International Journal of Power Electronics and Drive Systems (IJPEDS)

Vol. 15, No. 2, June 2024, pp. 1276~1289

ISSN: 2088-8694, DOI: 10.11591/ijpeds.v15.i2.pp1276-1289  1276

Optimal coordination of directional over current relays for

distribution systems using hybrid GWO-CSA

Arathi Pothakanahalli Bheemasenarao, Shankaralingappa Channappa Byalihal

Department of Electrical and Electronics Engineering, Dr. Ambedkar Institute of Technology
(Affiliated to Visvesvaraya Technological University), Karnataka, India

Article Info ABSTRACT

Article history: Coordination of protective relays is a critical aspect of electrical distribution
systems, ensuring effective and reliable protection against faults. In modern
Received Aug 5, 2023 power systems, the integration of distributed generation (DG) sources adds
Revised Nov 21, 2023 complexity to the coordination task. The dynamic nature of DG systems
Accepted Dec 7, 2023 requires adaptive relay settings that can swiftly detect and isolate faults while
minimizing potential damage and downtime. The purpose of this research is
to improve the coordination of directional over current relays in electrical
Keywords: distribution systems, particularly in DG systems. An optimization technique
combining the grey wolf optimization (GWO) and cuckoo search algorithm
Coordination (CSA) is developed to identify the best relay settings that reduce overall
Cuckoo search algorithm operation time while ensuring excellent fault identification and isolation. To
Distributed generation systems address relay faults caused by DG integration, a suitable primary and backup
Grey wolf optimization relay design is chosen, and the influence of time multiplier settings (TMS) on
Optimal relay setting system performance and reliability is investigated. The proposed GWO-CSA
Over current relays technique is evaluated and implemented on IEEE 3, 8 and 15-bus systems
using MATLAB. Simulation results show that the GWO-CSA strategy
outperforms well compared to previous algorithms, enabling optimal
coordination and increased protection in DG systems while drastically
lowering relay operating time.
This is an open access article under the CC BY-SA license.

Corresponding Author:
Arathi Pothakanahalli Bheemasenarao
Department of Electrical and Electronics Engineering, Dr. Ambedkar Institute of Technology
(Affiliated to Visvesvaraya Technological University)
Belagavi, Karnataka, India
Email: [email protected]

1. INTRODUCTION
The current paradigm of electrical power distribution network (EPDN) planning, management, and
control is based on the premise of unidirectional power flows transported from higher voltage levels to
consumption levels. EPDNs are typically designed in a radial configuration, facilitating the implementation of
straightforward and cost-effective protection methods that ensure selective operation of the protection system
[1]. Over-current relays (OCRs) play a crucial role in EPDNs, serving as both primary and secondary protection
mechanisms. If the primary relay fails to operate within the specified time period, the backup protective relay is
activated. When a fault occurs within the backup relay's protection zone, it must conform to the coordination
time interval (CTI) [2]. It is essential for OCRs to operate swiftly and be capable of coordination with other
relays to enhance system dependability, thereby improving the speed and reliability of the overall protection
strategy. A variety of methods for OCR coordination have been suggested. Three categories can be made out of
these techniques: optimization, topological analysis, and trial-and-error [3]. However, in the first two classes'

Journal homepage: http://ijpeds.iaescore.com

Int J Pow Elec & Dri Syst ISSN: 2088-8694  1277

solutions, the plug multiplier setting (PMS) or pickup setting (PS) and time multiplier setting (TMS) were
manually determined using expert knowledge or analytical computation to ensure the optimal relay coordination.
Today, the EPDN is so complicated that using the traditional relay coordination mechanism is not
advised. As a result, effective relay coordination is crucial for adaptation [4]. The practical advantage of Meta
heuristic methods resides in their efficacy and broad applicability. Meta heuristic approaches are designed to
address difficult optimization problems by providing near-optimal solutions in reasonable computation times.
The seeker algorithm, differential evolution, genetic algorithm, and particle swarm optimization have all been
used to solve the DOCR coordination problem in more recent work. Analytical optimization approaches such
as linear (LP) and non-linear programming (NLP), in contrast to meta heuristics, ensure convergence to a local
or global optimum depending on the problem model and the supplied initial point. In the linear programming
approach, only TMS is employed as a continuous variable, whilst PS is kept constant based on prior knowledge
[5]. Using the linear programming methodology, the Big-M [6], [7], dual simplex [8], [9], modified jaya
algorithm [10], and other optimization techniques have been used to discover the optimal TMS values. In the
non-linear programming approach, TMS and PS both serve as decision variables.
Researchers used the following techniques to find the best relay settings, which include both TMS and
PS, by taking into account the non-linear programming methodology. In this field, genetic approach (GA)
quickly rose to prominence in the early 1990s. The continuous genetic approach (CGA), an improvement to this
approach, has been developed in [9]. As the chromosome doesn't have to be decoded, CGA produces findings
faster than binary GA. Alkaran et al. [10] developed the fuzzy-based Genetic Algorithm technique as a remedy
to the miscoordination problem, which caused the weighting parameters to shift throughout simulation. Moravej
et al. [11] introduces particle swarm optimization (PSO), the following phase in the bio-nature inspired
technique. A modified differential evolution algorithm with an information exchange strategy was proposed.
Instead of using a single population, several sub populations which enhance the explorative power of the
algorithm were used [12]. The invasive weed optimization (IWO) algorithm has been adapted for the high
dimension coordination problem [13]. It has been demonstrated that the PSO produces better results than both
traditional and contemporary GA. Differential evolution (DE), modified differential evolution (MDE), and
invasive weed optimization (IWO) continue the algorithm's revolution. Hybrid methods have been developed in
[14], [15] to improve MDE performance. Ahmarinejad et al. [16], the cuckoo search algorithm (CSA) is devised.
Improved GSO and the electromagnetic field optimization (EFO) approach have also been introduced in recent
years [17], [18]. All of these algorithms were developed in an effort to identify the best over current relay setting.
The grey wolf optimization (GWO) technique, a new reliable and trustworthy algorithm, was recently
introduced. This GWO approach has been applied to issues in optimal reactive power dispatch [19], integrated
economic emission dispatch [20], and biomedical engineering [21]. Mirjalili et al. [22] presented GWO, which
was motivated by a pack of wolves' hunting techniques. To increase the rate at which the searching agents
explore, some modifications to the traditional GWO have been made. Conventional GWO has been found to
have slow convergence and frequently become stuck in local optimal. Instead of acting as the top three agents'
followers, the suggested enhancement expanded the number of searching agents.
This study seeks to find the best TMS and PS value in order to reduce the objective function. This
study investigates the system's performance and dependability in connection to the time multiplier settings
(TMS) of directional over current relays. To overcome relay difficulties caused by DG penetration, an
appropriate primary along with backup relay configuration is chosen. The proposed GWO-CSA technique is
tested on IEEE 3,8 and 15 bus systems using MATLAB. The simulation results indicate that using the GWO-
CSA hybrid technique significantly reduces relay operation time and performs using the GWO and CSA
methods individually. This result demonstrates how well the suggested technique works to achieve the best
coordination and enhanced security in DG systems.

2. PROBLEM FORMULATION
The primary goal of DOCR issue formulation is to determine the best peak-up current setting and
time-dial setting for primary relays, that are often positioned closest to the fault's site. If the primary relay
coordination fails, one or more of the other relays can serve as a backup, keeping the fault current from
spreading to the rest of the network. Primary relay close-in and far bus fault coordination is frequently sub
classified. This is widely used in literature to describe an acceptable relay coordination objective function, as
will be demonstrated later in this work. This imposes a limitation on relay behaviour, ensuring that a primary
relay always take precedence over its corresponding back-up equivalent. Because the DOCR problem is non-
convex, a hybrid GWO-CSA approach is created in MATLAB and tested on IEEE 3, 8, and 15-bus systems.

2.1. Objective function

Overall, the coordination of directional over-current relay in power systems has been stated as an
optimization problem, with the goal of minimizing the sum of the working times of the system's relays during
Optimal coordination of directional over current relays for … (Arathi Pothakanahalli Bheemasenarao)
1278  ISSN: 2088-8694

a malfunction, denoted by the symbol 'tz'. Near end faults are typically taken into account. It can be expressed
mathematically as given in (1).

Minimize 𝑡𝑧 = ∑𝑛𝑚=1 𝑊𝑚 𝑡𝑚,𝑙 (1)

In this context, n represents the total number of relays in the distribution system. The goal is to
minimize the target function tz. For this purpose, the operating time of the primary relay for faults in zone l and
near end faults is denoted as tm,l, where m corresponds to a specific relay and l represents a particular zone.
Each relay Rm is assigned a weight Wm to quantify its importance in the overall scheme. Considering that
distribution system lines are short and have nearly equal lengths, the probability of a fault occurring on any
line is assumed to be the same. Based on this assumption, equal weights are assigned to all relays, denoted by
W = 1 for each relay [23], [24]. This approach is a typical strategy used in such scenarios.

2.2. Constraints
2.2.1. Coordination constraints
The system's relays must be coordinated according to the requirement as given in (2). Where t m,l
represents the operating time of the primary relay at zone l for a near-end fault, while tp,l represents the operating
time of the backup relay in the same zone for the same fault. The selective time interval (STI) is a time range
that lies between 0.1 and 0.5. The value of STI is determined by how quickly the circuit breakers operate in
the system. This interval is crucial to ensure proper selectivity in relay coordination, where the primary relay
should operate faster than the backup relay within the STI to isolate the fault without any interference from the
backup relay.

𝑡𝑝,𝑙 −𝑡𝑚,𝑙 ≥ 𝑆𝑇𝐼 (2)

Relays require a specific amount of time to operate. However, the time it takes for them to operate
should not be excessive. This places a constraint on the operating time as shown in (3).

𝑡𝑚, 𝑛𝑚,𝑙,𝑚𝑎𝑥 𝑚,𝑙,𝑚𝑖𝑛 (3)

Where tm,l,max is the maximum working period of the relay at bus P for a local fault and t m,l,min is the lowest
operating time. This additionally pertains to other relays.

2.2.2. Relay characteristics

The relays are taken to be identical and to possess typical inverse definite minimum time (IDMT)
properties, which are expressed by (4).

0.14∗(𝑇𝑀𝑆)
𝑡𝑜𝑝𝑡 = (4)
𝑃𝑆𝑀(0.02) −1

Where topt is the duration of the relay. Standard IDMT relay characteristics use the constants 0.14 and 0.02.
PSM stands for the plug setting multiplier which is calculated using (5), and TMS is for the time multiplier
setting.
𝐼𝑟𝑒𝑙𝑎𝑦
𝑃𝑆𝑀 = (5)
𝑃𝑆

In the form of the previously indicated (4), the issue is non-linear. The relay coordination problem can
be expressed as a linear programming problem if the plug setting (PS) of the relays is fixed as well as the
working time of the relays is regarded to be a linear function of the duration multiplier setting [25], Because
PSM is fixed in the linear problem, the operating time is calculated using (6), where α𝑚 is shown in (7).

𝑡𝑜𝑝𝑡 = α𝑚 (𝑇𝑀𝑆) (6)

Where,
0.14
𝛼𝑚 = (7)
𝑃𝑆𝑀 (0.02) −1

As a result, the objective function may be expressed as shown in (8).

Int J Pow Elec & Dri Syst, Vol. 15, No. 2, June 2024: 1276-1289
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1279

𝑡𝑧 = ∑𝑛𝑚=1 𝛼𝑚(𝑡𝑚𝑠)𝑚 (8)

Subject to the limitations specified in (2) and (3).

3. HYBRID GREY WOLF OPTIMIZATION ALGORITHM AND CUCKOO SEARCH

ALGORITHM (GWO-CSA)
It has become more and more common to combine two or more algorithms to find the best answers to
optimization issues. Numerous well-known optimization techniques have been included into hybrid optimized
algorithms, making them more effective in addressing the problems. The GWO algorithm simulates the
leadership structure and hunting behavior of grey wolves. In the wild, grey wolves coexist in packs comprising
four different species. The dominant wolf, known as the alpha (α), holds the highest position in the pack's
hierarchy. The alpha is responsible for critical social decisions, such as food distribution and dealing with
potential threats. Next in the hierarchy is the beta, who serves as the alpha's aide and assists in maintaining
control over the group. The beta commands the other wolves but must always respect and follow the alpha's
leadership. The third level wolf is the delta (δ), who must comply with the requests of both the alpha and the
beta. After their useful lifetimes are over, the alpha and beta are degraded to delta status. At the base of the
pyramid is the omega (ω), and the rest of the group must accept omega's submission.
The GWO algorithm is inspired by this grey wolf pack's hunting behavior and utilizes three types of
wolves for searching the solution space: the alpha wolf, the beta wolf, and the delta wolf. On the other hand,
the cuckoo search or CS algorithm is a different meta heuristic approach. It is inspired by the behavior of
cuckoos, specifically their compulsory brood parasitism behavior. CS employs Lévy flying to generate new
ideas and explore the solution space. Both GWO and CS are popular meta heuristic algorithms with distinct
search strategies, making them effective in different optimization scenarios [26], [27].
Numerous studies have highlighted the strengths of the cuckoo search (CS) algorithm in global
exploration and the grey wolf optimization (GWO) algorithm in exploitation [28]–[31]. In this study, the GWO
approach was employed to optimize the size of system components while minimizing costs and meeting load
demands. However, GWO's tendency to fall into local optima when inspecting individuals with high fitness
values can limit its global search ability. To overcome this limitation and enhance the GWO algorithm, a hybrid
approach called grey wolf optimization with cuckoo search optimization (GWCSO) was proposed. In GWCSO,
the cuckoo search algorithm (CSA) is used to modify the location of nests, allowing for more efficient jumping
between different search areas. By integrating CSA, GWCSO achieves robustness and fast optimization for
sizing units of solar PV, wind turbines, storage batteries, and biomass gasifiers in a grid-connected micro grid
(MG). The collaboration between GWO and CSA is depicted in Figure 1, where the positions, speeds, and
convergence accuracy of grey wolf agents are adjusted using CSA's position update equation. This hybrid
approach ensures an effective and swift solution to optimization problems, making it well-suited for the
application at hand. The GWO algorithm's fitness function determines the "degree of grey" to identify the best
solutions represented by alpha, beta, and delta wolves. These wolves form the key-group, while the remaining
members are represented by the omega wolf. Mathematical models inspired by grey wolf social hierarchy and
hunting behavior are used to design and optimize the GWO algorithm. Encircling behavior during hunting is
expressed using specific in (9) and (10).

𝐸 = |𝐷. 𝑌𝑝(𝑡) − 𝑌(𝑡)| (9)

𝑌(𝑡 + 1) = |𝑌𝑝(𝑡) − 𝐵. 𝐸| (10)

In both of these equations, Y represents the location of a single wolf, and t + 1 represents the future
iteration. The prey's location is represented by Yp, while the coefficient vectors are represented by B and E.
The (11) and (12) model the calculation process.

𝐵 = 2𝑏. 𝑟1 − 𝑏 (11)

𝐷 = 2. 𝑟2 (12)

During the iterative procedure of the GWO algorithm, random integers r1 and r2, ranging between 0 and 1, are
used. If vector b is assigned a value between 2 and 0, the iterative process results in a linear decrease in the
number of iterations.

Optimal coordination of directional over current relays for … (Arathi Pothakanahalli Bheemasenarao)
1280  ISSN: 2088-8694

Figure 1. Flow chart of hybrid grey wolf optimization algorithm and cuckoo search algorithm

When the wolf pack discovers prey, the alpha, beta, and delta wolves take the lead in encircling it.
They possess knowledge of the prey's location. These best three solutions are combined to form a key group,
which is then employed to update the position of each wolf within the group. The following (13)-(16) are used
to update the position of the wolves:

𝑌(𝑡 + 1) = (𝑌1 + 𝑌2 + 𝑌3)/3 (13)

𝑌1 = |𝑌𝑎 − 𝐵1 − 𝐸𝑎| (14)

𝑌2 = |𝑌𝑏 − 𝐵2. 𝐸𝑏| (15)

𝑌3 = |𝑌𝑠 − 𝐵3. 𝐸𝑠| (16)

Where Ya, Yb, and Ys represent the top three results from the iterative process thus far, which comprise the
key group. The mathematical (17)-(19) that follow describe additional parameters.

𝐸𝑎 = |𝐷1. 𝑌𝑎 − 𝑌| (17)

𝐸𝑏 = |𝐷2. 𝑌𝑏 − 𝑌| (18)

𝐸𝑠 = |𝐷3. 𝑌𝑠 − 𝑌| (19)

The movements of grey wolves when they approach their prey with respect to the maximum number of
iterations (Max) can be represented by (20).
𝑡
𝐵 = 2 − 2( ) (20)
𝑀𝑎𝑥

The (13) depicts how the GWO algorithm adjusts the positions of the key group individuals with high
fitness values. However, it may lack effective global search capabilities and might get trapped in local optima,
especially when dealing with large datasets. On the other hand, the CS approach uses random walks and Levy
flights to move nests, with the search path length being arbitrary and having nearly equal probabilities of being
longer or shorter than previous paths. This randomness makes the movement from one location to another more
straightforward. Initially, cuckoo birds select their nests at random and only lay one egg at a time. Second, only
the best nests will endure for upcoming generations. Thirdly, both the quantity of bird nests and the likelihood
that the eggs will be found are fixed. When a foreign egg is discovered, the host bird abandons the nest and

Int J Pow Elec & Dri Syst, Vol. 15, No. 2, June 2024: 1276-1289
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1281

builds another. The following equations are used to update the nests throughout iteration while adhering to
these three requirements:
The CS process evolves a population of N individuals from a starting point (k = 0) to an overall number
of iterations (gen) known as Ek(Y k 1, Yk2,...., Y kN). Each individual Yki (i [1,..., N]) has the dimensions
(Yki,1, Yki,2,...., Yki,n), where (Yki,1, Yki,2,...., Yki,n) is a decision variable in the optimization problem to
be solved. Each individual, Yki (candidate solution), is assessed using an objective function, f(Y ki), the output
of which represents Y ki's fitness value. The use of Levy flights to generate new candidate solutions is one of
the most advanced components of cuckoo search. By shifting the current Yk i, a new candidate solution, Yki
(i[1,...., N]), is generated. A symmetric Levy distribution produces qi, a random step, to produce ci. To produce
qi, Mantegna's algorithm is employed which is by (21).
𝑣
𝑞𝑖 = (21)
|𝑤|1/𝑏

Where w is an n-dimensional vector having 3/2 dimension and v is an n-dimensional vector with v1, v2
dimensions. The normal distributions shown in (22) and (23) are used to determine each element of w and v.

v~N(0,δv2),w~N((0,δw2) (22)
𝑏
г(1+𝑏).sin⁡(𝜋.2)
δv = (1+𝑏) ⁡δw = 1
, (23)
г( ).𝑏.2(𝑏−1)/2
2

Where Γ (.) represents the gamma distribution. After determining si, the necessary adjustment to position Di
is determined as in (24):

Di=0.01si+(Yik-Ybest) (24)

Where Ybest denotes the best solution identified thus far in regard to fitness value, whereas the product is
composed of entry-wise multiplications. Finally, as shown in (25) is used to identify the new candidate solution,
Yk+1i.

Yik+1=Yik+Di (25)

The CS algorithm's step changes, characterized by short-distance detection and infrequent long-
distance walking, allow for efficient exploration of the solution space. As the algorithm progresses, the step
length increases significantly with time, leading to rapid and extensive exploration of potential solutions. This
adaptive behavior enables CS to quickly cover a large search space and find promising regions with potential
optimal solutions. By balancing short-distance local exploration and occasional long-distance global
exploration, the CS algorithm can effectively and rapidly locate solutions to optimization problems. The hybrid
of GWO-CSA algorithm is developed and pseudo code is as given in the following:

Initialize GWO and CSA parameters (population size, and iterations)

Initialize the positions of grey wolves (GWO population)
Initialize the positions of cuckoos (CSA population)
Evaluate fitness of each GWO wolf and CSA cuckoo
While stopping criteria are not met:
Sort GWO and CSA populations by fitness
Perform GWO operations:
For each grey wolf:
Update wolf's position using GWO equations
Boundary checking and fitness evaluation
Perform CSA operations:
For each cuckoo:
Explore a new solution using Levy flight or other CSA methods
Boundary checking and fitness evaluation
Select top solutions from both populations based on fitness
Combine selected solutions to form a merged population
Apply GWO and CSA operations to the merged population:
For each solution:
With a probability P_gwo, perform GWO operation (e.g., search for prey)
With a probability P_csa, perform CSA operation (e.g., lay an egg)
Optimal coordination of directional over current relays for … (Arathi Pothakanahalli Bheemasenarao)
1282  ISSN: 2088-8694

Evaluate fitness of solutions in the merged population

Select the top solutions from the merged population
End while
Return the best solution found

The pseudo code of the GWO-CSA hybrid algorithm is developed for the total operating time minimization
and results are analyzed. The following section details the results and discussion obtained from the optimization
solution.

4. RESULTS AND DISCUSSION

Relay coordination for 3 bus, 8 bus and 15 bus system is developed using the GWO-CS algorithm.
Minimization of total operating time using the TMS as the independent variable for the optimization problem
using GWO-CSA algorithm is developed and the results are tabulated to compare with the previous
implementation. MATLAB based implementation is carried out to obtain the optimized TMS values for the
minimized total operating time objective function given in (8) considering the inequality constraints as given
in (2) and (3). The 3 bus system has 6 relays, 8 bus system has 14 relays and 15 bus system has 42 relays. Three
cases are considered for the implementation of relay coordination of 3 bus system, 3 bus (Case 1: Without DG,
Case 2: With synchronous based distributed generation (SBDG), and Case 3: With inverter based distributed
generation (IBDG); 15 bus (Case 1: With DG); and 8 bus (Case 1: Without DG).
Figure 2 depicts the IEEE 3 bus test feeder with three buses, generators, three lines, and six relays.
Shunt faults occurring on the power supply are immediately detected by both the primary and backup relays.
The primary relay responds faster due to its shorter operating time compared to the backup relay. The system
uses this protection scheme to ensure rapid fault detection and isolation, maintaining the stability of the
distribution system. The faults points for the 3-bus system are denoted as A, B, C and D, while relays are
denoted as R1, R2, R3, R4, R5, and R6 in Figure 2.
Figure 3 [32] illustrates the single line diagram of the 8-bus distribution system, which serves as
second test system. The current transformer ratio of the 8-bus system is as given in the Table 1 [32]. The 8-bus
system with 14 relays denoted as R1, R2 …R14 with two generators are as given in Figure 3. Within this
system, there are two primary generation sources accompanied by two transformers, seven lines, and a total of
14 relays for protection and control. Specifically, the current transformer ratios, representing the minimum and
maximum values of TMS, have been precisely set at 0.1 and 1.2, respectively. The efficiency and stability of
the system depend on proper coordination based on fault current levels. The fault current of both the primary
and the backup relay of the 8-bus system is as given in Table 2. The examination of a much more complex
system to show the flexibility and efficacy of the suggested technique using 15 buses is discussed. This 15-bus
system is protected by a network of 42 relays and connected by 21 distribution lines which is shown graphically
in Figure 4 [33]. This system serves as a model for DG integrated distribution network. The fault currents for
primary relay and secondary relays are shown in Table 3.

Figure 2. Single line diagram of IEEE Figure 3. Relay coordination diagram for
3 bus system 8 bus system [32]

Int J Pow Elec & Dri Syst, Vol. 15, No. 2, June 2024: 1276-1289
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1283

Table 1. Relay and its current transformer ratio (CTR) for 8 bus system [32]
Relay no. CTR Relay no. CTR
1 1200/5 8 1200/5
2 1200/5 9 800/5
3 800/5 10 1200/5
4 1200/5 11 1200/5
5 1200/5 12 1200/5
6 1200/5 13 1200/5
7 800/5 14 800/5

Table 2. Relay and its fault current for 8 bus system

Primary relay Fault current (A) Backup relay Fault current(A)
1 3232 6 3232
2 5924 1 996
2 5924 7 1890
3 3556 2 3556
4 3783 3 2244
5 2401 4 2401
6 6109 5 1197
6 6109 14 1874
7 5223 5 1197
7 5223 13 987
8 6093 7 1890
8 6093 9 1165
9 2484 10 2484
10 3883 11 2344
11 3707 12 3707
12 5899 13 987
12 5899 14 1874
13 2991 8 2991
14 5199 1 996
14 5199 9 1165

Figure 4. Relay coordination diagram for 15 bus system [33]

Optimal coordination of directional over current relays for … (Arathi Pothakanahalli Bheemasenarao)
1284  ISSN: 2088-8694

Table 3. Relay and its fault current for 15 bus system [33]
Primary Fault Backup Fault Primary Fault Backup Fault
relay current (A) relay current (A) relay current (A) relay current (A)
1 3621 6 1233 20 7662 30 681
2 4597 4 1477 21 8384 17 599
2 4597 16 743 21 8384 19 1372
3 3984 1 853 21 8384 30 681
3 3984 16 743 22 1950 23 979
4 4382 7 1111 22 1950 34 970
4 4382 12 1463 23 4910 11 1475
4 4382 20 1808 23 4910 13 1053
5 3319 2 922 24 2296 21 175
6 2647 8 1548 24 2296 34 970
6 2647 10 1100 25 2289 15 969
7 2497 5 1397 25 2289 18 1320
7 2497 10 1100 26 2300 28 1192
8 4695 3 1424 26 2300 36 1109
8 4695 12 1463 27 2011 25 903
8 4695 20 1808 27 2011 36 1109
9 2943 5 1397 28 2525 29 1828
9 2943 8 1548 28 2525 32 697
10 3568 14 1175 29 8346 17 599
11 4342 3 1424 29 8346 19 1372
11 4342 7 1111 29 8346 22 642
11 4342 20 1808 30 1736 27 1039
12 4195 13 1503 30 1736 32 797
12 4195 24 753 31 2867 27 1039
13 3402 9 1009 31 2867 29 1828
14 4606 11 1475 32 2069 33 1162
14 4606 24 753 32 2069 42 907
15 4712 1 853 33 2305 21 1326
15 4712 4 1477 33 2305 23 979
16 2225 18 1320 34 1715 31 809
16 2225 26 905 34 1715 42 907
17 1875 15 969 35 2095 25 903
17 1875 26 905 35 2095 28 1192
18 8426 19 1372 36 3283 38 882
18 8426 22 642 37 3301 35 910
18 8426 30 681 38 1403 40 1403
19 3998 3 1424 39 1434 37 1434
19 3998 7 1111 40 3140 41 745
19 3998 12 1463 41 1971 31 809
20 7662 17 599 41 1971 33 1162
20 7662 22 642 42 3295 39 896

4.1. Result analysis of 3- bus system

The relay coordination problem was solved using GWO-CSA. In each case the number of search
agents were taken as 50 and the number iterations were taken to be 100. From Tables 4 and 5 it can be inferred
that the operating time of each relay is reduced when DG is connected to the system since the fault current
increases. The total operating time with SBDG reduces by 44.36% when DG is not connected and 5.53% with
IBDG.

Table 4. Operating time (OT), time multiplier setting (TMS) of relays for 3 bus system
OT in Sec OT in Sec OT in Sec TMS TMS TMS
Case 1 Case 2 Case 3 Case 1 Case 2 Case 3
1.1457 1.1211 1.9226 0.0502 0.0497 0.0502
0.3100 0.2517 0.2842 0.0512 0.0481 0.0506
4.9533 1.5935 3.8774 0.0502 0.0495 0.0502
1.2220 1.1204 1.2370 0.0501 0.0485 0.0501
0.4426 0.3373 0.3927 0.0501 0.0530 0.0501
0.5795 0.3903 0.4602 0.0503 0.0488 0.0502

Table 5. Total operating time (TOT) for all cases 3 bus system
Case TOT in Seconds
1 8.6531
2 4.8143
3 8.1741

Int J Pow Elec & Dri Syst, Vol. 15, No. 2, June 2024: 1276-1289
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1285

These tables offer valuable insights into how the introduction of a SBDG influences the currents and
associated constants during fault conditions at different locations within the power system. The comparative
data aids in assessing the impact and benefits of SBDG integration on system performance and fault
management. The operating time of all the relays for all the faults are optimized and are compared for all the
three cases like without DG, with SBDG and IBDG. The results shows that operating time is reduced since the
TMS is optimized.
From the convergence graphs shown in Figures 5-7, it is clear that the proposed GWO-CSA technique
Convergence fast for the better TMS with minimized total operating time. The total operating time for all the
above said 3 cases were compared. Results show that the operating time with SBDG is reduced by 3.71 seconds
when compared with the system when DG not connected and with IBDG its reduced by 0.47seconds.

Figure 5. Convergence graph for case 1 Figure 6. Convergence graph for case 2

Figure 7. Convergence graph for case 3

4.2. Result analysis of 15- bus system

Table 6 summarizes the results achieved for the total operating time minimization problem for 15 bus
system. The optimized TMS values are both tabulated. Each relay's operating time during the fault situation is
recorded TMS is the independent variable, and operating time is calculated from TMS values. The convergence
graph of the total operating time minimization is shown in Figure 8. The objective function converges at 160 th
iteration and out performs better by reducing the operating time which is shown in Figure 8.

4.3. Result analysis of 8 bus system

Table 7 gives the optimized operating time and TMS for the 8 bus system for different faults. Figure
9 shows the convergence graph for the 8 bus system. An optimized TMS can be able to obtain using GWO-
CSA when compared with WOA and HWOA techniques, this comparison can be seen Table 8. GWO-CSA
gives better optimized operating time by optimizing TMS when compared to other well-known techniques
which can be observed in the Table 9. A reduction of 45% in the operating time can be observed compared to
WOA and HWOA techniques.

Optimal coordination of directional over current relays for … (Arathi Pothakanahalli Bheemasenarao)
1286  ISSN: 2088-8694

In essence, GWO-CSA employs advanced optimization techniques to fine-tune TMS values for relays,
ultimately leading to minimized operating times and improved system performance. The tables serve as
valuable references for analyzing the impact of these optimizations and comparing results across various
scenarios. These outcomes are presented across distinct scenarios, shedding light on the effectiveness of the
optimization techniques employed in each case.

Table 6. Operating time and TMS

Relay number Operating time in seconds TMS
1 0.3191 0.0508
2 0.4074 0.0548
3 0.3134 0.0599
4 0.1832 0.0783
5 0.3461 0.0642
6 0.3881 0.0758
7 0.8978 0.0785
8 0.4826 0.0810
9 0.2879 0.0729
10 0.5251 0.0565
11 0.1559 0.0666
12 0.2373 0.1001
13 0.4219 0.0720
14 0.3615 0.0546
15 0.4811 0.0666
16 0.5631 0.0572
17 0.7916 0.0885
18 0.2523 0.0609
19 0.2461 0.1172
20 0.1204 0.0565
21 0.3072 0.0502
22 0.4755 0.0723
23 0.8181 0.0679
24 0.7226 0.0502
25 0.5237 0.0577
26 0.4648 0.0512
27 0.386 0.0692
28 0.4053 0.1826
29 0.135 0.0652
30 0.3206 0.1457
31 0.5253 0.0564
32 0.4979 0.0503
33 0.2075 0.0906
34 0.2594 0.1174
35 0.4792 0.0833
36 0.3162 0.0547
37 0.4453 0.0831
38 0.3293 0.0632
39 0.4646 0.0615
40 0.1368 0.06
41 0.2932 0.0567
42 0.1278 0.0571
Total operating time in seconds 16.8

Figure 8. Convergence graph for 15 bus system

Int J Pow Elec & Dri Syst, Vol. 15, No. 2, June 2024: 1276-1289
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1287

Table 7. Comparison of operating time

Relay number Operating time in seconds TMS
1 0.1377 0.0526
2 0.2498 0.0539
3 0.2558 0.0537
4 0.2775 0.0511
5 0.1537 0.0517
6 0.2086 0.0998
7 0.3549 0.0476
8 0.1951 0.0407
9 0.3347 0.0564
10 0.2557 0.0477
11 0.2959 0.0540
12 0.2492 0.0543
13 0.1348 0.0499
14 0.1187 0.0612
Total operating time in seconds 3.22

Figure 9. Convergence graph for 8 bus system

Table 8. Comparison of TMS

Relay number GWO-CSA Whale optimization algorithm (WOA) Hybridized whale optimization algorithm (HWOA)
1 0.0526 0.1 0.1
2 0.0539 0.5929 0.5381
3 0.0537 0.1007 0.1
4 0.0511 0.1 0.2164
5 0.0517 0.3581 0.1
6 0.0998 0.249 0.2689
7 0.0476 0.1018 0.1
8 0.0407 0.343 1.1
9 0.0564 0.1 0.1
10 0.0477 0.1 0.1
11 0.054 0.1004 0.1
12 0.0543 0.1521 0.1
13 0.0499 0.1 0.1
14 0.0612 0.1 0.1

Table 9. Comparison of operating time

Methods TOT in Seconds
GWO-CSA 3.221
WOA 5.95
HWOA 5.85

5. CONCLUSION
Relay coordination using hybrid optimization algorithm GWO-CSA is developed and results are
observed for analysis. The convergence for the better TMS with the minimized total operating time is
developed using MATLAB based implementation. The analysis for relay coordination with the constraint,
without any constraint and with both constraint and DG is observed. The analysis developed on 8, 15, and 3
bus system inferred that the introduction of DG in the bus system improves the total operating time. The results
Optimal coordination of directional over current relays for … (Arathi Pothakanahalli Bheemasenarao)
1288  ISSN: 2088-8694

obtained is satisfactory and is with par with the recent optimization algorithm recently implemented. The hybrid
optimization of GWO-CSA method has shown good improvement than the recent optimization algorithms. To
evaluate its efficacy and robustness the method can be applied further complex systems in the future
implementations.

REFERENCES
[1] D. Alcala-Gonzalez, E. M. G. Del Toro, M. I. Más-López, S. García-Salgado, and S. Pindado, “Linear programming coordination
for overcurrent relay in electrical distribution systems with distributed generation,” Applied Sciences (Switzerland), vol. 12, no. 9,
p. 4279, Apr. 2022, doi: 10.3390/app12094279.
[2] A. Korashy, S. Kamel, L. Nasrat, and F. Jurado, “Developed multi-objective grey wolf optimizer with fuzzy logic decision-making
tool for direction overcurrent relays coordination,” Soft Computing, vol. 24, no. 17, pp. 13305–13317, Sep. 2020, doi:
10.1007/s00500-020-04745-7.
[3] M. Rojnić, R. Prenc, H. Bulat, and D. Franković, “A comprehensive assessment of fundamental overcurrent relay operation
optimization function and its constraints,” Energies, vol. 15, no. 4, p. 1271, Feb. 2022, doi: 10.3390/en15041271.
[4] M. A. Elsadd, T. A. Kawady, A. M. I. Taalab, and N. I. Elkalashy, “Adaptive optimum coordination of overcurrent relays for
deregulated distribution system considering parallel feeders,” Electrical Engineering, vol. 103, no. 3, pp. 1849–1867, 2021, doi:
10.1007/s00202-020-01187-0.
[5] R. Tiwari, R. K. Singh, and N. K. Choudhary, “Performance analysis of optimization technique for protection coordination in single
and multi-loop distribution system,” in Proceedings - 2019 International Conference on Electrical, Electronics and Computer
Engineering, UPCON 2019, Nov. 2019, pp. 1–6, doi: 10.1109/UPCON47278.2019.8980151.
[6] A. Gupta, O. V. G. Swathika, and S. Hemamalini, “Optimum coordination of overcurrent relays in distribution systems using Big-
M and dual simplex methods,” in Proceedings - 2015 International Conference on Computational Intelligence and Communication
Networks, CICN 2015, Dec. 2016, pp. 1540–1543, doi: 10.1109/CICN.2015.341.
[7] S. V. Khond and G. A. Dhomane, “Optimum coordination of directional overcurrent relays for combined overhead/cable distribution
system with linear programming technique,” Protection and Control of Modern Power Systems, vol. 4, no. 1, p. 9, Dec. 2019, doi:
10.1186/s41601-019-0124-6.
[8] P. P. Bedekar and P. N. Korde, “Determining optimum time multiplier setting of overcurrent relays using modified Jaya algorithm,”
in 2017 Innovations in Power and Advanced Computing Technologies (i-PACT), Apr. 2017, vol. 2017-Janua, pp. 1–6, doi:
10.1109/IPACT.2017.8245106.
[9] P. P. Bedekar and S. R. Bhide, “Optimum coordination of overcurrent relay timing using continuous genetic algorithm,” Expert
Systems with Applications, vol. 38, no. 9, pp. 11286–11292, Sep. 2011, doi: 10.1016/j.eswa.2011.02.177.
[10] D. Solati Alkaran, M. R. Vatani, M. J. Sanjari, G. B. Gharehpetian, and M. S. Naderi, “Optimal overcurrent relay coordination in
interconnected networks by using fuzzy-based GA method,” IEEE Transactions on Smart Grid, vol. 9, no. 4, pp. 3091–3101, Jul.
2018, doi: 10.1109/TSG.2016.2626393.
[11] Z. Moravej, M. Jazaeri, and M. Gholamzadeh, “Optimal coordination of distance and over-current relays in series compensated
systems based on MAPSO,” Energy Conversion and Management, vol. 56, pp. 140–151, Apr. 2012, doi:
10.1016/j.enconman.2011.11.024.
[12] M. Singh, B. K. Panigrahi, A. R. Abhyankar, and S. Das, “Optimal coordination of directional over-current relays using informative
differential evolution algorithm,” Journal of Computational Science, vol. 5, no. 2, pp. 269–276, Mar. 2014, doi:
10.1016/j.jocs.2013.05.010.
[13] C. A. Castillo, A. Conde, and M. Y. Shih, “Improvement of non-standardized directional overcurrent relay coordination by invasive
weed optimization,” Electric Power Systems Research, vol. 157, pp. 48–58, Apr. 2018, doi: 10.1016/j.epsr.2017.11.014.
[14] J. A. Sueiro, E. Diaz-Dorado, E. Míguez, and J. Cidrás, “Coordination of directional overcurrent relay using evolutionary algorithm
and linear programming,” International Journal of Electrical Power and Energy Systems, vol. 42, no. 1, pp. 299–305, Nov. 2012,
doi: 10.1016/j.ijepes.2012.03.036.
[15] V. A. Papaspiliotopoulos, T. A. Kurashvili, and G. N. Korres, “Optimal coordination of directional overcurrent relays in distribution
systems with distributed generation based on a hybrid PSO-LP algorithm,” in MedPower 2014, 2014, vol. 2014, no. CP665, pp. 64
(6 pp.)-64 (6 pp.), doi: 10.1049/cp.2014.1697.
[16] A. Ahmarinejad, S. M. Hasanpour, M. Babaei, and M. Tabrizian, “Optimal overcurrent relays coordination in microgrid using
Cuckoo algorithm,” Energy Procedia, vol. 100, pp. 280–286, Nov. 2016, doi: 10.1016/j.egypro.2016.10.178.
[17] H. R. E. H. Bouchekara, M. Zellagui, and M. A. Abido, “Optimal coordination of directional overcurrent relays using a modified
electromagnetic field optimization algorithm,” Applied Soft Computing Journal, vol. 54, pp. 267–283, 2017, doi:
10.1016/j.asoc.2017.01.037.
[18] M. Alipour, S. Teimourzadeh, and H. Seyedi, “Improved group search optimization algorithm for coordination of directional
overcurrent relays,” Swarm and Evolutionary Computation, vol. 23, pp. 40–49, Aug. 2015, doi: 10.1016/j.swevo.2015.03.003.
[19] A. Sahoo and S. Chandra, “Multi-objective grey wolf optimizer for improved cervix lesion classification,” Applied Soft Computing
Journal, vol. 52, pp. 64–80, 2017, doi: 10.1016/j.asoc.2016.12.022.
[20] M. H. Sulaiman, Z. Mustaffa, M. R. Mohamed, and O. Aliman, “Using the gray wolf optimizer for solving optimal reactive power
dispatch problem,” Applied Soft Computing Journal, vol. 32, pp. 286–292, 2015, doi: 10.1016/j.asoc.2015.03.041.
[21] H. M. Song, M. H. Sulaiman, and M. R. Mohamed, “An application of grey wolf optimizer for solving combined economic emission
dispatch problems,” International Review on Modelling and Simulations, vol. 7, no. 5, pp. 838–844, Oct. 2014, doi:
10.15866/iremos.v7i5.2799.
[22] S. Mirjalili, S. M. Mirjalili, and A. Lewis, “Grey wolf optimizer,” Advances in Engineering Software, vol. 69, pp. 46–61, Mar.
2014, doi: 10.1016/j.advengsoft.2013.12.007.
[23] B. Chattopadhyay, “An online relay coordination algorithm for adaptive protection using linear programming technique,” IEEE
Power Engineering Review, vol. 16, no. 1, pp. 47–48, 1996.
[24] H. H. Zeineldin, E. F. El-Saadany, and M. M. A. Salama, “Optimal coordination of overcurrent relays using a modified particle
swarm optimization,” Electric Power Systems Research, vol. 76, no. 11, pp. 988–995, 2006, doi: 10.1016/j.epsr.2005.12.001.
[25] P. P. Bedekar and S. R. Bhide, “Optimum coordination of directional overcurrent relays using the hybrid GA-NLP approach,” IEEE
Transactions on Power Delivery, vol. 26, no. 1, pp. 109–119, Jan. 2011, doi: 10.1109/TPWRD.2010.2080289.

Int J Pow Elec & Dri Syst, Vol. 15, No. 2, June 2024: 1276-1289
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1289

[26] C. S. Karavas, K. G. Arvanitis, and G. Papadakis, “Optimal technical and economic configuration of photovoltaic powered reverse
osmosis desalination systems operating in autonomous mode,” Desalination, vol. 466, pp. 97–106, Sep. 2019, doi:
10.1016/j.desal.2019.05.007.
[27] H. Y. Mahmoud, H. M. Hasanien, A. H. Besheer, and A. Y. Abdelaziz, “Hybrid cuckoo search algorithm and grey wolf optimiser‐
based optimal control strategy for performance enhancement of HVDC‐based offshore wind farms,” IET Generation, Transmission
& Distribution, vol. 14, no. 10, pp. 1902–1911, May 2020, doi: 10.1049/iet-gtd.2019.0801.
[28] W. Long, J. Jiao, X. Liang, and M. Tang, “An exploration-enhanced grey wolf optimizer to solve high-dimensional numerical
optimization,” Engineering Applications of Artificial Intelligence, vol. 68, pp. 63–80, Feb. 2018, doi:
10.1016/j.engappai.2017.10.024.
[29] S. Gupta and K. Deep, “A novel random walk grey wolf optimizer,” Swarm and Evolutionary Computation, vol. 44, pp. 101–112,
2019, doi: 10.1016/j.swevo.2018.01.001.
[30] A. Saxena, R. Kumar, and S. Das, “β-Chaotic map enabled grey wolf optimizer,” Applied Soft Computing Journal, vol. 75, pp. 84–
105, Feb. 2019, doi: 10.1016/j.asoc.2018.10.044.
[31] U. Mlakar, I. Fister, and I. Fister, “Hybrid self-adaptive cuckoo search for global optimization,” Swarm and Evolutionary
Computation, vol. 29, pp. 47–72, Aug. 2016, doi: 10.1016/j.swevo.2016.03.001.
[32] T. Khurshaid, A. Wadood, S. Gholami Farkoush, J. Yu, C. H. Kim, and S. B. Rhee, “An improved optimal solution for the directional
overcurrent relays coordination using hybridized whale optimization algorithm in complex power systems,” IEEE Access, vol. 7,
pp. 90418–90435, 2019, doi: 10.1109/ACCESS.2019.2925822.
[33] H. R. E. H. Bouchekara, M. Zellagui, and M. Abido, “Coordination of directional over current relays using backtracking search
algorithm,” Journal of Electrical Systems, vol. 12, no. 2, pp. 387–405, 2016.

BIOGRAPHIES OF AUTHORS

Arathi Pothakanahalli Bheemasenarao received B.E degree in Electrical and

Electronics Engineering from Visvesvaraya Technological University. Belagavi, Karnataka
in 2003 and M. Tech degree in Digital Electronics from Visvesvaraya Technological
University-Belagavi, Karnataka in 2011.Currently She is working as Assistant Professor at
the Department of Electrical and Electronics Engineering. Ballari Institute of Technology and
Management, Ballari. Her research interest includes relay co-ordination, application of
evolutionary and heuristic optimization techniques to power systems. She can be contacted
at email: [email protected].

Shankaralingappa Channappa Byalihal was born in Raichur, Karnataka, India

on June 01, 1969. He received his B.E (Electrical) and M.E (Energy systems) degrees from
Karnataka University Dharwad, India in 1993 and 1994 respectively and Ph.D. (Power
Systems) from Visveswaraya Technological University, Belgaum, Karnataka, India in 2011.
Currently he is working as a professor in the department of Electrical and Electronics
Engineering at Dr. Ambedkar Institute of Technology, Bengaluru, India. His current research
interests include renewable integration, electric vehicle and meta heuristic algorithms. He is
a life member of the Indian Society for Technical Education (ISTE). He can be contacted at
email: [email protected].

Optimal coordination of directional over current relays for … (Arathi Pothakanahalli Bheemasenarao)