pipe
components as the total cost of a design solution. The unit cost of a specific pipe diameter
for each problem is derived from the paper by [26] (Equation (1) below).
Resilience in the context of WDS design refers to the ability of the system to continue
functioning in the presence of failures or disturbances. The specific definition of resilience
may vary depending on the WDS design optimization problem. In this study, the resilience
index proposed by [10] (Equation (2) below) is used as a measure to optimize the reliability
of the network. It is recognized that some authors have proposed updates to Todini’s index
of resilience and that other indexes have been proposed [28–30], however it is used in this
paper for the following reasons: (1) it is widely used in the literature on water distribution
systems optimization and (2) it enables a comparison with the work of Wang et al. [26].
The resilience index is defined as the ratio of power dissipated in the network to the
maximum power that would be dissipated in order to satisfy the design demand and head
requirements at the junction nodes.
Water 2023, 15, 2561 4 of 17

The objectives to be optimized to obtain the optimal WDS design solution are pre-
sented in mathematical form as Equations (1) and (2), where the cost is minimized while
maximizing the resilience index.
N
Min Cost = ∑ij Cij × Lij + ∑ Cp (1)

 
req
∑nn C Q
j=1 j j H j − H j
Max Ir =  npu Pi

req
(2)
nr
∑k=1 Qk Hk + ∑i=1 γ − ∑nn j = 1 Q j Hj

npj
∑i=1 Di
Cj = (3)
npj × max{Di}
where Ir = network resilience; nn = number of demand nodes; Cj, Qj, Hj, and Hj req = uniformity,
demand, actual head, and minimum head of node j; nr = number of reservoirs; Qk and
Hk = discharge and actual head of reservoir k; npu = number of pumps; Pi = power of pump i;
γ = specific weight of water; npj = number of pipes connected to node j; Di = diameter of pipe i
connected to demand node j.

2.2. Improved NSGA II

The NSGA II algorithm is widely used for solving multi-objective optimization prob-
lems. However, the quality of the solutions generated by the algorithm can be improved.
Therefore, the main objective of improving the NSGA II algorithm is to enhance the quality
of its solutions. This involves increasing the density of solutions, especially in the im-
portant regions of the Pareto front, such as the knee area, and improving the diversity of
solutions by expanding their range. To achieve this, the improved NSGA II introduces
novel improvements in both the diversity and convergence of the algorithm.
One of the main developments of the improved NSGA II algorithm is the implemen-
tation of new methods for offspring population generation. Unlike the original NSGA II,
which uses all parent population solutions in tournament selection to generate the offspring
population, the improved NSGA II focuses on specific solutions among the nondominated
solutions found so far to generate new solutions in addition to the original method. By fo-
cusing on specific solutions, the improved NSGA II can generate better candidate solutions,
ultimately leading to a better quality of solutions in the Pareto front.
The improved NSGA II algorithm introduces different generation methods based
on [27], which is crucial to achieving fast convergence and a high diversity of solutions
that are uniformly distributed in the objective space. The different generation methods,
including those that target the maximum and minimum regions of the Pareto front and
the uncrowded area of the front, are used to increase the range and diversity of solutions.
The knee area of the Pareto front is given special attention, as it is a critical region with
solutions that exhibit a small improvement in one objective causing a large deterioration in
the other.

2.3. Generation Methods

The algorithm generates an offspring population using four main processes (Gs), which
are selected in each iteration starting from different starting iterations (ITm) proposed to
control the process. Additionally, different probabilities (represented by Ps) are used to
select the generation methods. The selection of these probabilities aims to produce new
solutions that can perform a global search in the early stages of the process and then include
a more localized search towards the end in addition to the global search.
The following are the types of searching methods presented as offspring generation
methods:
1. G1 method: This is the original method used in NSGA II. It involves selecting N
(population size) number of parent populations and N offspring populations, evalu-
 eter as in Equation (4), where Rand is a random number between 0 and 1.
Offspring = Rand × parent + (1 − Rand) × Maximum diameter, (4)
On the other hand, if the selected points are from the minimum and uncrowded ar-
Water 2023, 15, 2561 eas, the offspring population is generated in Equation (5) by randomly assigning a diam-5 of 17
eter value of the next higher or lower possible diameter value to the selected pipes, where
“gene” is the randomly selected pipe.
ating their objective values, and sorting
Gene = them
rand(),based on non-dominated sorting and
crowd distancing. The parent populations are then randomly paired to create child (5)
Offspring(gene)
populations through crossover = Parent(gene) ± 1,
and mutation.
2.
where G2
Genemethod: This generated
= randomly method saves the new
number, offspring
Parent = selectedpopulation generated in each
points, Offspring=generated
iteration to the archive.
population from selected points. The archive contains all the populations generated from the
start of the iteration. Parent populations are randomly selected from the archive to
4. G4 method:
create This method
new offspring generatesthrough
populations offspring populations
crossover using the knee area of the
and mutation.
3. Pareto front asThis
G3 method: the parent
methodpopulation.
focuses onThe knee is
specific found
areas of by
thecalculating thesuch
Pareto front, Euclidean
as the
distance of all Pareto front points to the corner, shown in Figure 1 and
extreme and uncrowded areas shown in Figure 1. At each iteration, a number selecting points
of
with the leastare
points distance.
selectedAfter
fromselecting the parent
the required region,population, the offspring
and N (population population
number) is
offspring
generated using Equation
populations (5). from these points.
are generated

Theselected
[Link]
Figure selectedregions
regionsof
ofpareto
paretofront
frontfor
forG3
G3and
andG4.
G4.

If the selected points are from the maximum region, the offspring population is
generated via bound type crossover and weighted average with the maximum possible
diameter as in Equation (4), where Rand is a random number between 0 and 1.

Offspring = Rand × parent + (1 − Rand) × Maximum diameter, (4)

On the other hand, if the selected points are from the minimum and uncrowded areas,
the offspring population is generated in Equation (5) by randomly assigning a diameter
value of the next higher or lower possible diameter value to the selected pipes, where
“gene” is the randomly selected pipe.

Gene = rand(), (5)

Offspring(gene) = Parent(gene) ± 1,
where Gene = randomly generated number, Parent = selected points, Offspring=generated
population from selected points.
4. G4 method: This method generates offspring populations using the knee area of the
Pareto front as the parent population. The knee is found by calculating the Euclidean
distance of all Pareto front points to the corner, shown in Figure 1 and selecting points
with the least distance. After selecting the parent population, the offspring population
is generated using Equation (5).
 2.4. Pseudocode
2.4. Pseudocode
The pseudocode provided in Figure 2 outlines the steps of the improved NSGA II
The pseudocode provided in Figure 2 outlines the steps of the improved NSGA II
procedure. The maximum number of iterations is divided into four stages and each stage
procedure. The maximum number of iterations is divided into four stages and each stage
includes a different combination of generation methods, as shown in Figure 3. The algo-
includes a different combination of generation methods, as shown in Figure 3. The algo-
rithm starts with the original NSGA II method and includes additional methods one by
Water 2023, 15, 2561 rithm starts with the original NSGA II method and includes additional methods one by17
6 of
one into each stage. In the early phases, only the G1 method is used, but as the second
one into each stage. In the early phases, only the G1 method is used, but as the second
stage is initiated, G2 is introduced. In the third stage, G3 becomes available for use, and
stage is initiated, G2 is introduced. In the third stage, G3 becomes available for use, and
likewise, G4 is brought into the fold in the fourth stage.
likewise, G4 is brought into the fold in the fourth stage.
2.4. Pseudocode
In each iteration, a generation method from the available methods in each stage is
In each iteration, a generation method from the available methods in each stage is
picked usingThe a probability provided
pseudocode decision factor, and 2that
in factor,
Figure method thegeneratesof N offspring popu-
picked using a probability decision andoutlines
that method steps
generatestheNimproved NSGA
offspring popu-
lations
II in that
procedure. iteration.
The The
maximum number of
number function
of evaluations
iterations is each
divided method
into four uses is
stages deter-
and each
lations in that iteration. The number of function evaluations each method uses is deter-
mined byincludes
stage the probability
a assigned
different to each of
combination method and the
generation stage’s as
methods, starting
shown iteration.
in Figure The
3. The
mined by the probability assigned to each method and the stage’s starting iteration. The
fouralgorithm
generating processes
starts with thecan be selected
original NSGA according
II method to the
and starting
includes iterationmethods
additional number one
four generating processes can be selected according to the starting iteration number
(ITm1, ITm2,
by one ITm3,
into each ITm4)
stage. and probabilities
In the (P1,only
P2, the
P3, G1
P4). These additional parameters
(ITm1, ITm2, ITm3, ITm4) andearly phases,
probabilities (P1, P2, P3, method
P4). Theseis used, but
additional asparameters
the second
(explained
stage is below in detail)
initiated, G2 is are used to In
introduced. control
the the number
third stage, G3 ofbecomes
functionavailable
evaluationsfor used
use,used
and
(explained below in detail) are used to control the number of function evaluations
by the generation
likewise, G4 isprocesses
brought intoandthe
thefold
results they
in the can contribute.
fourth stage.
by the generation processes and the results they can contribute.

1. Begin
1. Begin
1.1 Input P, N, maximum_generation
1.1 Input P, N, maximum_generation
1.2 Divide the maximum_generation into four stages (set parame-
1.2 Divide the maximum_generation into four stages (set parame-
ters)
ters)
1.3 Initialize parent population- Po
1.3 Initialize parent population- Po
1.4 Evaluate objective values Po
1.4 Evaluate objective values Po
1.5 Initialize offspring population- No
1.5 Initialize offspring population- No
2. FOR i=1 to maximum generation DO
2. FOR i=1 to maximum generation DO
2.1 Evaluate objective values of -No
2.1 Evaluate objective values of -No
2.2 objective value of parent + offspring
2.2 objective value of parent + offspring
2.3 Sort based on domination and crowd distance.
2.3 Sort based on domination and crowd distance.
2.4 Select generation method (G1, G2, G3, G4)
2.4 Select generation method (G1, G2, G3, G4)
2.5 Generate offspring population based on selected method:
2.5 Generate offspring population based on selected method:
2.6 Combine offspring + parent.
2.6 Combine offspring + parent.
2.7 END FOR
2.7 END FOR
3 Output: extract the best pareto front.
3 Output: extract the best pareto front.
Figure 2. Pseudocode
Figure of Improved
Pseudocode NSGA
ofImproved
Improved II. II.
NSGA
Figure [Link] of NSGA II.

Figure
Figure 3. The
3. The available
available options
options of generation
of generation methods
methods at every
at every stagestage throughout
throughout the algorithm.
the algorithm.
Figure 3. The available options of generation methods at every stage throughout the algorithm.
In each iteration, a generation method from the available methods in each stage is
picked using a probability decision factor, and that method generates N offspring popula-
tions in that iteration. The number of function evaluations each method uses is determined
by the probability assigned to each method and the stage’s starting iteration. The four
generating processes can be selected according to the starting iteration number (ITm1,
ITm2, ITm3, ITm4) and probabilities (P1, P2, P3, P4). These additional parameters (ex-
plained below in detail) are used to control the number of function evaluations used by the
generation processes and the results they can contribute.
Water 2023, 15, 2561 7 of 17

2.5. Parameters
The selection of the four generating processes is based on three additional parameters
to the NSGA II algorithm: the starting iteration number (ITm1, ITm2, ITm3, ITm4), proba-
bilities (P1, P2, P3, P4), and the number of selected points as a percentage of population
size. These parameters are used to control the number of function evaluations used by the
generation processes and the results they can contribute.
1. ITm1—iteration number where the G1 starts, which is equal to zero.
2. ITm2—iteration number where the G2 starts.
3. ITm3—iteration number where the G3 starts.
4. ITm4—iteration number where the G4 starts.
The ITm parameter represents the iteration number at which the generation methods
start and can take values from 0 to 1, where 0 represents the beginning of the iterations
and 1 represents the end of the iterations. The starting iteration of a generation method is
determined by multiplying ITm by the maximum number of iterations as in Equation (6).
For example, the starting iteration of G2 can be calculated as ITm2 multiplied by the
maximum number of iterations.

Starting iteration of a generation method = ITm × maximum number of iterations (6)

where ITm = a parameter between 0 and 1.

The other parameter is the probability assigned to each method to be selected. In each
iteration, a random number between 0 and 1 is generated to determine the probability of
selecting a particular generation method.
1. P2—probability of G2 being selected starting ITm2.
2. P3—probability of G3 being selected starting ITm3.
3. P4—probability of G4 being selected starting ITm4.
The probabilities of G2, G3, and G4 are determined by P2, P3, and P4, respectively. The
probabilities of these methods are constant from their starting iteration to the maximum
number of iterations. The probability of G1, on the other hand, changes throughout the
stages. In stage one, G1 is the only method used, and therefore has a probability of 1. In
stage two, there are two options: G2 with a probability of P2 and G1 with a probability
of (1 − P2). Similarly, in stage three, there are three options: G2 with a probability of
P2, G3 with a probability of P3, and G1 with a probability of (1 − P2 − P3). The sum of
probabilities for G2, G3, and G4 is less than 1, with G1 taking up the remaining probabilities
according to the stage.
As an instance, during stage 3, the selection of a method from G1, G2, and G3 is
determined based on specific conditions outlined in Equations (7) through (9). This involves
generating a random number between 0 and 1. If this number is less than (1 − P2 − P3)
as in Equation (7), G1 is selected. If the random number falls between (1 − P2 − P3) and
(1 − P3) as in Equation (8), G2 is chosen. On the other hand, if the random number is
greater than (1 − P3) as in Equation (9), G3 is the selected method.

0 < Rand < (1 − P2 − P3) (7)

(1 − P2 − P3) < Rand < (1 − P3) (8)

(1 − P3) < Rand < 1 (9)

In addition, the number of points selected by each method is a crucial parameter in
the improved NSGA II algorithm. G3 and G4 choose the parent population from the search
region and generate the offspring population. The number of points to be selected must
be determined in advance. The number of selected points is equal to sp multiplied by
 Water 2023, 15, 2561 8 of 17

Water 2023, 15, x FOR PEER REVIEW 8 of 17

the population size (N) as in Equation (10), where sp is number of selected points as a
percentage of the population size (N).

Number
Number ofof selected
selected points
points = sp
= sp ×N×N (10) (10)

1. 1.sp3max—number
sp3max—number of selected
of selected points
points in maximum
in the the maximum region
region as percentage
as percentage of [Link] N.
2. 2.sp3min—number
sp3min—number
of selected points in the minimum region as percentage of [Link] N.
of selected points in the minimum region as percentage
3. 3.sp3uc—number
sp3uc—number of selected
of selected points
points in uncrowded
in the the uncrowded region
region as percentage
as percentage of [Link] N.
4. 4.sp4—number
sp4—number of selected points in the knee area as percentage
of selected points in the knee area as percentage of N. of N.
G3 selects
G3 selects three
three types
types of points
of points based
based on three
on three additional
additional parameters:
parameters: P3max,
P3max, P3min,
P3min,
and P3uc. These parameters control the number of selected points in the maximum
and P3uc. These parameters control the number of selected points in the maximum region, region,
minimum region, and uncrowded region, respectively.
minimum region, and uncrowded region, respectively.
The The pseudocode
pseudocode of generation
of generation methods
methods is depicted
is depicted in Figure
in Figure 4, where
4, where Rand2
Rand2 is a is a
second
second random
random number
number between
between 0 and
0 and 1. 1.

IF (Rand < P2 AND It > ITm2) THEN

G2: Generate offspring using randomly selected points from archive.
ELSEIF ((P3+P2) > Rand > P2 AND It > ITm3) THEN
Rand2=rand();
IF (Rand2<P3min)
Select sp3min *N
ELSEIF ((P3max+ P3min) > Rand2 > P3min)
Select sp3uc *N
ELSE
Select sp3max*N
G3: Generate offspring population from the selected points.
ELSEIF (It > ITm4 && (P4+P3+P2) > Rand > (P3+P2))
Select sp4 *N
G4: Generate offspring using the selected points in the knee area.
ELSE
G1: Generate offspring using ALL points in parent population.

Figure 4. The
Figure 4. Pseudocode of random
The Pseudocode process
of random of selecting
process generation
of selecting methods.
generation methods.

The The
algorithm
algorithm starts by using
starts G1 in
by using G1the in initial iterations
the initial iterationsto create the Pareto
to create the Pareto frontfront
and and
initialize the processes.
initialize the processes. This generation process
This generation can becan
process revisited
be revisitedthroughout the entire
throughout the it-
entire
eration. Then,Then,
iteration. G2 is G2 introduced, and itand
is introduced, brings backback
it brings somesome populations
populations fromfrom the archive that that
the archive
werewere eliminated
eliminated in the in process
the process but but
maymay havehave the the potential
potential to findto find
betterbetter solutions
solutions when when
combined
combined withwith the current
the current population.
population. Once Once a front
a front is established,
is established, the algorithm
the algorithm usesuses
G3 G3
and and
G4 toG4target
to target specific
specific regions
regions of the
of the front. front.
G4 is G4executed
is executed last last because
because it requires
it requires an an
intensive
intensive evaluation,
evaluation, and and its characteristics
its characteristics are appropriate
are appropriate for thefor knee
the kneeregionregion
late late
in thein the
progression of the front. These strategies help the algorithm
progression of the front. These strategies help the algorithm converge more quickly to the converge more quickly to the
Pareto optimal front and increase the range
Pareto optimal front and increase the range of solutions it can [Link] solutions it can find.
In Figure
In Figure 5, the5, the generation
generation methods
methods usedininthe
used thealgorithm
algorithm are illustrated
illustratedfor forananexample
ex-
ampleof one
of onerun, along
run, along with thethe
with number
number ofof iterations
iterationseach eachgeneration
generationwas wasemployed.
employed. At At the
outset, G1
the outset, G1 isis employed
employedat atevery
everyiteration,
iteration,and andthetheline
lineforforG1 G1increases
increaseswith with a slope
a slope ofof 1.
When the iteration reaches ITm2 = 500, G2 is added as an
1. When the iteration reaches ITm2 = 500, G2 is added as an option, and the line for G2 option, and the line for G2 begins
to increase.
begins to increase. The Theslope of of
slope G1G1 becomes
becomes lower
lower atatthis
thispoint
pointsincesinceG2 G2isisalso
alsoininuse.
[Link]
The low
slope of G2 is related to the probability it is given, which is lower
low slope of G2 is related to the probability it is given, which is lower than that of G1 (P1 than that of G1 (P1 > P2).
> P2). At the iteration of ITm3 = 750, G3 is introduced, and the line for G3 starts increasing. The
At the iteration of ITm3 = 750, G3 is introduced, and the line for G3 starts increasing.
The slope
slope of
of G3
G3 isis higher
higher thanthan G2G2 because
because itit hashasaahigher
higherprobability
probability(P3 (P3>>P2).P2).Lastly,
Lastly,G4 G4 is
includedatatthe
is included thelast
laststage
stagewhen
whenthetheiteration
iterationisisatatITm4ITm4= =1250,1250,and andthe theline
lineforforG4G4begins
begins to
increase. The slope of G4 is higher than G2 but lower than G3,
to increase. The slope of G4 is higher than G2 but lower than G3, indicating that its prob- indicating that its probability
ability is higher than G2 but lower than G3. As new methods are added, the slope of G1
decreases, and the probability of G1 is a function of the other probabilities (1 − P2 − P3 −
P4).
 Water 2023, 15, 2561 9 of 17

Water 2023, 15, x FOR PEER REVIEW 9 of 17

is higher than G2 but lower than G3. As new methods are added, the slope of G1 decreases,
and the probability of G1 is a function of the other probabilities (1 − P2 − P3 − P4).

Anexample
[Link]
Figure exampleofofprogress
progressofofgeneration
generationmethods
methodsthroughout
throughoutthe
theiterations.
iterations.

2.6. Case Studies

2.6. Case Studies
To test the improved algorithm in this study, five networks of different sizes were used.
To test the improved algorithm in this study, five networks of different sizes were
These networks are taken from the benchmark case studies of [26] and range from small to
used. These networks are taken from the benchmark case studies of [26] and range from
intermediate in size. Table 1 summarizes the benchmark problems, including the number
small to intermediate
of water in [Link],
sources, decision Table 1 summarizes the benchmark
pipe diameter options, theproblems, including
search space the
calculated
number of water sources, decision variables, pipe diameter options, the search space
as pipe diameter options to the power of decision variables and the number of function cal-
culated as pipe
evaluation (NFE)diameter
used. options to the power of decision variables and the number of
function evaluation (NFE) used.
Table 1. Benchmark network characteristics, WS = water sources, DV = decision variables, PD = pipe
Table 1. Benchmark network characteristics, WS = water sources, DV = decision variables, PD = pipe
diameter options, NFE = number of function evaluation.
diameter options, NFE = number of function evaluation.
Problem
Problem Name WS
Name WS DVDV PD PD SearchSearch
SpaceSpace NFE
NFE
9
Two-Loop Network
Two-Loop Network TLN TLN1 1 8 8 14 14 109 × 10
1.48 × 1.48 100,000
100,000
26
Hanoi Hanoi
Network
Network HAN HAN1 1 34 34 6 6 1026× 10
2.87 ×2.87 600,000
600,000
GoYang Network GOY 1 30 8 1.2427× 1027 600,000
GoYang Network GOY 1 30 8 1.24 × 10 600,000
Fossolo Network FOS 1 58 22 7.25 × 1077 1,000,000
PescaraFossolo
NetworkNetwork PES FOS3 1 99 58 22 13 7.25 ×1.91
1077× 10110 1,000,000
1,000,000
Pescara Network PES 3 99 13 1.91 × 10110 1,000,000

InIn[26]
[26]different
differentcomputational
computationalbudgets
budgetsareareused
usedfor
forthe
thenetworks
networksbased
basedon
ontheir
theirsize
size
and complexity. To ensure fair comparison, the same computational budget
and complexity. To ensure fair comparison, the same computational budget is used for is used for
each network in this study. The problems are solved using different population
each network in this study. The problems are solved using different population sizes, withsizes, with
1010runs
runsforforeach
eachnetwork
networkininthree
threegroups,
groups,totaling
totaling3030runs
runsfor
foreach
eachproblem.
[Link]
Theresults
results
from each run are combined and sorted to obtain the final non-dominated Pareto front.
from each run are combined and sorted to obtain the final non-dominated Pareto front.
The benchmark problems are of varying complexity, and therefore different computational
The benchmark problems are of varying complexity, and therefore different computa-
budgets are required for different cases. This approach helps to ensure that the results are
tional budgets are required for different cases. This approach helps to ensure that the re-
reliable and comparable across different networks and problems.
sults are reliable and comparable across different networks and problems.
In the paper of [26], the Pareto front for each multi-objective optimization problem was
In the paper of [26], the Pareto front for each multi-objective optimization problem
obtained by collecting raw data reported by each multi-objective evolutionary algorithm
was obtained by collecting raw data reported by each multi-objective evolutionary algo-
(MOEA) for 30 runs. Duplicates in the dataset of each group, which were obtained using a
rithm (MOEA) for 30 runs. Duplicates in the dataset of each group, which were obtained
specific population size, were checked, and removed. The data from different groups were
using a specific population size, were checked, and removed. The data from different
then merged and duplicates were checked and removed once again. The non-dominated
groups were then merged and duplicates were checked and removed once again. The non-
sorting procedure was applied to the aggregated dataset to produce the best Pareto front
dominated sorting procedure was applied to the aggregated dataset to produce the best
obtained by the current MOEA.
Pareto front obtained by the current MOEA.
For each problem, the Pareto front obtained by each MOEA was first aggregated, and
duplicates in the merged dataset were checked and removed. The non-dominated sorting
procedure was then used to generate the best-known Pareto front of the current problem
Water 2023, 15, 2561 10 of 17

For each problem, the Pareto front obtained by each MOEA was first aggregated, and
duplicates in the merged dataset were checked and removed. The non-dominated sorting
procedure was then used to generate the best-known Pareto front of the current problem
that is presented in [26] and it will be referred in this study as all evolutionary algorithms
(AEAs). Lastly, the contribution from each MOEA to the best-known Pareto front was
identified. This process ensured that the AEA was obtained by considering the results from
all MOEAs and removing duplicates.
The Best-Known Pareto Front (BKPF) solutions for these problems, which are pre-
sented on the Exeter university website “[Link]
resources/benchmarks/pareto/ (accessed on 6 July 2023)”, were obtained through a sec-
ondary stage of extensive computation on the Pareto front initially obtained by five state-
of-the-art MOEAs. For the smaller networks, the true Pareto front was obtained through
full enumeration, while for the larger networks, the goal was to approximate the true
Pareto fronts.
The improved version of NSGA II is used to generate a Pareto front. The dataset from
10 independent runs with the same population size is collected, and the results are merged
to remove any duplicates. Next, the non-dominated sorting method is applied to select
the best solutions and form the Pareto front of the group. This process is repeated for each
population size to obtain the Pareto front for each group. Finally, the Pareto front for each
population size (3 groups) is merged, and any duplicates are removed. The non-dominated
sorting method is then applied to the merged solutions to produce the final Pareto front.

3. Results
Water 2023, 15, x FOR PEER REVIEW 1
Water 2023, 15, x FOR PEER REVIEW The proposed methodology was tested on benchmark case studies using the same 11 of 17
optimization model and hydraulic simulator to enable fair comparison. Table 2 presents the
results obtained from 30 runs using the improved NSGA II and compares them to the three
different resultscomprehensive
for the five caseevaluation,
studies (CS). The
it is first settoofdetermine
possible solutions, the
as described in [26],
effectiveness and efficie
comprehensive evaluation, it is possible to determine the effectiveness and efficiency of
consists of the the
results obtained
improved by five
NSGA II state-of-the-art
algorithm in solvingMOEAs thewithin the computational
benchmark problems, considerin
the improved NSGA II algorithm in solving the benchmark problems, considering the
budget. By combining
quality ofand theeliminating duplicates
solutions obtained byfrom these
the five results, the values
state-of-the-art MOEAs for and
the the ext
quality
AEA of the
column were solutions obtained
determined. by the five
The invested
second sourcestate-of-the-art
of solutions isMOEAs and the extensive
the Best-Known Pareto
computational effort in obtaining the BKPFs. This approach helps to pro
computational
Front effortfor
(BKPF) solutions invested
these in obtaining
problems, which the BKPFs. This approach helpscalculations
to provide a
more accurate assessment of the were obtained
effectiveness afterefficiency
and extensive of the improved NS
more accurate
on the raw solutionsassessment
initially of the
obtained effectiveness and efficiency of the improved NSGA II
algorithm in solving theby the same five
benchmark MOEAs as in the AEA.
problems.
algorithm in solving the benchmark problems.
Table 2. Results of 2.
Table the improved
Results of the NSGA
improvedII and
NSGA theII and
five the
MOEAS in [26]inCS
five MOEAS [26]=CScase study;
= case study; CWS =
Table 2. Results of the improved NSGA II and the five MOEAS in [26] CS = case study; CWS = center
CWS = center forfor water systems; BKPF = best known pareto front; AEA = all the evolutionary
water systems; BKPF = best known pareto front; AEA = all the evolutionary algorithms; T
for water systems; BKPF = best known pareto front; AEA = all the evolutionary algorithms; TNDS =
algorithms; TNDS total
= non-dominated solutions
total non-dominated from improved
solutions from improvedNSGA NSGA
II; ENDS II;= ENDS
equal to= BKPF;
equal DS
to = domina
total non-dominated solutions from improved NSGA II; ENDS = equal to BKPF; DS = dominated by
BKPF; DS = BKPF solutions;
dominated by BKPF NDS = solutions
solutions; NDS that
= are non-dominated
solutions that are to BKPF; BNDS =to
non-dominated solutions
BKPF; that dom
BKPF solutions; NDS = solutions that are non-dominated to BKPF; BNDS = solutions that dominated
BNDS by
= solutions BKPF, TF
that Front. = True Front.
dominated by BKPF, TF = True Front.
by BKPF, TF = True
Wang et al. [26] Results Improved NSGA II-Results
CWS Wang et al.
Wang et al.[26]
[26]Results
Results ImprovedNSGA
Improved NSGA II-Results
II-Results
CWS
CWS CS ɛ-NSGA-
CSCS NSGA-BKPF ɛ-NSGA-
NSGA-II ɛ-MOEA
-NSGA- AM- AMAGAM Borg AEA TNDS ENDS DS NDS B
BKPF NSGA-II
BKPF ɛ-MOEA
-MOEA AMAGAM
II Borg AEA
Borg AEA TNDS
TNDS ENDS ENDS DS DS NDS NDSBNDS BNDS
II IIII AGAM
TLN 114 TF 77 64 64 76 65 77 114 114 0 0
TLN 114
TLN 114TFTF 77
77 64
64 6464 76
76 65 65 7777 114
114 114
114 00 0 0 0 0
HAN 575 39 8 9 35 10 39 716 10 115 664
HAN 575
HAN 575 39
39 88 99 35
35 10 10 3939 716
716 10
10 115 664664 40 40
115
GOY 489 29 2 39 57 15 67 341 303 97 27
GOY
GOY
489 29
489 FOS 29 474
2
2 48
3939
57
57
15
15 21 67
67 341 303 97 27 11 11
10 42 31 341 140303 53297 327 58 39
FOS 474 48 10 42 21 31 140 532 3 58 39 490
FOS 474 PES 48 782 10 82 42
58 21
24 31 41 140 49 532 215 3 24758 1 39 490
437 100
PES 782 82 58 24 41 49 215 247 1 437 100 146
PES 782 82 58 24 41 49 215 247 1 437 100 146
In the benchmark case studies, the first set of case studies, TLN, the true Pareto
In the benchmark case studies, the first set of case studies, TLN, the true Pareto solu-
tion represents the Pareto front that was obtained by enumerating the entire solution
tion represents
The the Pareto
last five columns offront
Tablethat was obtained
2 contain by enumerating
the results obtained by the theimproved
entire solution
NSGA space
II
of the problem. Thus, no better solution can be found for this network. With the pro
of the problem.
algorithm using NFEThus, no better
(number ofsolution
functioncan be found for
evaluations) this network.
limitations. The With
numberthe of
proposed
solu-
improved NSGA II algorithm, all the true Pareto front points were found within the
improved
tions obtainedNSGA II algorithm,
by comparing the all the true NSGA
improved Pareto II
front pointswith
solutions wereBKPF
foundwere
within the com-
recorded,
putational budget. On the other hand, the original NSGA II algorithm only finds 77
putational budget. On the other hand, the original NSGA II algorithm only
including the number of solutions that are equal to BKPF (ENDS), the number of solutions finds 77 points
out of the 114 Pareto front solutions, even when combining all the five MOEAs, the
out of the
obtained by 114
the Pareto
improved frontNSGA
solutions, even
II that arewhen combining
dominated by the allBKPF
the five MOEAs,
solutions the the
(DS), num-
ber of solutions found is still lower than that of the BKPF. With the improvements
ber of solutions found is still lower than that of the BKPF. With the improvements made
to the NSGA II algorithm, all the possible solutions, the true Pareto front solutions
to the NSGA II algorithm, all the possible solutions, the true Pareto front solutions, were
found. In terms of dominance, the DS (dominated solutions), NDS (nondominated
found. In terms of dominance, the DS (dominated solutions), NDS (nondominated solu-
tions), and BNDS (solutions dominated by BKPF) are all equal to zero because the
tions), and BNDS (solutions dominated by BKPF) are all equal to zero because there are
no solutions that can dominate the true Pareto front. Furthermore, since all the sol
Water 2023, 15, 2561 11 of 17

number of solutions that dominate BKPF (BNDS), the number of solutions obtained by
the improved NSGA II that are nondominated to the BKPF (NDS), and the total number
of nondominated solutions obtained by the improved NSGA II (TNDS) for each of the
case studies. The aim of these comparisons is to evaluate the effectiveness of the proposed
improvement to the NSGA II algorithm in obtaining a good approximation of the true
pareto front with the same function evaluations than other state-of-the-art MOEAs.
In the context of evaluating the performance of the proposed improved NSGA II
algorithm, it is crucial to consider the benchmark Pareto fronts solution, which serves as
a reference for the quality of the results obtained. Therefore, the results obtained by the
improved NSGA II will be compared with its predecessor, the original NSGA II, using the
same computational budget. This comparison is the primary point of comparison, as it
helps determine if the improved NSGA II outperforms the original NSGA II under the
same computational budget. If the improved NSGA II can outperform the original NSGA
II, it would be a significant improvement.
After comparing the results of the improved NSGA II with the original NSGA II, the
algorithm’s performance will be further evaluated by comparing it to AEA and BKPF. AEA
and BKPF are both combinations of the five MOEAs, but it is essential to note that the
BKPFs are obtained with a much larger computational budget than the number of function
evaluations (NFEs) used for the improved NSGA II. By conducting this comprehensive
evaluation, it is possible to determine the effectiveness and efficiency of the improved
NSGA II algorithm in solving the benchmark problems, considering the quality of the
solutions obtained by the five state-of-the-art MOEAs and the extensive computational
effort invested in obtaining the BKPFs. This approach helps to provide a more accurate
assessment of the effectiveness and efficiency of the improved NSGA II algorithm in solving
the benchmark problems.
In the benchmark case studies, the first set of case studies, TLN, the true Pareto
solution represents the Pareto front that was obtained by enumerating the entire solution
space of the problem. Thus, no better solution can be found for this network. With the
proposed improved NSGA II algorithm, all the true Pareto front points were found within
the computational budget. On the other hand, the original NSGA II algorithm only finds
77 points out of the 114 Pareto front solutions, even when combining all the five MOEAs,
the number of solutions found is still lower than that of the BKPF. With the improvements
made to the NSGA II algorithm, all the possible solutions, the true Pareto front solutions,
were found. In terms of dominance, the DS (dominated solutions), NDS (nondominated
solutions), and BNDS (solutions dominated by BKPF) are all equal to zero because there are
no solutions that can dominate the true Pareto front. Furthermore, since all the solutions
are found, there are no solutions that are dominated by BKPF.
The performance of the improved NSGA II algorithm was evaluated on two medium-
sized networks: HAN and GOY. For the HAN network, the BKPF was found to be 575, and
using the improved NSGA II algorithm, a total of 716 (TDNS) solutions were found. Among
these 716 solutions, 10 are equal to the BKPF and the non-dominated solutions found by
the improved NSGA II algorithm were 664, which is higher than the BKPF. Additionally,
40 solutions that dominate the BKPF were found by the improved NSGA II algorithm,
which is higher than the 39 solutions found by the original NSGA II and AEA. For the GOY
network, the improved NSGA II algorithm found a total of 341 solutions (TDNS). Among
these solutions, 303 were equal to the BKPF and 27 non-dominated solutions compared to
the BKPF were found. The number of solutions found by the improved NSGA II algorithm
is much higher than the solutions found by the original NSGA II algorithm, which is only
29 solutions, and AEA, which found 67 solutions. The improved NSGA II algorithm was
able to find 11 solutions that dominate the BKPF. The improved NSGA II algorithm found
solutions lower than the BKPF, but it is important to note that the BKPF was obtained
through an extensive computational effort that utilized a combination of five different
state-of-the-art multi-objective evolutionary algorithms (MOEAs).
Water 2023, 15, 2561 12 of 17

The FOS and PES networks are two intermediate-sized cases with higher solution
spaces compared to previous cases. For the FOS network, the algorithm found 532 so-
lutions, which is higher than the 474 solutions found by the BKPF and 91% higher than
the 48 solutions found by the original NSGA II and 140 solutions found by the AEA (a
combination of the five MOEAs). The solutions that are equal to the BKPF are only three of
the solutions found, and only 39 were non-dominated solutions. However, the solutions
that dominate the BKPF compromise 490 solutions. This suggests that from the solutions
found by the improved NSGA II, there are significantly more solutions that dominate
the BKPF.
Another intermediate network is PES with three sources, which increases the difficulty
of solving the optimization. The improved NSGA II found 247 solutions, which is higher
than the 82 solutions found by the original NSGA II and 215 AEA solutions. We found
100 non-dominated solutions compared to BKPF solutions, and 146 solutions dominated
the BKPF. From the solutions found, 59% dominated the BKPF solutions and most of the
solutions found are better.
To evaluate the quality of the solutions obtained by the improved NSGA II algorithm,
the solutions are compared with the benchmark Pareto fronts (BKPFs). In some cases, parts
of the front are zoomed in to highlight the relationship between these two sets of solutions.
The graphs presented below depict the results from the improved NSGA II algorithm
with the new methods and the best-known Pareto front. The blue points on the graphs
represent the best-known Pareto front of networks presented in [26]. The yellow points
represent the solutions found by the improved NSGA II algorithm with the new methods
(PF). The red points indicate the total number of non-dominated solutions compared to
the BKPF.
Figure 6a displays the result of the TLN network obtained by utilizing the improved
NSGA II algorithm with the new methods and comparing them to the best-known Pareto
front. It can be observed that there is a complete overlap between all the true Pareto front
points and the solutions obtained by the improved NSGA II algorithm. This is because the
improved NSGA II algorithm has successfully found all the true Pareto optimal solutions.
As can be seen in Figure 6b, for the HAN network the new method has expanded the
range of the BKPF and increased the number of solutions. Specifically, the improved NSGA
II has discovered more points with minimum resilience and cost, as highlighted in the
zoomed-in portion. In the maximum region, the highest possible resilience for HAN has
already been achieved, which is 0.3538, and further expansion towards the maximum is not
feasible. In the minimum region, 33 additional solutions have been identified. Since more
solutions than the BKPF have been uncovered, it indicates that the density of solutions in
the remaining part of the Pareto front has been increased.
Figure 6c shows the results of GOY network, illustrating that the improved NSGA II
algorithm has successfully generated solutions across the entire range of the pareto front for
the GOY network. The improved NSGA II also outperforms the original NSGA II algorithm
in terms of solution density, having found a greater number of solutions.
The results obtained from the improved NSGA II for FOS network, Figure 6d reveal
that a considerable number of solutions that dominate the BKPF are found, leading to
substantial improvements across most parts of the pareto front. Zooming in on specific
parts of the front, it becomes apparent that a significant gap exists between BKPF solutions
and the improved NSGA II solutions in the lower resilience region and the knee area,
implying that for the same cost better resilience can be achieved. A significant enhancement
is observed in the knee area, which is a crucial part of the pareto front. The fourth method
(G4) is specifically designed to address this region and generate more solutions. The
positive impact of the G4 method can be clearly observed in this network as the solutions
have demonstrated significant improvement.
Water 2023, 15, 2561 13 of 17
Water 2023, 15, x FOR PEER REVIEW 13 of 17

(a) (b)

(c) (d)

(e)
Figure 6. Pareto front results of Improved NSGA II and BKPF of: (a) TLN; (b) HAN; (c) GOY; (d)
Figure 6. Pareto front results of Improved NSGA II and BKPF of: (a) TLN; (b) HAN; (c) GOY; (d) FOS;
FOS; (e) PES.
(e) PES.
In comparison to the BKPF, the results of the improved NSGA II for PES network in
In comparison to the BKPF, the results of the improved NSGA II for PES network
Figure 6e show that there exist some uncovered sections of the front due to the lower
in Figure 6e show that there exist some uncovered sections of the front due to the lower
number of solutions found by the improved method. However, it has been observed that
number of solutions found by the improved method. However, it has been observed that
Water 2023, 15, x FOR PEER REVIEW 14 of 17
Water 2023, 15, 2561 14 of 17

the improved NSGA II has found solutions that were not discovered by the original NSGA
the
II. Inimproved NSGA II has
Figure 7 presented fromfound
[26],solutions that wereof
the contributions not discovered
each MOEA to bythe
thefinal
original NSGA
result can
II. In Figure 7 presented from [26], the contributions of each MOEA to the final result
be seen, and it is evident that the original NSGA II found solutions only in the lower can to
be
seen, and it is evident that the original NSGA II found solutions only in the lower to middle
middle part of the pareto front. On the other hand, the improved NSGA II solution indi-
part of the pareto front. On the other hand, the improved NSGA II solution indicates that
cates that solutions have been discovered in the maximum region of the front. This is par-
solutions have been discovered in the maximum region of the front. This is particularly
ticularly notable in the zoomed-in section, which is located around the knee area, where
notable in the zoomed-in section, which is located around the knee area, where densely
densely populated solutions are found, and those solutions dominate the BKPF.
populated solutions are found, and those solutions dominate the BKPF.

Figure
Figure7.7. Distribution
Distribution of non-dominated
non-dominatedsolutions
solutionsfrom
fromeach
each MOEA
MOEA in in
thethe BKPF
BKPF of PES
of PES network
network [26].
[26].
4. Discussion
Table 2 presents the results of the improved NSGA II algorithm and compares them
4. Discussion
in terms
Tableof2 the number
presents theof solutions
results obtained
of the improved for NSGA
each ofIIthe five caseand
algorithm studies. The results
compares them
demonstrate that the improved NSGA II performs well for
in terms of the number of solutions obtained for each of the five case studies. The resultsall case studies since it has
found a large number of total non-dominated solutions.
demonstrate that the improved NSGA II performs well for all case studies since it hasIn multi-objective optimization
problems,
found a largethe number
density and diversity
of total of the solutions
non-dominated across In
solutions. themulti-objective
Pareto front andoptimization
convergence
to the best possible solutions are important factors to consider.
problems, the density and diversity of the solutions across the Pareto front and conver-
In general, it was observed that the improved NSGA II algorithm discovered a greater
gence to the best possible solutions are important factors to consider.
number of solutions compared to the original NSGA II for all benchmark networks. Fur-
In general, it was observed that the improved NSGA II algorithm discovered a greater
thermore, the improved NSGA II found more solutions than the combination of five
number of solutions compared to the original NSGA II for all benchmark networks. Fur-
state-of-the-art MOEAs all networks. Even though the BKPF was discovered through exten-
thermore, the improved NSGA II found more solutions than the combination of five state-
sive computation and the use of multiple MOEAs, the improved NSGA II was able to find
of-the-art MOEAs all networks. Even though the BKPF was discovered through extensive
more solutions for the TLN, HAN, and FOS networks, which is a significant achievement
computation and the use of multiple MOEAs, the improved NSGA II was able to find
that demonstrates its effectiveness in solving multi-objective optimization problems. In
more solutions for the TLN, HAN, and FOS networks, which is a significant achievement
addition, the improved NSGA II was able to increase the range of the pareto front for
that demonstrates its effectiveness in solving multi-objective optimization problems. In
HAN and discover a significant number of solutions that dominate the BKPF, especially for
addition, the improved NSGA II was able to increase the range of the pareto front for
intermediate-sized networks, resulting in a better pareto front.
HAN and discover a significant number of solutions that dominate the BKPF, especially
Moreover, the improved NSGA II found solutions in the parts of the pareto front
for intermediate-sized
that were not found bynetworks, the original resulting
NSGA in II,athereby
better pareto
improving [Link] diversity of solutions
Moreover, the improved NSGA II found
that can be discovered. These findings suggest that the improved solutions in the parts of the pareto
NSGA front that
II algorithm is a
were not found by the original NSGA II, thereby improving
promising approach for solving multi-objective optimization problems and can be appliedthe diversity of solutions that
can
to abe discovered.
wide range ofThese findings
real-world suggest that the improved NSGA II algorithm is a prom-
applications.
ising approach for solving multi-objective
It is important to note that although the optimization
number ofproblems
solutionsand foundcanbybethe
applied to a
improved
wide
NSGA range
II is of real-world
lower than that applications.
of the BKPF for PES network, many of the solutions found are
better than the BKPF solutions. although
It is important to note that However,the number
a larger of solutionsbudget
computational found isbyrequired
the improved
for the
NSGA II is lower than that of the BKPF for PES network,
improved NSGA II algorithm to make a fair and complete comparison with the many of the solutions found
BKPF are
in
better than the BKPF solutions. However, a larger computational
terms of the number of solutions. In summary, the improved NSGA II algorithm shows budget is required for
the improved
promising NSGA
results in II algorithm
solving to make a fair and
intermediate-sized complete
problems suchcomparison
as FOS andwith PES the BKPF
networks,
in terms
and furtherof the number of solutions.
improvements In summary,
can be made by increasingthe improved NSGA II algorithm
the computational budget. shows
promising resultsof
The results in the
solving intermediate-sized
improved NSGA II algorithm problems such as FOS
presented andstudy
in this PES networks,
are highly
and furtherand
promising improvements
warrant further can be made by increasing
exploration in the fieldthe computational optimization.
of multi-objective budget. Al-
The results of the improved NSGA II algorithm presented
though the algorithm has demonstrated success in solving a two-objective problem, there is in this study are highly
promising and warrantthrough
room for improvement further further
exploration
parameterin thecalibration.
field of multi-objective optimization.
Future work should focus on
Although
fine-tuning thethealgorithm
implementation has demonstrated
parameters to success
optimizein solving a two-objective
the algorithm’s performanceproblem,
on all
there is roomnetworks.
benchmark for improvement
This canthrough further
be achieved parameter
through calibration.
extensive Future
testing and work should
experimentation,
focus
leading onto fine-tuning
a more efficient the and implementation
effective algorithm. parameters to optimize the algorithm’s
Water 2023, 15, 2561 15 of 17

Furthermore, it is recommended that additional tests be conducted on the remaining

benchmark networks to further evaluate the algorithm’s capabilities and limitations. These
tests will allow researchers to gain a deeper understanding of the algorithm’s performance
across a wider range of problem sets. In addition, further research could be conducted to
define a new best-known Pareto front through extensive computation.
This study compared the solutions obtained using the improved NSGA II algorithm
with the original NSGA II and AEA in terms of the number of solutions and graphical
representation. However, for a more comprehensive analysis, it may be useful to evaluate
these solutions using additional metrics such as data dispersion. To achieve a better
comparison, future research could consider using alternative metrics such as box plots to
provide insight into the spread of solutions in the front.

5. Conclusions
This research presents an improved version of the NSGA II algorithm that can effec-
tively tackle multi-objective optimization problems. The NSGA II algorithm is a well-
established state-of-the-art approach that has been widely used in the field of multi-
objective optimization. The improved NSGA II algorithm employs various generation
processes motivated by MOSA-GR to enhance the optimization search. The algorithm
is designed to target different regions of the Pareto front to achieve better convergence,
diversity, and density. The four generation processes, namely G1, G2, G3, and G4, serve
different purposes, such as increasing diversity, searching for specific parts of the Pareto
front, and finding solutions in the knee area. By utilizing different generation processes,
the algorithm can access parts of the Pareto front that were previously unobtainable using
a single generation process.
The improved NSGA II algorithm was tested on a well-known two-objective WDN
problem from the literature, and the results were compared with those obtained by the
original NSGA II, the combination of five state-of-the-art MOEAs and with the best know
pareto front. The comparison demonstrated that the proposed improvement to the NSGA
II algorithm was effective in finding more non-dominated solutions and dominating more
solutions than the original NSGA II algorithm and other MOEAs in most cases. The
algorithm was able to find all true Pareto front solutions for the small case study, widen
the range of solutions for the intermediate-sized problem (HAN), and found solutions that
dominate the best-known Pareto front, particularly in the knee area of the Pareto front for
the intermediate-sized networks.
In conclusion, the proposed modifications to the NSGA II algorithm have the potential
to be a significant contribution to the field of multi-objective optimization. The improved
NSGA II algorithm is capable of finding solutions that cover all parts of the Pareto front with
a single algorithm without increasing the computational effort. The good performance of
the improved NSGA II algorithm compared to the original NSGA II algorithm demonstrates
the effectiveness of the proposed approach.
In summary, the improved NSGA II algorithm exhibits significant promise in tackling
multi-objective optimization problems. Subsequent research endeavors should concentrate
on enhancing its implementation and assessing its performance across a broader spectrum
of problem sets, while considering larger computational resources.

Author Contributions: Conceptualization, M.C. and A.O.; methodology, R.A.K., M.C., A.O. and
E.S.; software, R.A.K., M.C., A.O. and E.S.; validation, A.O., M.C. and E.S.; writing—original draft
preparation, R.A.K.; writing—review and editing, R.A.K., M.C., A.O. and E.S.; visualization, R.A.K.,
M.C., A.O. and E.S.; supervision, M.C., A.O. and E.S.; funding acquisition, A.O. All authors have
read and agreed to the published version of the manuscript.
Funding: This research was supported by the Grand Water Research Institute at the Technion–Israel
Institute of Technology.
Data Availability Statement: The data presented in this study are available on request from the
corresponding author.
Water 2023, 15, 2561 16 of 17

Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design
of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or
in the decision to publish the results.

References
1. Filion, Y.R.; MacLean, H.L.; Karney, B.W. Life-Cycle Energy Analysis of a Water Distribution System. J. Infrastruct. Syst. 2004, 10,
120–130. [CrossRef]
2. Kleiner, Y.; Adams, B.J.; Rogers, J.S. Water Distribution Network Renewal Planning. J. Comput. Civ. Eng. 2001, 15, 15–26.
[CrossRef]
3. Cunha, M.D.C.; Sousa, J. Water Distribution Network Design Optimization: Simulated Annealing Approach. J. Water Resour. Plan.
Manag. 1999, 125, 215–221.
4. Zecchin, A.C.; Simpson, A.R.; Maier, H.R.; Leonard, M.; Roberts, A.J.; Berrisford, M.J. Application of two ant colony optimisation
algorithms to water distribution system optimisation. Math. Comput. Model. 2006, 44, 451–468. [CrossRef]
5. Bi, W.; Chen, M.; Shen, S.; Huang, Z.; Chen, J. A Many-Objective Analysis Framework for Large Real-World Water Distribution
System Design Problems. Water 2022, 14, 557.
6. Wu, W.; Maier, H.R.; Simpson, A.R. Multiobjective optimization of water distribution systems accounting for economic cost,
hydraulic reliability, and greenhouse gas emissions. Water Resour. Res. 2013, 49, 1211–1225. [CrossRef]
7. Awe, O.M.; Okolie, S.T.A.; Fayomi, O.S.I. Optimization of Water Distribution Systems: A Review. J. Phys. Conf. Ser. 2019, 1378,
022068. [CrossRef]
8. Farmani, R.; Savic, D.A.; Walters, G.A. Evolutionary multi-objective optimization in water distribution network design. Eng.
Optim. 2005, 37, 167–183. [CrossRef]
9. Prasad, T.D.; Park, N.-S. Multiobjective genetic algorithms for design of water distribution networks. J. Water Resour. Plan. Manag.
2004, 130, 73–82.
10. Prasad, T.D.; Hong, S.-H.; Park, N. Reliability based design of water distribution networks using multi-objective genetic
algorithms. KSCE J. Civ. Eng. 2003, 7, 351–361. [CrossRef]
11. Saldarriaga, J.; Takahashi, S.; Hernández, F.; Escovar, M. Multi-objective water distribution system design using an expert
algorithm. In Proceedings of the Urban Water Management Challenges Oppurtunities—11th International Conference on
Computing and Control for the Water Industry, CCWI 2011, Exeter, UK, 5–7 September 2011; Volume 3.
12. Todini, E. Looped water distribution networks design using a resilience index based heuristic approach. Urban Water 2000, 2,
115–122. [CrossRef]
13. Prasad, T.D.; Tanyimboh, T.T. Entry Bassed Design of “Anytown” Water Distribution Network; Kruger: Montreal, QC, Canada, 2008.
14. Maier, H.R.; Kapelan, Z.; Kasprzyk, J.; Kollat, J.; Matott, L.S.; Cunha, M.C.; Dandy, G.C.; Gibbs, M.S.; Keedwell, E.; Marchi,
A.; et al. Evolutionary algorithms and other metaheuristics in water resources: Current status, research challenges and future
directions. Environ. Model. Softw. 2014, 62, 271–299. [CrossRef]
15. Nicklow, J.; Asce, F.; Reed, P.; Asce, M.; Savic, D.; Dessalegne, T.; Harrell, L.; Chan-Hilton, A.; Karamouz, M.; Minsker, B.; et al.
State of the Art for Genetic Algorithms and Beyond in Water Resources Planning and Management ASCE Task Committee on
Evolutionary Computation in Environmental and Water Resources Engineering. J. Water Resour. Plan. Manag. 2010, 136, 412–432.
16. Zhou, A.; Qu, B.Y.; Li, H.; Zhao, S.Z.; Suganthan, P.N.; Zhang, Q. Multiobjective evolutionary algorithms: A survey of the state of
the art. Swarm Evol. Comput. 2011, 1, 32–49. [CrossRef]
17. Wang, M.; Dai, G.; Hu, H. Improved NSGA-II algorithm for Optimization of Constrained Functions. In Proceedings of the 2010
International Conference on Machine Vision and Human-Machine Interface, Kaifeng, China, 24–25 April 2010; pp. 673–675.
[CrossRef]
18. D’Souza, R.G.L.; Sekaran, K.C.; Kandasamy, A. Improved NSGA-II Based on a Novel Ranking Scheme. J. Comput. Civ. Eng. 2010,
2, 91–95.
19. Verma, S.; Pant, M.; Snasel, V. A Comprehensive Review on NSGA-II for Multi-Objective Combinatorial Optimization Problems.
IEEE Access 2021, 9, 57757–57791. [CrossRef]
20. Liu, T.; Gao, X.; Wang, L. Multi-objective optimization method using an improved NSGA-II algorithm for oil-gas production
process. J. Taiwan Inst. Chem. Eng. 2015, 57, 42–53. [CrossRef]
21. Wang, Q.; Wang, L.; Huang, W.; Wang, Z.; Liu, S.; Savić, D.A. Parameterization of NSGA-II for the optimal design of water
distribution systems. Water 2019, 11, 971. [CrossRef]
22. Deb, K.; Pratap, A.; Agarwal, S.; Meyarivan, T. A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE Trans. Evol.
Comput. 2002, 6, 182–197. [CrossRef]
23. Kapelan, Z.S.; Savic, D.A.; Walters, G.A. Multiobjective design of water distribution systems under uncertainty. Water Resour. Res.
2005, 41, 1–15. [CrossRef]
24. Deb, K.; Jain, H. An evolutionary many-objective optimization algorithm using reference-point-based nondominated sorting
approach, Part I: Solving problems with box constraints. IEEE Trans. Evol. Comput. 2014, 18, 577–601. [CrossRef]
25. Mohamad Shirajuddin, T.; Muhammad, N.S.; Abdullah, J. Optimization problems in water distribution systems using Non-
dominated Sorting Genetic Algorithm II: An overview. Ain Shams Eng. J. 2022, 14, 101932. [CrossRef]
Water 2023, 15, 2561 17 of 17

26. Wang, Q.; Guidolin, M.; Dragan, S.; Zoran, K. Two-Objective Design of Benchmark Problems of a Water Distribution System via
MOEAs: Towards the Best-Known Approximation of the True Pareto Front. J. Water Resour. Plan. Manag. 2015, 141, 04014060.
[CrossRef]
27. Cunha, M.; Marques, J. A New Multiobjective Simulated Annealing Algorithm—MOSA-GR: Application to the Optimal Design
of Water Distribution Networks. Water Resour. Res. 2020, 56, e2019WR025852. [CrossRef]
28. Bin Mahmoud, A.A.; Piratla, K.R. Comparative evaluation of resilience metrics for water distribution systems using a pressure
driven demand-based reliability approach. J. Water Supply Res. Technol. 2018, 67, 517–530. [CrossRef]
29. Zhan, X.; Meng, F.; Liu, S.; Fu, G. Comparing Performance Indicators for Assessing and Building Resilient Water Distribution
Systems. J. Water Resour. Plan. Manag. 2020, 146, 06020012. [CrossRef]
30. Jayaram, N.; Srinivasan, K. Performance-based optimal design and rehabilitation of water distribution networks using life cycle
costing. Water Resour. Res. 2008, 44, 1417. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.