Energy Systems

https://doi.org/10.1007/s12667-022-00521-8

ORIGINAL PAPER

An intelligent water drop approach for simultaneous

reconfiguration and DG integration in distribution system

Nimish Bhatt1 · Ashwani Kumar Chandel1

Received: 16 December 2021 / Accepted: 12 May 2022

© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2022

Abstract
The inclusion of distributed generator units (DGs) has significantly aided in address-
ing the exponentially incrementing electric power demand, reducing the losses, and
enhancing the voltage profile of the distribution system. However, the DGs must be
placed at optimal locations with appropriate outputs to extract the maximum bene-
fits. Typically, classical optimization techniques are employed to size DGs for small
and simple systems; nevertheless, large and complex distribution systems require
intelligent approaches for optimal sizing of DGs. Consequently, the present work
utilizes a nature-inspired optimization algorithm known as intelligent water drop to
find the optimal locations and sizes of the DGs. Additionally, the optimal sizing of
the DGs for reconfigured distribution systems is also performed to upgrade the sys-
tem performance further. The algorithm has been implemented on IEEE 33 bus and
IEEE 69 bus distribution systems. The results explicitly demonstrate that the pro-
posed algorithm is superior in addressing the problem of reconfiguration combined
with DGs installation.

Keywords Distribution system · Distributed generator units · Intelligent water-

droplet algorithm · Reconfiguration

1 Introduction

The distribution system (DS) is the last functional part of the power system as the
voltage is stepped down at lower values and fed to industrial and domestic consum-
ers. To employ a simpler protection scheme, the DSs are generally radial, have a
higher R∕X ratio, and are fed from a single source. Owing to these attributes, the DS
accounts for a higher amount of power loss; in particular, it shares nearly 70% of the
power loss of the complete power system. Further, the progressive upsurge in the

* Nimish Bhatt
[email protected]
1
Department of Electrical Engineering, National Institute of Technology, Hamirpur,
Himachal Pradesh, India

13
Vol.:(0123456789)
 N. Bhatt, A. K. Chandel

load demand has imposed a challenge on the utility operators for a reliable and effi-
cient operation of the DS [1]. Consequently, to tackle these operational conditions,
the utilities are bound to assess the loadability margin of the DS to avert the risk of
voltage collapse. Therefore, the utilities have now focused on exploiting the prevail-
ing infrastructure’s effective usage in a well-planned manner. Over the years, recon-
figuration has been employed as a competent tool to handle the problems of the DS,
such as poor voltage profile and higher power losses. Additionally, the emphasis
on cutting down the emissions due to fossil fuels and the growth of the renewable-
based distributed generation units (DGs) has further encouraged its application in
improving the performance of the DS [2].
The DS reconfiguration (DSR) refers to an arrangement of the topology of the
DS by operating the sectionalizers and tie-switches to optimize the APL and simul-
taneously meet out the operational constraints [3]. It is employed with APL mini-
mization and voltage profile as constraints [4]. Further, it should be ensured that
the radiality of the DS remains unchanged irrespective of the new topology along
with the condition of meeting associated load. The first attempt in solving the prob-
lem of DSR was the branch and bond technique suggested by Merlin and Back [5].
Further, the branch exchange technique[6] and the improved branch and bond tech-
nique [7] were suggested by the authors for DSR. Although the methods are easy
to implement, these do not lead to optimal solutions. Moreover, the computational
time is also a big concern with these methods. Later the authors in [8] addressed the
problem of DSR by applying the two-minimum current neighbour chain updating
methods. In [9], the authors introduced a new approach based on benders decompo-
sition for DSR combined with optimal power flow. Later the authors in [10] advo-
cated using a differential evolution algorithm to solve the DSR and mitigate the
power quality issues. Afterwards, several metaheuristic-based techniques have been
applied to solve the problem of DSR. The methods are quick to find a solution with
easier implementation. Some of metaheuristic techniques include genetic algorithm
[11, 12], Binary Particle Swarm Optimization (BPSO) algorithm [13, 14], Harmony
search algorithm (HSA) [15], Cuckoo search algorithm (CSA) [16, 17] fireworks
algorithm (FWA) [18], and root runner algorithm [19].
With the deregulation of the power system, gigantic requirement of power, and
limitations in transmission capacities, numerous researchers suggested the integra-
tion of capacitors, D-STATCOM, and DGs to reinforce the performance of the DS.
Primarily the DGs are preferably integrated with the DS owing to the advantage
of providing active and reactive power support. This integration relives the over-
burdened power system and imparts numerous benefits such as APL minimization,
voltage profile enhancement, and reliability of the DS. Regardless of the gains, the
unfitting size and location of the DG lead to contrasting results, such as increased
losses and reduction of reliability. Hence, it is vital to select the precise location and
suitable size of the DG. Impacts of the DGs are explicitly explored in [20]. Since
the allotment of the DGs exhibits mixed-integer non-linear nature, a variety of opti-
mization techniques implemented by the researchers are quantified in the present
literature.
PSO has been utilized to obtain the optimal size of DG for the DS to minimize
the APL [21]. CSA is employed to cut down the APL in the DS for non-dispatchable

13
An intelligent water drop approach for simultaneous…

integrated DGs [22]. Likewise, evolutionary programming has been employed to

locate the DGs in a precise manner [23]. In [24], the researchers proposed a Bat-
inspired algorithm that deals with optimal sizing of DGs for radial as well as meshed
networks. Some researchers have focused on the sizing of DG to enhance the volt-
age stability margin based on the current conditions of utilities. In [25], the authors
have utilized the concept of voltage stability index (VSI) to determine the suitable
location, and the FWA has been considered for optimizing the size of DG. Nonethe-
less, the algorithm has a slow convergence speed, which is not recommended when
the system is large and complex. The authors in [26] have employed the VSI and
bus participation factor together to determine the stability and location of DG. An
improved imperialistic competitive algorithm (ICA) that determines the optimal
size of the DG based on the reduction of APL and improvement in voltage stability
margin is implemented [27]. Esmaili [28] has formulated a fuzzified multi-objective
approach to optimize the voltage stability margin, number of DGs, and APL. Later,
a novel strategy that encompasses power loss sensitivity, power stability index, and
distinct VSI to optimize DG location and sizing for increment in load has been pro-
posed by Murthy and Kumar [29].
In both previous types of research, the emphasis is only on loss minimization,
which has been achieved either by reconfiguration or employing appropriate pow-
ered and located DG. Nevertheless, the new topology achieved after reconfigura-
tion of the DS sharply affects the placement and dimension of the DG. Addition-
ally, the amalgamation of reconfiguration and DG integration further complicates
the problem.
Therefore, to deal with the complex problem of DSR combined with DG instal-
lation, the researchers have exploited the metaheuristic techniques. In ref. [30],
the authors suggested HSA to reconfigure the DS and optimize the size of the DG.
Additionally, sensitivity analysis has been used to identify the suitable bus for the
placement of the DG. However, HSA suffers from premature convergence, which
results in a suboptimal solution. The authors in [31] have implemented the Big-Bang
Big-Crunch algorithm for DG allotment and reconfiguration. The method handles
both the problems, although the technique has been applied on the IEEE 25 bus and
IEEE 33 bus system. Recently the researchers advocated new techniques such as
Salp Swarm optimization algorithm (SSA) [29], Grey-Wolf optimizer combined
with PSO (GW-PSO) [32], and Stochastic Fractal search algorithm (SFS) [33]. SSA
depicts the swarm formation behaviour of salps to fetch solutions for every run, yet
premature convergence is possible when the objective function becomes complex
[34]. GW-PSO is superior to PSO, HSA, FWA CSA, etc., but striking an appropriate
balance between exploration and exploitation of solution is a difficult task with the
increment in the complexity of the objective function [35].
It is to be noted that a metaheuristic algorithm may yield better results for recon-
figuration but unsatisfactory output for DG sizing or vice versa. Hence, the selec-
tion of a proper algorithm that handles both the problems in an appropriate manner
becomes a tricky task. In the present work, the problem of optimal configuration of
DS and appropriate size of DG has been dealt with by an intelligence-based algo-
rithm known as the intelligent water drop algorithm (IWDA). IWDA portrays the
nature of the flow of water in the river, which picks out an easier path to flow from

13
 N. Bhatt, A. K. Chandel

the available routes to reach from starting to the endpoint. IWDA has proved its
worth by successfully solving the knapsack and travelling salesmen problem. This
algorithm has well-handled the issue of global optimization (e.g., robot path plan-
ning). Besides, the problem of economic load dispatch in the power system has been
efficiently addressed by the IWDA. Less computational time and capability of deter-
mining the viable global optimal solution are the highlights of this algorithm. The
IWDA has been used for appropriate sizing of DG; however, no such literature is
available that employs the algorithm for reconfiguring DS and sizing DG concur-
rently. Hence, the present paper showcases the performance of the IWDA for both
the cases mentioned above. Moreover, the algorithm has been applied for recon-
figuration, DG installation, DG installation after reconfiguration to demonstrate its
robustness.
The organization of the rest of the article is as follows: The problem formulation
has been illustrated in Sect. 2. The idea of sensitivity which helps identify the proper
location of buses in DS to mount the DG, is also discussed in the section. Further,
the Bus injection to branch current and branch current method to bus voltage (BIBC-
BCBV) is illustrated. The IWDA has been described in Sect. 3. Implementation and
scrutinization of the proposed method are elaborated in Sect. 4. The contribution of
the algorithm and conclusion of the article is present in Sect. 5.

2 Problem formulation

The proposed algorithm aims to reconfigure the DS and assign the appropriate
DGs that aid in minimizing the APL occurring in the distribution network. In other
words, the aim is to maximize power loss reduction by reconfiguring the DS and
allocating the appropriate DGs. Hence, the objective function is as follows:
Maximize P = Maximize (ΔPloss )
Reconfig
ΔPloss = PDG
loss
+ Ploss (1)

Subjected to:
(
Vmin ≤ Vn ≤ Vmax
|Ik | ≤ Ik_ max
| |
where ΔPDG is the reduction of power loss by appropriate allocation of DG and
Reconfig loss
ΔPloss is the reduction of power loss by reconfiguring the network Ik and Vk are
the current and voltage values of kth bus respectively. Vmax and Vmin are the upper
and lower limits of voltage on kth bus. Ik_ max is the maximum permissible current
value in the kth bus.

2.1 Sensitivity analysis

To account for the impact on the performance of the system due to the variation
of parameters associated with the DS, sensitivity analysis is exceedingly helpful in

13
An intelligent water drop approach for simultaneous…

assessing the system condition. Additionally, the changes in bus voltages, losses,
branch current, variation of generation, and loads are also detected by determin-
ing the sensitivity. The problem of the suitable location of the capacitor for reactive
power compensation is solved by selecting the real power loss sensitivity concerning
reactive power injection at the bus.
The heuristic-based sensitivity analysis helps cut down the search space by pick-
ing out the most sensitive buses. Sensitivity analysis comprises active power loss
sensitivity (APLS) and reactive power loss sensitivity (RPLS). APLS is obtained by
calculating the real power loss corresponding to active power injection at each bus,
while RPLS is found by computing the APL for reactive power injection. The for-
mer is used to identify the location for assimilating the DGs, and the latter is used to
mount the capacitors in the DS.
The present paper has utilized the APLS technique to identify the optimal loca-
tion of buses. The procedure to compute the APLS is as follows:

1. Determine the real power loss with respect to APL corresponding to every bus
of the DS.
2. Organize all the buses in a downward sequence of their sensitivities.
3. The buses with peak values of APLS qualify for a suitable location for DG place-
ment.

The mathematical expression to compute the APLS for every bus in the DS is
given as:
𝜕Pline_loss 2Pl+1,effe × Rl,l+1
APLS(l, l + 1) =
𝜕Pl+1,effe
=
|Vl+1 |2 (2)
| |
where Pl+1,effe is the net effective active power supplied beyond lth + 1 the bus,Vl+1
corresponds to the voltage of lth + 1.

2.2 Bus injection to branch current and branch current method to bus voltage
Method

As stated in the previous section, the ratio of resistance to reactance R∕X is signifi-
cantly higher than the transmission systems, leading to an ill-conditioned admittance
matrix. As a result, conventional methods like Newton–Raphson, Gauss–Seidel,
Fast Decoupled, etc., cease to converge during the load flow solution. The conven-
tional methods are modified, or other methods like Backward/Forward sweep and
BIBC-BCBV methods are employed to solve the load flow problems of the DS. The
present article utilizes the BIBC-BCBV method for computing the load flow. BIBC-
BCBV matrix stands for Bus Injection Branch Currents and Branch Currents Bus
Voltages. The former matrix establishes the relationship between the current injec-
tion at each bus and the current flowing in the branches of the DS. The latter matrix
describes the association of branch currents with the voltages of each bus. Figure 1

13
 N. Bhatt, A. K. Chandel

Fig. 1  Single line diagram of IEEE 6 bus DS

shows a single line diagram 6 bus DS to illustrate the formation of BIBC and BCBV
matrix for IEEE 6 bus DS is as follows:
By employing KCL in Fig. 1, we get:
Br1 = I2 + I3 + I4 + I5 + I6 (3)

Br2 = I3 + I4 + I5 + I6 (4)

Br3 = I4 + I5 (5)

Br4 = I5 (6)

Br5 = I6 (7)
The association of branch currents and bus injections can be expressed in matrix
form.

⎡ Br1 ⎤ ⎡ 1 1 1 1 1 ⎤⎡ I2 ⎤
⎢ Br2 ⎥ ⎢ 0 1 1 1 1 ⎥⎢ I3 ⎥
⎢ Br3 ⎥ = ⎢ 0 0 0 1 0 ⎥⎢ I4 ⎥
⎢ ⎥ ⎢ ⎥⎢ ⎥
⎢ Br4 ⎥ ⎢ 0 0 0 1 0 ⎥⎢ I5 ⎥
⎣ Br5 ⎦ ⎣ 0 0 0 0 1 ⎦⎣ I6 ⎦

[Br] = [BIBC][I] (8)

where Br corresponds to branch current and I corresponds to current injections in
the bus.
By employing KVL in Fig. 1 we get:
V2 = V1 − Br1 Z12 (9)

13
An intelligent water drop approach for simultaneous…

V3 = V1 − Br1 Z12 − Br2 Z23 (10)

V4 = V1 − Br1 Z12 − Br2 Z23 − Br3 Z34 (11)

V5 = V1 − Br1 Z12 − Br2 Z23 − Br3 Z34 − Br4 Z45 (12)

V6 = V1 − Br1 Z12 − Br2 Z23 − Br5 Z36 (13)

The association of branch currents and bus voltages can be expressed in matrix
form

⎡ V1 ⎤ ⎡ V2 ⎤ ⎡ Z12 0 0 0 0 ⎤⎡ Br1 ⎤
⎢ V1 ⎥ ⎢ V3 ⎥ ⎢ Z12 Z23 0 0 1 ⎥⎢ Br2 ⎥
⎢ V1 ⎥ − ⎢ V4 ⎥ = ⎢ Z12 Z23 Z34 0 0 ⎥⎢ Br3 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎥
⎢ V1 ⎥ ⎢ V5 ⎥ ⎢ Z12 Z23 Z34 Z45 0 ⎥⎢ Br4 ⎥
⎣ V1 ⎦ ⎣ V6 ⎦ ⎣ Z12 Z23 Z34 0 Z36 ⎦⎣ Br5 ⎦

[ΔV] = [BCBV][Br] (14)

where V corresponds to the voltage of the bus and Z corresponds to impedance in
between the bus.

3 Intelligent water drop algorithm

3.1 Introduction

IWDA is a swarm or population-based optimization algorithm given by Shah-Hos-

seini [30]. It characterizes the flow of water in a river, where it chooses a path that
offers fewer hindrances. The flow of water has a specific speed and modifies the
overall surrounding attributes of the river that influence the course of water. IWDA
utilizes this phenomenon to develop a computational tool to solve the optimization
problem. As the water drops form elemental components of the river, similarly, the
water drops are the computational instruments to realize the IWDA. The algorithm
starts by considering water drops with some initial state that eventually follow a
fixed discretized movement. The next step includes the coordination of the water
drops among themselves, which modifies the environmental properties. These envi-
ronmental properties are:

1. Soil content in IWD.

2. The velocity of IWD.

The IWDA begins with a random number of water drops having initial velocity
with zero soil content. As the algorithm proceeds, the velocity and the soil values vary
with the movement of the water drop. The variation in the velocity is non-linear and

13
 N. Bhatt, A. K. Chandel

inversely proportional to the soil content in between the current and previous location
of the drop. The droplet also transmits soil in every movement whose accumulation
nature is non-linear. It is inversely correlated to the time lapsed in the movement of the
water drop to reach a new position from its former position. The time elapsed in the
movement of water drop and velocity is proportional to each other. Conversely, the time
taken for movement is inversely related to the distance between the two locations.

3.2 Procedure for IWDA

The primary step while utilizing the IWDA for optimization problems is to represent
the problem in the form of a wholly linked graph known as a construction graph. The
construction graph is denoted by F(U, D), where U corresponds to a set of nodes,
and D corresponds to the set of edges that links with the nodes. The number of IWD,
i.e.,MIWD and the number of nodes, are equal.

3.2.1 Initialization of IWDA

The IWD starts by initializing two parameters: static and dynamic.

3.2.2 Static parameters

Static parameters, as the name indicate that these parameters remain unchanged after
the initialization at the beginning of the algorithm. Further, these parameters are bifur-
cated as per their respective functions. The first includes the following parameters:

1. The size of the swarm or the number of water drops is denoted by MIWD.
2. The maximum number of iterations denoted by max_iter
3. The initial value of soil in IWD is marked by init_soil.

These parameters regulate the swarm dynamics of the IWDA. The second category
includes the parameters that regulate the updating of the velocity and soil throughout
the course of the search. These parameters are:

1. Soil updating parameters employed to regulate the soil function of IWDA are
denoted by as,bs and cs shown in Eq. (18).
2. Velocity updating parameters employed to regulate the velocity function of IWDA
are denoted by av,bv and cv shown in Eq. (21).

4 Dynamic parameters

These parameters are employed to regulate the construction course of the solution,
and the initialization takes place at the beginning of the algorithm. All the values of
these parameters are updated for every iteration. Dynamic parameters comprise of.

13
An intelligent water drop approach for simultaneous…

1. Drvisited corresponds to list of edges visited by water drop r.

2. init_velr corresponds to the velocity of water drop r at the initial stage.
3. soilk corresponds to the amount of soil laden by the water drop r at the initial
stage.

4.1 Steps of IWDA

The complete procedure to obtain a solution by incorporating the IWD has been
explained in the form of steps which is as follows:

1. Assign the values to both static and dynamic parameters. The soil update param-
eters are assigned with values as = 1, followed by bs = 0.001, and cs = 1. Simi-
larly, the parameters corresponding to velocity are set as av = 1, bv = 0.001 and
cv = 1. The soil available between the two nodes is updated as 𝜎m = 0.9, and
the global soil update 𝜎IWD is equal to 0.9. soili,j represents the soil present in
the consecutive paths.
2. Initialize the init_soil and Drvisited as 0.
3. To start the algorithm, the IWD is spread at random across the nodes present in
the constructed graph of the problem.
4. Since every IWD is now located at a node, therefore updating Drvisited for the cor-
responding IWD is done. The IWD located at the present node (e.g.,ith ) moves
to the node which has less soil in such a way that there is no violation of any
constraint corresponding to the given problem. The probability of the selection
of a particular node is given by:

g(soili,j )
Pri (j) = ∑ (15)
l≠Dr g(soili,l )
visited

5. The probability of relocation of IWD from the current position to a new position
is inversely proportional to the soil present between these two locations. The
constant 𝜍 = 0.001 is added to avoid the division of function g by zero.
1
g(soili,j ) = (16)
𝜍 + h(soili,j )

{
soili,j if min(soili,l ) ≥ 0
i,j
h(soil ) = i,j i,l
soil − min(soil ) otherwise (17)
l≠Drvisited

6. The IWD shall visit the nodes that have the maximum probability. Update the
set of visited nodes. The movement of IWD causes changes in velocity and soil
content which is updated using the following function.
r av r
velIWD (t + 1) =
bv + cv × (soili,j )2
+ velIWD (t) (18)

13
 N. Bhatt, A. K. Chandel

 where velIWDr
(t + 1) corresponds to the velocity of drop r at t + 1 time t. The
soil value, i.e., the soil contained in the IWD (soilr ) and the soil available in
between the two locations, are updated using the equations below:

soilr = Δsoili,j + soilr (19)

soili,j = (1− ∈) × soili,j − ∈ ×Δsoili,j (20)

 where ∈ is a positive constant whose value is more significant than zero
and lesser than one and Δsoili,j is the amount of soil removed from the edge in
between i and j respectively, which derived by the following equation:
as
Δsoili,j = (21)
bs + cs × time2 (i, j)

 where,timei,j is the time required by the rth IWD to travel from ith position
to jth position. The time taken for the movement of the drop between the two
positions is correlative to the velocity and inversely correlated to the distance.
 The timei,j is calculated using the expression shown below:

HUDi,j
timei,j = (22)
velr (i, j)

 The HUDi,j represents a heuristic function that calculates the desirability of

an IWD to relocate to a new jth location from a previous ith location. It also
represents the searchability of the algorithm at the local level.
7. Apply the above steps for all the IWD available in the system. Obtain the iter-
ation-best solution from the available solutions using the following equation:
( )
SIB = arg maxIWD
w SIWD (23)
∀S

 where SIB is the iteration best solution,SIWD is the solution obtained from
IWDA, and w(.) is a function that corresponds to the quality of the solution.
8. The global update of soil is obtained as:
( ) 1
soili,j = 1 + 𝛽IWD × soili,j − 𝛽IWD × ( ) × soilIB
r
(24)
M IB

 where 𝛽IWD is a positive constant and M IB represents the number of nodes in

the solution.
9. For every iteration the global best-solution STB is either kept same or updated
by present best-solution using:
{ TB ( )
S ifw(STB ) > w SIB
(25)
TB
S =
SIB otherwise

10. Continue with the procedure above till the iteration count is equal to max_iter.

13
An intelligent water drop approach for simultaneous…

11. The termination of the algorithm occurs when the total best-solution STB is
reached.

4.2 Procedure for IWDA for the current problem

This article utilizes the IWDA for reconfiguring DS and deriving the appropriate
size of DG. The problem of reconfiguration and optimal sizing of DG are separately
explained for better clarity (Fig. 2).

4.2.1 IWDA for reconfiguration

1. Read the system data, i.e., bus voltages, branch number, receiving and sending
buses, resistance, and reactance of the branch, active and reactive power flow
corresponding to the respective bus.
2. The BIBC-BCBV method is exploited to compute the power flow of the DS.
For reconfiguration, generate an initial set of tie switches corresponding to a
configuration of DS using IWDA. Each water drop is a potential candidate for
reconfiguration. The soil content in every water drop is the set of tie-switches
3. The probability of choosing the next set of tie-switches is determined using
Eq. (15).
4. Modify the velocity using Eq. (18). Calculate the changes in the set of tie switches
using Eq. (19). Update the tie-switches set using Eq. (20).
5. Obtaining the best solution for the current iteration employing the objective func-
tion as well as updating of tie switch set is done using Eq. (23).
6. Find the global best solution as per Eq. (25).

4.2.2 IWDA for optimal sizing of DG

1. Read the system data, i.e., bus voltages, branch number, receiving and sending
buses, resistance, and reactance of the branch, active and reactive power flow
corresponding to the respective bus.
2. The BIBC-BCBV method is exploited to compute the power flow of the DS
(including the reconfigured DS).
3. The candidate buses for mounting the DGs are identified using the sensitivity
factor as per the Eq. (2).
4. Implement the IWDA for finding the appropriate size of DG. The water drops
present in the algorithm acts as a potential candidate for the DG sizing problem.
The value of soil in each IWD represents the size of DG.
5. The soil values, i.e., the DG values, are initialized at random. The computation
of probability of selecting the following DG values is done as per Eq. (15).
6. The modification in the velocity is done using Eq. (18). Alterations in the size of
DG are computed using Eq. (19). The DG values are updated through Eq. (20).

13
 N. Bhatt, A. K. Chandel

Fig. 2  Flowchart of IWDA for reconfiguration and DG placement in DS

7. Determination of the best solution for the current iteration is done employing the
objective function of the problem and updating the DG size as per Eq. (23).
8. Find the global best solution as per Eq. (25).

13
An intelligent water drop approach for simultaneous…

5 Simulation results and discussion

The proposed IWDA is implemented on two IEEE standard test systems viz., IEEE
33 bus and IEEE 69 bus system to examine its capability to solve simultaneous DSR
and DG sizing problems. Moreover, a comparative analysis has been carried out
with the other schemes to exhibit the superiority of the algorithm. It is to be noted
that the maximum number of DGs is restricted to three as a further increase in the
number does not reduce the power loss in the DS [36]. The results and discussion
have been done in the subsequent section. MATLAB 2018a has been employed to
carry out the task of simulation. Five different entities have been simulated to dem-
onstrate the effectiveness of the projected algorithm, which are as follows:

1. Base case: The DS is unaltered, and no DGs are mounted.

2. Reconfiguration: The DS is only reconfigured with the help of the sectionalizers
and tie-switches (Base case + Reconfiguration).
3. DG Placement: The DGs are mounted at the suitable buses of the DS. (Base
case + DG Installation).
4. Reconfiguration and DG installation: The reconfiguration of the DS is done, fol-
lowed by installing the DGs at suitable buses. (Case 2 + Case 3) Simultaneous
reconfiguration and integration of the DG: The process of reconfiguration and
DG installation is simultaneously done.

The various parameters of IWD such as initial velocity (init_vel = 200), maxi-
mum iteration (max_iter = 1000), velocity update parameters (av = 1,bv = 0.001 and
cv = 1), and soil update parameters (as = 1,bs = 0.001, and cs = 1) are standard for
both the IEEE DS.

5.1 Test system I: IEEE 33 bus

The first test DS considered is IEEE 33-bus DS that comprises 37 branches in which
there are 32 sectionalizing switches, and the rest are tie switches. All the necessary
data have been derived from [37]. The net active and reactive power loads connected
to the DS are 3.72 MW and 2.3 MVAr, respectively. The voltage value at the sub-
station grid level is considered as 1 p.u. The DS has a base capacity of 100MVA
and a base voltage of 12.66 kV. However, the voltage has been converted into per
unit (p.u) to ease the plotting of the voltage profile. The voltage profile of the IEEE
DS under all the five entities is shown in Fig. 3. The algorithm’s performance has
been mentioned in Table 1, which includes the comparative analysis with other algo-
rithms such as FWA, HAS, SFS and SSA. The acquired results mentioned in the
table also include the new configuration of tie-switches, minimum voltage and cor-
responding bus, APL, decrement in loss, and the location and corresponding size of
the DGs.
Case 1: The base case is simulated by applying the load flow algorithm, i.e.,
the BIBC-BCBV method. The total APL is 202.6 kW, and the minimum voltage

13
 N. Bhatt, A. K. Chandel

Fig. 3  Illustration of variation of voltage for five cases of IEEE 33 bus DS

is 0.9131 p.u., which occurs at bus 18. Since the tie-switches are not operated, they
remain open, and the only sectionalizers are connected in the DS.
Case 2: The second case corresponds to DSR, where an initial configuration of
sectionalizers and tie-switches is assumed the same as the base case. The IWDA is
applied to obtain the new structure of switches and sectionalizers. It is visible from
the voltage profile (shown in Fig. 3), the overall voltage value across each bus of the
DS is higher than the case 1. Additionally, the minimum voltage value substantially
improves from 0.9131 p.u. (case 1) to 0.9465 p.u. at bus 32, and the net real power
loss is 139.48 kW. Further, Table 1 indicates that the proposed method gives a new
configuration of switches, with 7–9–14–32–37 as open switches and tenders power
loss similar to other schemes.
Case 3: The third case deals with the mounting of appropriately powered DGs in
the DS. To derive the optimal location of the DGs, APLS is employed. The APLS is
computed for all the buses of the DS and sorted in descending order of their sensi-
tivities. The bus with the highest value of APLS is considered for mounting the DG.
For the present scenario, the three buses with the highest sensitivity are 8, 29, and
24, as only three DGs are mounted. The size of the DGs computed by the IWDA
is 0.1059 MW, 0.5526 MW, and 1.0458 MW. With the integration of the DGs, the
minimum voltage now increases from 0.9465 p.u. (reconfiguration) to 0.9619 p.u. at
bus 33. The APL also reduces from 139.48 kW (reconfiguration) to 76.42 kW.
Case 4: In the fourth case, the integration of DGs is done after the reconfigu-
ration of the DS. The IWDA is employed for reconfiguration as well as sizing
of the DGs. Since the DGs are mounted after the reconfiguration, the suitable
location determined by APLS is different from the third case. By referring to
Fig. 3, the voltage profile corresponding to the present case indicates further

13
An intelligent water drop approach for simultaneous…

Table 1  Performance analysis and comparison of the proposed IWDA for IEEE 33 bus DS
Cases Consid- Entity HSA[22] FWA[23] SFS[27] SSA[25] Proposed
ered method

Base Case 37 37 37 37 37
36 36 36 36 36
35 35 35 35 35
34 34 34 34 34
33 33 33 33 33
Position of DG NR
in the DS
DG size
Minimum 0.9131(18) 0.9131(18) 0.9131(18) 0.9131(18) 0.9131(18)
voltage(bus)
APLS (kW) 202.68 202.68 202.68 202.68 202.68
Decrement in NR
loss (%)
Reconfiguration Disconnected 37 7 37 32 37
switches
32 9 32 28 32
14 14 14 14 14
9 28 9 9 9
7 32 7 7 7
Position of DG NR
in the DS
DG size
Minimum 0.9423 0.9413 0.9378 0.9378 (32) 0.9465 (32)
voltage(bus)
APLS (kW) 138.06 139.98 139.55 139.55 139.48
Decrement in 31.88 30.93 31.15 31.94 31.67
loss (%)

13
 N. Bhatt, A. K. Chandel

Table 1  (continued)
Cases Consid- Entity HSA[22] FWA[23] SFS[27] SSA[25] Proposed
ered method

DG Placement Disconnected 37 37 37 37 37
switches
36 36 36 36 36
35 35 35 35 35
34 34 34 34 34
33 33 33 33 33
Position of DG 18 14 14 13 8
in the DS
17 18 24 23 29
33 32 30 29 24
DG size 0.1070 0.5897 0.7540 0.7536 0.1059
0.5724 0.1895 1.0994 0.11004 0.5526
1.0462 1.0146 1.0714 0.10706 1.0458
Minimum 0.9670 0.9680 0.9687 0.9686 (32) 0.9619(33)
voltage(bus)
APLS (kW) 96.76 88.68 71.47 71.45 70.42
Decrement in 52.26 56.24 64.74 64.74 65.255
loss (%)
DG placement Disconnected 37 37 37 32 37
after Recon- switches
figuration 32 32 32 28 32
14 14 14 14 14
9 9 9 9 9
7 7 7 7 7
Position of DG 32 32 24 8 30
in the DS
31 33 30 24 15
30 18 8 30 12
DG size 0.2686 0.5996 1.0682 0.932 0.2578
0.1611 0.3141 0.9503 0.1068 0.1623
0.6612 0.1591 0.9317 0.950 0.7421
Minimum 0.9479 0.9612 0.9741 0.9741(18) 0.97568(32)
voltage(bus)
APLS (kW) 97.13 83.91 58.88 58.87 58.472
Decrement in 52.07 58.59 70.95 70.95 71.150
loss (%)

13
An intelligent water drop approach for simultaneous…

Table 1  (continued)
Cases Consid- Entity HSA[22] FWA[23] SFS[27] SSA[25] Proposed
ered method

Simultaneous Disconnected 32 32 34 28 32
reconfigura- switches
tion and DG
placement 28 28 32 17 28
14 14 28 14 14
10 10 11 11 11
7 7 6 8 7
Position of DG 32 32 8 8 29
in the DS
31 29 25 24 15
33 18 13 31 21
DG size 0.5258 0.5367 0.6951 0.1027 1.5831
0.5586 0.6158 1.5718 0.1180 0.6723
0.5840 0.5315 0.6318 0.8370 0.5017
Minimum 0.9700 0.9713 0.9724 0.9762(18) 0.97642(32)
voltage(bus)
APLS (kW) 73.05 67.11 55.28 56.42 55.242
Decrement in 63.95 66.89 72.73 72.16 72.74
loss (%)

NR not relevant

improvement compared to the first three entities. The minimum voltage rises
from 0.9619 p.u. (DG integration) to 0.9756 p.u. at bus 32 APL of 65.472 kW.
Table 1 shows that the configuration of open switches is the same as the recon-
figuration case, but the DGs have an optimal size of 0.2578 MW, 0.1623 MW,
and 0.7421 MW at buses 32,31, and 30.
Case 5: The last case incorporates the simultaneous reconfiguration of the DS
along with the placement of the DG. IWDA gives out an optimal configuration
with 7–11–14–28–32 as opened switches. Subsequently, the DGs are integrated
at buses 29, 15, and 21 with an optimal size of 1.5831 MW, 0.6723 MW, and
0.5017 MW. The minimum voltage and the APL are 0.97641(at 33 bus) p.u and
55.242 kW, respectively. The results signify a subsequent improvement in mini-
mum voltage and APL reduction from case 2 to the present case. Likewise, the
best values of the decline in the APL and the minimum voltage belong to the
present case for all the methods cited in Table 1.

13
 N. Bhatt, A. K. Chandel

5.2 Test system I: IEEE 69 bus

The IEEE 69 bus DS is relatively a larger DS with 68 sectionalizers and 5 tie

switches. The sectionalizers are indexed from 1 to 68, and 69 to 73 represent the
tie-switches. The line and load data have been derived from [38]. The net load
connected is 3.802 MW and 2.694 MVAr. Like the IEEE 33 bus test system, this
system is also subjected to five scenarios. The initial configuration of the opened
switches is 69–70–71–72. Figure 4 shows the voltage profile under five cases, and
Table 2 comprises the results obtained by IWDA. The base case gives the total loss
of 225 kW with the minimum voltage of 0.9092 at bus 65. This APL reduces to
98.58 kW, 76.42 kW, 37.57 kW, and 35.74 kW from case 2 to case 5 by IWDA. Fur-
ther, it is perceived from Table 2 that the minimum voltage rises from 0.9092 for the
base case to 0.9801 for simultaneous reconfiguration and DG placement. The results
obtained from IWDA are also compared with HSA, FWA, SFS, and SSA and are
mentioned in Table 2. In the second case, i.e., the DSR, the new configuration has
13, 18, 56, 61, and 69 as open tie switches. The overall voltage profile of the DS for
this case has a better response than the.
base case, and the minimum voltage is 0.9489 p.u. The APL is 98.58 kW which is
better than all the other algorithms. In the third case, the DGs are mounted at buses
65, 64, and 61 derived by APLS. Since only DGs are placed in the present case,
the DS configuration is the same as the base case. The optimal sizes of the DGs
obtained by IWDA are 0.1059 MW, 0.5526 MW, and 1.0458 MW. The minimum
voltage value is 0.9619 p.u. and the voltage profile is significantly flatter than both
the previous cases. The placement of DG after reconfiguration is mentioned in case

Fig. 4  Illustration of variation of voltage for five cases of IEEE 69 bus DS

13
An intelligent water drop approach for simultaneous…

Table 2  Performance analysis and comparison of the proposed IWDA for IEEE 69 bus DS
Cases Consid- Entities HSA[22] FWA[23] SFS[27] SSA[25] Proposed
ered method

Base Case Disconnected 73 73 73 73 73

switches
72 72 72 72 72
71 71 71 71 71
70 70 70 70 70
69 69 69 69 69
Position of DG NR
in the DS
DG size
Minimum 0.9092 (65) 0.9092(65) 0.9092 (65) 0.9092 (65) 0.9092(65)
voltage(bus)
APLS (kW) 225 225 225.03 224.96 225
Decrement in NR
loss (%)
Reconfiguration Disconnected 69 70 70 71 69
switches
61 69 69 69 61
56 61 61 61 56
18 56 55 58 18
13 14 14 14 13
Position of DG NR
in the DS
DG size
Minimum 0.9428 0.9495 0.9495 0.9492(61) 0.9489 (61)
voltage(bus)
APLS (kW) 99.35 98.59 98.62 98.63 98.58
Decrement in 55.85 56.17 56.17 56.15 56.18
loss (%)

13
 N. Bhatt, A. K. Chandel

Table 2  (continued)
Cases Consid- Entities HSA[22] FWA[23] SFS[27] SSA[25] Proposed
ered method

DG Placement Disconnected 73 73 73 73 73
switches
72 72 72 72 72
71 71 71 71 71
70 70 70 70 70
69 69 69 69 69
Position of DG 65 65 11 10 65
in the DS
64 62 18 17 64
63 27 61 60 17
DG size 0.1018 0.4085 0.5268 0.527 0.1059
0.3690 1.1986 0.3804 0.380 0.5526
1.3024 0.2258 1.7190 1.718 1.0458
Minimum 0.9677 0.9740 0.9790 0.9789(65) 0.9619(61)
voltage(bus)
APLS (kW) 86.77 77.85 69.44 69.41 76.42
Decrement in 61.43 65.39 69.14 69.14 66.03
loss (%)
DG placement Disconnected 69 70 70 71 69
after Recon- switches
figuration 61 69 69 69 61
56 61 61 61 56
18 56 55 58 18
13 14 14 14 13
Position of DG 61 61 11 11 61
in the DS
60 62 61 27 11
58 64 64 61 64
DG size 1.0666 1.0014 0.5376 0.538 1.067
0.3525 0.2145 1.4340 0.566 0.2246
0.4257 0.1425 0.4903 1.434 0.1384
Minimum 0.9619 0.9720 0.9813 0.9812(61) 0.97568(32)
voltage(bus)
APLS (kW) 51.30 43.88 35.17 35.54 35.57
Decrement in 77.20 80.49 84.37 84.20 84.19
loss (%)

13
An intelligent water drop approach for simultaneous…

Table 2  (continued)
Cases Consid- Entities HSA[22] FWA[23] SFS[27] SSA[25] Proposed
ered method

Simultaneous Disconnected 69 70 70 70 69
reconfigura- switches
tion and DG
placement 61 69 69 69 61
58 63 55 63 58
17 55 26 58 17
13 13 12 14 13
Position of DG 61 61 59 11 61
in the DS
60 62 61 27 17
62 65 60 61 11
DG size 1.0666 1.1272 0.3802 0.650 0.5263
0.3525 0.2750 1.1811 0.490 0.6921
0.4257 0.4159 0.3802 1.467.5 1.5124
Minimum 0.9736 0.9796 0.9810 0.9808(61) 0.9801(62)
voltage(bus)
APLS (kW) 40.30 39.25 42.84 35.81 35.74
Decrement in 82.08 82.55 80.94 84.08 84.11
loss (%)

4. The DGs are installed after performing the reconfiguration by IWDA. The APLS
suggests buses 62, 61, and 64 mount the DGs. The size of the DGs for these respec-
tive buses as per the IWDA is 1.067 MW, 0.2246 MW, and 0.1384 MW. The APL
after mounting the DGs is 35.57 kW which is lesser than HSA and FWA. For the
last case, the network configuration has 13–17–58–61–69 as opened switches deter-
mined by IWDA. The DGs are placed at bus 61 (0.5263 MW), 17 (0.6921 MW),
and 11 (1.5124 MW). Figure 4 shows that the voltage profile for the present case has
the flattest response compared to the other 4 cases. Additionally, from Table 2, it is
observed that the power loss is minimum among all the cases.

5.3 Discussion

The scrutiny of the obtained results demonstrates that the case of DSR along DG
placement is more potent for reducing APL and improving the voltage profile than
the simple DSR. For the case of DSR amalgamated with sizing and placement of
DG, the IWD yields an optimal solution that improves the minimum voltage to
0.9764 p.u. and reduces the APL by 72.74% for IEEE 33 bus and 0.9801 p.u. and
84.11% for IEEE 69 bus systems. However, considering the case of DSR only, the
minimum voltage and APL for IEEE 33 bus are 0.9465 p.u. and 31.67% and 0.9489

13
 N. Bhatt, A. K. Chandel

p.u. and 56.18% for 69 bus DS. Additionally, it is also imperative that the IWDA is
a profound and credible method to resolve the simultaneous DG placement and DSR
in all the accounted DS. Besides, the IWDA determines the new configuration of tie
switches and DGs’ sizes, resulting in improved voltage values and reduced APL to
other methods.

6 Conclusion

In the present paper, IWDA has been implemented to address the complex problem
of simultaneous DSR and DG placement. The algorithm portrays the flow of water
droplets in the rivers to achieve an optimal solution. Additionally, four other opera-
tional situations (reconfiguration, integration of the DGs, and installation of DGs
after reconfiguration) are also simulated to demonstrate the preeminence of the pro-
jected method. The results are obtained after implementing the algorithm on IEEE
33 bus and IEEE 69 bus DS. The highlights of the present work are summarized as:

(a) The algorithm solves the primary objective of minimizing APL and the enhance-
ment of the voltage profile of the DS, complying with all the constraints.
(b) BIBC-BCBV method is employed to perform the load flow analysis.
(c) The sensitivity analysis has been utilized to locate the feasible buses to mount
the DGs.
(d) The algorithm shows the best figures of voltage and APL for simultaneous DSR
and DG placement.

Additionally, the voltage profile becomes relatively flat for the concluding case,
which is evident in the results given in Tables 1 and 2. In compliance with the
improved voltage profile, the minimum voltage value also becomes 1.069 times (for
IEEE 33 bus DS) and 1.077 times (IEEE 69 bus DS) its nominal value (base case).
Further, the comparative analysis manifests that the IWDA surpasses the other
methods such as HSA, FWA, SFA in solving the complex problem of simultaneous
DSR combined with DG allocation.

References
1. Muhammad, M.A., Mokhlis, H., Naidu, K., Amin, A., Franco, J.F.: Distribution network planning
via network reconfiguration and DG integration using dataset approach and water cycle algorithm. J.
Mod. Power Syst. Clean Energy 8(1), 86–93 (2020). https://​doi.​org/​10.​35833/​MPCE.​2018.​000503
2. Karimi, M., Atashbar, M., Ravadanegh, S.N., Karimi, M., Atashbar, M., Ravadanegh, S.N.: Risk-
based modeling of simultaneous reconfiguration of power distribution networks and allocation of
distributed generation. Int. J. Ambient Energy 2018, 1–14 (2018). https://​doi.​org/​10.​1080/​01430​
750.​2018.​14513​72
3. Lotfi, H., Ghazi, R., Bagher-Naghibi-Sistani, M.: Multi-objective dynamic distribution feeder recon-
figuration along with capacitor allocation using a new hybrid evolutionary algorithm. Energy Syst.
11(3), 779–809 (2020). https://​doi.​org/​10.​1007/​s12667-​019-​00333-3

13
An intelligent water drop approach for simultaneous…

4. Ameli, A., Ahmadifar, A., Shariatkhah, M., Vakilian, M.: A dynamic method for feeder reconfig-
uration and capacitor switching in smart distribution systems. Int. J. Electr. Power Energy Syst.
85, 200–211 (2017). https://​doi.​org/​10.​1016/j.​ijepes.​2016.​09.​008
5. Merlin A.B.H.: Search for a minimal loss operating spanning tree configuration in an urban
power distribution system. In: Proceeding 5th Power Syst. Comput. Conf (PSCC), Cambridge,
UK; vol. 1, pp. 1–18 (1975)
6. Civanlar, L.S., Grainger, J.J., Yin, H.: Distribution feeder reconfiguration for loss reduction.
IEEE Trans. Power Deliv. 148(3), 148–162 (1988). https://​doi.​org/​10.​1109/​61.​193906
7. Shirmohammadi, D., Hong, H.W.: Reconfiguration of electric distribution networks for resistive
line losses reduction. IEEE Trans. Power Deliv. 4(2), 148–162 (1989). https://​doi.​org/​10.​1109/​
61.​25637
8. Ferdavani, A.K., Asuhaimi, A., Zin, M., Khairuddin, A.: Reconfiguration of distribution system
through two minimum-current neighbour-chain updating methods. IET Gener. Trans. Distrib.
7(12), 1492–1497 (2013). https://​doi.​org/​10.​1049/​iet-​gtd.​2012.​0737
9. Khodr, H.M., Matos, M.A., Pereira, J.: Distribution systems reconfiguration based on OPF using
benders decomposition. IEEE Trans. Power Deliv. 24(4), 2166–2176 (2009). https://​doi.​org/​10.​
1109/​TPWRD.​2009.​20275​10
10. Jazebi, S., Vahidi, B.: Reconfiguration of distribution networks to mitigate utilities power quality
disturbances. Electr. Power Syst. Res. 91, 9–17 (2012). https://​doi.​org/​10.​1016/j.​epsr.​2012.​04.​
008
11. Zhu, J.Z.: Optimal reconfiguration of electrical distribution network using the refined genetic
algorithm. Electr. Power Syst. Res. 62(1), 37–42 (2002). https://​doi.​org/​10.​1016/​S0378-​7796(02)​
00041-X
12. Ganesh-Vulasala, R.T., Sivanagaraju, S.: Feeder reconfiguration for loss reduction in unbalanced
distribution system using genetic algorithm. Int. J. Electr. Electron. Eng. 3(12), 754–762 (2009).
https://​doi.​org/​10.​5281/​zenodo.​10846​78
13. Amanulla, B., Chakrabarti, S., Singh, S.N.: Reconfiguration of power distribution systems consider-
ing reliability and power loss. IEEE Trans. Power Deliv. 27(2), 918–926 (2012). https://​doi.​org/​10.​
1109/​TPWRD.​2011.​21799​50
14. Othman, A.M., El-Fergany, A.A., Abdelaziz, A.Y.: Optimal reconfiguration comprising voltage sta-
bility aspect using enhanced binary Particle Swarm optimization algorithm. Electr. Power Compo-
nents Syst. 43(14), 1656–1666 (2015). https://​doi.​org/​10.​1080/​15325​008.​2015.​10416​23
15. Rao, R.S., et al.: Optimal network reconfiguration of large-scale distribution system using har-
mony search algorithm. IEEE Trans. Power Syst. 26(3), 1080–1088 (2011). https://​doi.​org/​10.​1109/​
TPWRS.​2010.​20768​39
16. Nguyen, T.T., Truong, A.V.: Distribution network reconfiguration for power loss minimization and
voltage profile improvement using Cuckoo Search algorithm. Int. J. Electr. Power Energy Syst. 68,
233–242 (2015). https://​doi.​org/​10.​1016/j.​ijepes.​2014.​12.​075
17. Nguyen, T.T., Nguyen, T.T.: An Improved Cuckoo search algorithm for the problem of electric
distribution network reconfiguration. Appl. Soft Comput. J. 84, 105720 (2019). https://​doi.​org/​10.​
1016/j.​asoc.​2019.​105720
18. Mohamed-Imran, A., Kowsalya, M.: A new power system reconfiguration scheme for power loss
minimization and voltage profile enhancement using fireworks algorithm. Int. J. Electr. Power
Energy Syst. 62, 312–322 (2014). https://​doi.​org/​10.​1016/j.​ijepes.​2014.​04.​034
19. Nguyen, T.T., Nguyen, T.T., Truong, A.V., Nguyen, Q.T., Phung, T.A.: Multi-objective electric dis-
tribution network reconfiguration solution using Runner-root algorithm. Appl. Soft Comput. J. 52,
93–108 (2016). https://​doi.​org/​10.​1016/j.​asoc.​2016.​12.​018
20. Vita, V., Alimardan, T.: The impact of distributed generation in the distribution networks’ volt-
age profile and energy losses. In: IEEE European Modelling Symposium (EMS); 2015. pp. 260–5.
https://​doi.​org/​10.​1109/​EMS.​2015.​46
21. Lalitha, M.P., Reddy, V.V., Usha, V.: Optimal DG placement for minimum real power loss in radial
distribution systems using PSO. J. Theor. Appl. Inf. Technol. 13, 107–116 (2010)
22. Moravej, Z., Akhlaghi, A.: A novel approach based on cuckoo search for DG allocation in distribu-
tion network. Int. J. Electr. Power Energy Syst. 44(1), 672–679 (2013). https://​doi.​org/​10.​1016/j.​
ijepes.​2012.​08.​009
23. Khatod, J.S.D.K., Pant, V.: Evolutionary programming based optimal placement of renewable
distributed generators. IEEE Trans. Power Syst. 28(2), 683–695 (2013). https://​doi.​org/​10.​1109/​
TPWRS.​2012.​22110​44

13
 N. Bhatt, A. K. Chandel

24. Candelo, J., Hernández, H.: Location and size of distributed generation to reduce power losses
using a bat-inspired algorithm. In: Proc. VII Simp. Int. sobre Calid. la Energía Eléctrica, Medellin,
Colomb. 2013. https://​doi.​org/​10.​13140/2.​1.​1594.​9603
25. Mohamed-Imran, A., Kowsalya, M., Kothari, D.P.: A novel integration technique for optimal net-
work reconfiguration and distributed generation placement in power distribution networks. Int. J.
Electr. Power Energy Syst. 63, 461–472 (2014). https://​doi.​org/​10.​1016/j.​ijepes.​2014.​06.​011
26. Mehta, P., Bhatt, P., Pandya, V.: Optimal selection of distributed generating units and its placement
for voltage stability enhancement and energy loss minimization. Ain Shams Eng. J. 9(2), 187–201
(2018). https://​doi.​org/​10.​1016/j.​asej.​2015.​10.​009
27. Poornazaryan, B., Karimyan, P., Gharehpetian, G.B., Abedi, M.: Optimal allocation and sizing of
DG units considering voltage stability, losses and load variations. Int. J. Electr. Power Energy Syst.
79, 42–52 (2016). https://​doi.​org/​10.​1016/j.​ijepes.​2015.​12.​034
28. Esmaili, M.: Placement of minimum distributed generation units observing power losses and volt-
age stability with network constraints. IET Gener. Transm. Distrib. 7(8), 813–821 (2013). https://​
doi.​org/​10.​1049/​iet-​gtd.​2013.​0140
29. Sambaiah, K.S., Jayabarathi, T.: Optimal reconfiguration and renewable distributed generation allo-
cation in electric distribution systems. Int. J. Ambient Energy 42(9), 1018–1031 (2021). https://​doi.​
org/​10.​1080/​01430​750.​2019.​15836​04
30. Rao, R.S., Ravindra, K., Satish, K., Narasimham, S.V.L.: Power loss minimization in distribution
system using network reconfiguration in the presence of distributed generation. IEEE Trans. Power
Syst. 28(1), 317–325 (2013). https://​doi.​org/​10.​1109/​TPWRS.​2012.​21972​27
31. Esmaeili, M., Sedighizadeh, M., Esmaili, M.: Multi-objective optimal reconfiguration and DG (Dis-
tributed Generation) power allocation in distribution networks using Big Bang-Big Crunch Algo-
rithm considering load uncertainty. Energy 103, 86–99 (2016). https://​doi.​org/​10.​1016/j.​energy.​
2016.​02.​152
32. Abd-El-salam, M.F., Beshr, E., Eteiba, M.B.: A new hybrid technique for minimizing power losses
in a distribution system by optimal sizing and siting of distributed generators with network recon-
figuration. Energies 11, 12 (2018). https://​doi.​org/​10.​3390/​en231​23351
33. The, T., Hoang, K., Ngoc, D.: Stochastic Fractal search algorithm for reconfiguration of distribution
networks with distributed generations. Ain Shams Eng. J. 11, 2 (2019). https://​doi.​org/​10.​1016/j.​
asej.​2019.​08.​015
34. Abualigah, L., Shehab, M., Alshinwan, M., Alabool, H.: Salp Swarm algorithm: a compre-
hensive survey. Neural Comput. Appl. 32(15), 11195–11215 (2020). https://​doi.​org/​10.​1007/​
s00521-​019-​04629-4
35. Wang, J.S., Li, S.X.: An improved Grey Wolf optimizer based on differential evolution and elimina-
tion mechanism. Sci. Rep. 9(1), 1–21 (2019). https://​doi.​org/​10.​1038/​s41598-​019-​43546-3
36. Mistry, K.D., Roy, R.: Enhancement of loading capacity of distribution system through distributed
generator placement considering techno-economic benefits with load growth. Int. J. Electr. Power
Energy Syst. 54, 505–515 (2014). https://​doi.​org/​10.​1016/j.​ijepes.​2013.​07.​032
37. Baran, W.F.: Network reconfiguration in distribution systems for loss reduction and load balancing.
IEEE Trans. Power Deliv. 4(2), 1401–1407 (1989). https://​doi.​org/​10.​1109/​61.​25627
38. Jean-Jumeau, R., Chiang, H.-D.: Optimal network reconfigurations in distribution systems: part
2: solution algorithms and numerical results. IEEE Trans. Power Deliv. 5(3), 1568–1574 (1990).
https://​doi.​org/​10.​1109/​61.​58002

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations.

13