ALGERIAN JOURNAL OF SIGNALS AND SYSTEMS (AJSS)

Improvement of Voltage Profile by Optimal

Integration of PV-DG in Distribution Power
System
(1)* (2) (3) (4) (5)
Raimon Bawazir , Numan Çetin , Mustafa Mosbah , Salem Arif , Rabie zine
(1)(2)
dept. Solar Energy Institute of Sciences, EGE University, Izmir, Turkey
(3)(4)
dept. LACoSERE Laboratory Amar Telidji University, Laghouat, Algeria
(5)
Department of Mathematics, Faculty of Sciences, Moulay Ismaïl University, Meknès, Meknès,
Morocco
*
[email protected]

Abstract: Integrating Distributed Generation (DG) based on renewable energy sources such as Photovoltaic
(PV), has become one of the governments priorities in the use of clean energy on a large scale. Since the
radial power system is not completely prepared for high injection of PV-DG, the optimization method has to be
applied to PV-DG. This paper aims to find the optimal location and size of the PV-DG to improve the voltage
profile using the Biogeography-Based Optimization algorithm (BBOA). Due to the voltage problems occurs
more at low voltage in nature than the high voltage, BBOA is implemented on the 10, 69, 85, and 118 bus
IEEE distribution systems using the MATLAB code. The results of this study show that the voltage profiles in
all different power systems are improved.
Keywords: voltage deviation, distributed generation, biogeography-based optimization
effective solutions used to mitigate the power
1. INTRODUCTION system problems.
The global electricity demand is rapidly
increased due to several factors such as
industrial development, population growth,
and household equipment evolution.
Therefore, there is a strong seek for
governments to use clean energy sources in
the electric sector widely[1]. This explains
why in the year 2018, the total global
capacity of renewable power (without
hydropower) jumped up to 1246 GW
compared with 657GW in the year 2014.
Among all RESs, solar PV (data are provided
in direct current), Concentrated Solar Power
(CSP) and solar hot water (only data are for Fig. 1 Solar power capacity of different application
glazed and unglazed collectors) are the most between 2014 and 2018
source used over the entire world. Solar
applications have a considerable increase Integrating DG in terms of location and
rapidly for the last years as illustrated in size in the distribution power system can
Figure (1)[2][3]. However, the radial show slight improvements at a particular
distribution power system is not completely location. However, this does not reflect the
designed to deal with DG based on Wind entire picture of the system's strength and
Turbine (WT) or Photovoltaic (PV). In performance. It also prevents the system to
addition, power loss and poor voltage profile exploit the full merits of DG integration. On
are other issues that stand up to a the other hand, the unsuitable location and
distribution power system [4]. Therefore, the size of DG may put the entire system at risk.
electric utility attempts to tackle such issues DG can cause negative sides to the
by enhancing the performance of the system. distribution power system such as voltage
Distributed Generation (DG) is one of the limit violation, more power loss, protection
dysfunction against short circuits, and
exceeding transit capacity of some

Vol. 6, Issue 4, December-2021| ISSN: 2543-3792- EISSN: 2676-1548 199

 ALGERIAN JOURNAL OF SIGNALS AND SYSTEMS (AJSS)

equipment[5][6][7]. In order to avoid and on low voltage, it is usually exposed to the

mitigate the aforementioned problems, DG voltage degradation along the feeder in radial
has to be integrated with the system network structure. Due to this issue, voltage
optimally[8][9][10][11]. drops is necessary to be reduced by
The integration of optimal location and improving the voltage system profiles as in
size of DG with the power system is a equation (1).
nonlinear problem along with constraints.
-Minimizing voltage deviation
Based on this, optimization algorithms are
required to handle such problem[12][13][14]. (1)
There are many studies conducted for
optimizing DG as in the literature. Thus, this objective function can be
The study in[15] attempted to inject DG in computed as in equation (2):
distributed power network optimally. The (2)
authors applied an innovative exhaustive
search method to find the most suitable where is voltage at the ith bus, is the
location. In a similar study, a hybrid algorithm bus number. The represents the
of market-based Optimal Power Flow (OPF) deviation between the voltage at bus and
and Particle Swarm Optimization Algorithm the reference voltage (1 pu).
(PSOA) was used. This algorithm was 2.2 Balance equations
applied to determine the optimal DGs in
terms of locations and sizes[16]. In[17], other The balance equations are the power
methods based on linear programming balance constraints with DG. They are two
optimization were also used to find the nonlinear recursive power equations, it can
appropriate location and size of DG. The be represented for bus i as bellow:
methods were used to get the high
penetration of DG inserted into the (3)
distribution power network associated with
constant line capacity taking into where and are the total active power of
consideration the network security. For the generator and load while and are the
study conducted in[18], the optimizing DG in total reactive power of generator and load,
terms of location and size was found by are presented the total active and
applying a probabilistic based planning reactive power system losses, and and
technique. are the total active and reactive power of
There are also other algorithms used to DG units respectively.
determine the best location and size of DG in
literature. For example; Particle Swarm 2.3 Inequality constraints
Optimization Algorithm (PSOA)[19][20], These constraints shows the boundaries of
Modified Teaching-Learning Based various components in the system.
Optimization[21], Cuckoo Search
Algorithm[22], a hybrid of PSOA and Genetic -Photovoltaic Generation constraints
Optimization Algorithm (GOA)[23][24], Stud (4)
Krill herd Algorithm[25], Artificial Bee
Colony[26], and Differential Evolution (5)
in[27][28][29][30].
-Voltage constraints
The improving voltage profile is
considered as an objective function for our (6)
study. Biogeography-Based Optimization
where and are the voltage limits
algorithm (BBOA) is applied to find both
the ith bus[32].
location and size of PV-DG in different IEEE
distribution systems 10-bus, 69-bus, 85-bus, -Thermal limits of line
and 118-bus.
(7)
2. PROBLEM FORMULATION where is a maximum admissible
2.1 Objective function loading limit of kth branch and is a line
loading.
One of the main reasons causes the
blackouts in the power system is referred to 2.4 Handing of constraints
as voltage collapse [31]. Since the By using random strategy, it is agreed upon
distribution power system is designed based that the variables of control are generated in

Vol. 6, Issue 4, December-2021| ISSN: 2543-3792- EISSN: 2676-1548 200

 ALGERIAN JOURNAL OF SIGNALS AND SYSTEMS (AJSS)

their allowable limits. In order to treat the represents the main steps of the BBOA
inequality constraints of the variables of state algorithm.
which are consisting of lines loading and load
bus voltage magnitudes, the objective Island n Island 2
function is can be extended and formulated Island 4

mathematically as in equations (8-10)

[33][34]: Island 1

Island 3

(8)
emigrating islands (individuals)

(9) ----
immigrating island (individual)

Fig. 2 Species migrate between islands via

(10) flotsam, wind, flying, swimming and etc.

where and are indicated the factors of

penalty. For this work, the penalty factors
values have been taking in consideration
10000.

3. BIOGEOGRAPHY-BASED
OPTIMIZATION ALGORITHM
Biogeography-Based Optimization Algorithm
(BBOA) algorithm is the study of the natural
phenomena or behavior of distributing
biological organisms geographically[35].
Fig. 3 BBOA algorithm steps[40][41]
BBOA is fırst introduced by the scientist Dan
Simon. It optimizes a function by iteratively
and stochastically, thus BBOA is considered 4. RESULTS AND DECISION
as a type of evolutionary algorithm[36]. Due More voltage deviation occurs in low
to emigration and immigration in every single voltage level as in nature. Therefore, this
habitat, the information is exchanged through study is conducted on the different IEEE
the migration operation of species. Therefore, power distribution systems as listed below:
BBOA has certain features in common with
other population-based optimization 1- 10-bus IEEE distribution network
algorithms like PSO and GAs in terms of 2- 69-bus IEEE distribution network
sharing information between
solutions[37][38]. 3- 85-bus IEEE distribution network

In BBOA, every individual is considered 4- 118-bus IEEE distribution network

as an island with a Habitat Suitability Index In addition, handling complex or large system
(HSI). Meanwhile, every variable of an island is one of the influence factors that indicates
is a Suitability Index Variable (SIV). The the good performance of any optimization
island with a low HSI indicates a small algorithm as mentioned in[24]. For that
number of species. This represents a poor reason, BBOA is applied on small, medium
solution. On the other hand, a high HSI has a and large scales of the systems. In this study,
large number of species which means a good a single PV-DG with unity power factor is
solution. Based on this, species often migrate injected on in each distribution system. The
from the high HSI islands to low HIS islands minimum and maximum limits of voltage
and the low HSI islands accept system is set in the range ±10 pu.
immigrants[39]. Figure (2) shows the species
migrate between islands while Figure (3) The study shows that the injection of PV-
DG yields to improve all power system

Vol. 6, Issue 4, December-2021| ISSN: 2543-3792- EISSN: 2676-1548 201

 ALGERIAN JOURNAL OF SIGNALS AND SYSTEMS (AJSS)

profiles in terms of voltage and real power PV-DG respectively. Meanwhile, the
loss ( ) as shown in Table (1) and Figure minimum voltage is improved from 0.9092 at
(4). Before PV-DG, the Voltage Profile Index bus 65 to 0.9789 at bus 27 when the BBOA
(VPI) is 0.6963 where the minimum voltage is is applied. Similar to 85-bus system, the VPI
0.8382 at the bus10 in the case of 10-bus and are minimized from 7.7843 and
system. After optimized PV-DG (size is 0.3151 to 1.9385 and 0.2429 respectively.
6.7958 and location is at bus 9), all voltage The minimum voltage is also improved from
buses in the system are improved. VPI is 0.8715 at bus 54 to 0.9459 at bus 76
reduced to 0.1557 while the minimum voltage respectively. In 118-bus system, the VPI is
is improved to 0.9702 at the bus 6. It is reduced from 5.2450 to 3.7830 after installing
worthy to mention that, is minimized up PV-DG optimally. The minimum voltage is
to 63.27%. enhanced from 0.8688 at bus 77 to 0.9053 at
bus 111 respectively as in Figure (4). The
In the case of 69-bus system, the VPI and
real power loss profile is also reduced by
are reduced from 1.8364 and 0.2251 14.64%.
before PV-DG to 0.4547 and 0.1785 after
Table 1 The simulated results of four distributed power network before and after PV-DG
installation using BBOA
Wıthout PV-DG ınstallatıon With PV-DG ınstallatıon
Test
Min Optimal Optimal size Min
system
(MW) voltage location (WM) (MVAr) MW % voltage

10-bus 0.7838 0.6963 0.8382 9 6.7958 0 0.2879 63.27 0.1557 0.9702

69-bus 0.2251 1.8364 0.9092 57 3.7946 0 0.1785 20.70 0.4547 0.9789

85-bus 0.3151 7.7843 0.8715 31 2.5698 0 0.2429 22.91 1.9385 0.9459

118-bus 1.2981 5.2450 0.8688 70 4.9963 0 1.1081 14.64 3.7830 0.9053

Vol. 6, Issue 4, December-2021| ISSN: 2543-3792- EISSN: 2676-1548 202

 ALGERIAN JOURNAL OF SIGNALS AND SYSTEMS (AJSS)

Fig. 4 System voltage profile of different network scales with and without PV-DG injection

eclectic distribution network was modeled as

5. CONCLUSION
a nonlinear constrained optimization problem.
This study aims to search for the optimal The simulated result is proved the optimized
location and size of PV-DG to minimize the allocation of PV-DG has a significant
voltage deviation in the four IEEE distribution contribution to improving the power system
power systems. Herein, the applied profile. It indicates that there is considerable
optimization algorithm is called BBOA. Where improvement in terms of voltage and real
the best location and size of PV-DGs in the power loss. this improvement is ensured by

Vol. 6, Issue 4, December-2021| ISSN: 2543-3792- EISSN: 2676-1548 203

 ALGERIAN JOURNAL OF SIGNALS AND SYSTEMS (AJSS)

the proximity of the PV production to the Network Conference, CAGRE 2019, 2019, no.
points of load. 1, pp. 55–60.
[12] M. Mosbah, S. Arif, R. Zine, R. D.
In a further study, the authors intend to Mohammedi, and S. H. Oudjana, ‘Optimal size
extend this work to include more DG models and location of PV based DG-unit in
like wind turbine and objective functions such transmission system using GA method for loss
as harmonics, and voltage stability. reduction’, J. Electr. Eng., no. January 2018,
2017.
[13] M. Mosbah, A. Khattara, M. Becherif, and
References
S. Arif, ‘Optimal PV Location Choice
[1] M. Mosbah, R. Bawazir, N. Cetin, S. Arif, K. Considering Static and Dynamic Constraints’,
Sebaa, ‘Voltage Stability Maximization by Int. J. Emerg. Electr. Power Syst., vol. 18, no.
Optimal Wind Turbine Insertion Using Genetic 1, 2017.
Algorithm’, Algerian Journal of Signals and [14] Y. Bakelli, M. Mosbah, A. Kaabeche, and
Systems, Vol. 6, No. 1, 2021. Acimi Manel, ‘Voltage Stability Improvement by
[2] REN21, ‘Renewables 2015 Global Status Optimal Location of Wind Sources’, in 2019 1st
Report’, 2015. International Conference on Sustainable
[3] N. Kabouche, F. Chellali, AEM. Recioui, ‘A Renewable Energy Systems and Applications
review on solar radiation assessment and (ICSRESA), 2019.
forecasting in Algeria’, Algerian Journal of [15] P. Lamaina, D. Sarno, P. Siano, A.
Signals and Systems, Vol. 6, No. 3, 2021. Zakariazadeh, and R. Romano, ‘A model for
[4] R. Bawazir, N. Cetin, M. Mosbah, and S. Arif, wind turbines placement within a distribution
‘Improvement of the Voltage Profile of the network acquisition market’, IEEE Trans. Ind.
Distribution Network by Optimal Integration of Informatics, vol. 11, no. 1, pp. 210–219, 2015.
PVbased Decentralised Source’, 2020, pp. 1– [16] P. Siano and G. Mokryani, ‘Assessing
4. wind turbines placement in a distribution
[5] W. El-Khattam, Y. Hegazy, and M. Salama, ‘An market environment by using particle swarm
integrated distributed generation optimization optimization’, IEEE Trans. Power Syst., vol. 28,
model for distribution system planning’, 2005 no. 4, pp. 3852–3864, 2013.
IEEE Power Eng. Soc. Gen. Meet., vol. 3, no. [17] D. J. Burke and M. J. O’Malley,
2, p. 2392, 2005. ‘Maximizing firm wind connection to security
[6] M. Mosbah, R. Zine, Hamid-Oudjana, and S. constrained transmission networks’, IEEE
Arif, ‘Optimal Location and Size of Wind Trans. Power Syst., vol. 25, no. 2, pp. 749–
Source in Large Power System for Losses 759, 2010.
Minimization’, Inter Confere Artif. Intellige [18] Y. M. Atwa and E. F. El-Saadany,
Renew Energ. Syst Smart Energy Empower. ‘Probabilistic approach for optimal allocation of
Smart Resilient Cities Springer, Cham, p. wind-based distributed generation in
566_574, 2019. distribution systems’, IET Renew. Power
[7] M. Mosbah, R. Zine, S. Arif, R. D. Mohammedi, Gener., vol. 5, no. 1, pp. 79–88, 2011.
and S. Bacha, ‘Optimal Power Flow for [19] M. P. Lalitha, V. C. V. Reddy, and V.
Transmission System with Photovoltaic Based Usha, ‘Optimal DG placement for minimum
DG Using Biogeography-Based Optimization’, real power loss in radial distribution systems
in Proceedings of 2018 3rd International using PSO’, J. Theor. Appl. Inf. Technol., vol.
Conference on Electrical Sciences and 13, no. 2, pp. 107–116, 2010.
Technologies in Maghreb, CISTEM 2018, [20] M. M. Aman, G. B. Jasmon, A. H. A.
2019, pp. 1–6. Bakar, and H. Mokhlis, ‘A new approach for
[8] M. Mosbah, A. Khattara, M. Becherif, and S. optimum DG placement and sizing based on
Arif, ‘Optimal PV Location Choice Considering voltage stability maximization and minimization
Static and Dynamic Constraints’, Int. J. Emerg. of power losses’, Energy Convers. Manag., vol.
Electr. Power Syst., vol. 18, no. 1, 2017. 70, pp. 202–210, 2013.
[9] R. D. Mohammedi, A. Hellal, S. Arif, and M. [21] J. A. Martín García and A. J. Gil Mena,
Mosbah, ‘Optimal DG placement and sizing in ‘Optimal distributed generation location and
radial distribution systems using NSGA-II for size using a modified teaching–learning based
power loss minimization and voltage stability optimization algorithm’, Int. J. Electr. Power
enhancement’, Int. Rev. Electr. Eng., vol. 8, no. Energy Syst., vol. 50, pp. 65–75, 2013.
6, pp. 1806–1814, 2013. [22] W. S. Tan, M. Y. Hassan, M. S. Majid,
[10] R. D. Mohammedi, M. Mosbah, and A. and H. A. Rahman, ‘Allocation and sizing of
Kouzou, ‘Multi-objective optimal scheduling for DG using Cuckoo Search algorithm’, in PECon
adrar power system including wind power 2012 - 2012 IEEE International Conference on
generation’, EEA - Electroteh. Electron. Power and Energy, 2012, pp. 133–138.
Autom., vol. 66, no. 4, pp. 102–109, 2018. [23] M. H. Moradi and M. Abedini, ‘A
[11] M. Mosbah, R. D. Mohammedi, and S. combination of genetic algorithm and particle
Arif, ‘Differential Evolution Method for Optimal swarm optimization for optimal DG location
Size and Localization of Photovoltaic in the and sizing in distribution systems’, Int. J.
Algerian Transmission Power System’, in Electr. Power Energy Syst., 2012.
Proceedings of Algerian Large Electrical [24] R. O. Bawazir and N. S. Cetin,

Vol. 6, Issue 4, December-2021| ISSN: 2543-3792- EISSN: 2676-1548 204

 ALGERIAN JOURNAL OF SIGNALS AND SYSTEMS (AJSS)

‘Comprehensive overview of optimizing PV-DG Technol. Maghreb, Cist. 2014, no. 1, pp. 1–6,
allocation in power system and solar energy 2014.
resource potential assessments’, Energy [34] M. Mosbah, R. D. Mohammedi, S. Arif,
Reports, vol. 6, pp. 173–208, 2020. and A. Hellal, ‘Optimal of shunt capacitor
[25] S. A. ChithraDevi, L. placement and size in Algerian distribution
Lakshminarasimman, and R. Balamurugan, network using particle swarm optimization’, in
‘Stud Krill herd Algorithm for multiple DG Proceedings of 2016 8th International
placement and sizing in a radial distribution Conference on Modelling, Identification and
system’, Eng. Sci. Technol. an Int. J., vol. 20, Control, ICMIC 2016, 2017, pp. 192–197.
no. 2, pp. 748–759, 2017. [35] R. Kumar, R. A. Gupta, and A. K. Bansal,
[26] F. S. Abu-Mouti and M. E. El-Hawary, ‘Economic analysis and power management of
‘Optimal Distributed Generation Allocation and a stand-alone wind/photovoltaic hybrid energy
Sizing in Distribution Systems via Artificial Bee system using biogeography based optimization
Colony Algorithm’, in Power Delivery, IEEE algorithm’, Swarm Evol. Comput., vol. 8, pp.
Transactions on, 2011, vol. 26, no. 4, pp. 33–43, 2013.
2090–2101. [36] D. Simon, ‘Biogeography-Based
[27] L. D. Arya, A. Koshti, and S. C. Choube, Optimization’, IEEE Trans. Evol. Comput., vol.
‘Distributed generation planning using 12, no. 6, pp. 702–713, 2008.
differential evolution accounting voltage [37] H. Ma, ‘An analysis of the equilibrium of
stability consideration’, Int. J. Electr. Power migration models for biogeography-based
Energy Syst., vol. 42, no. 1, pp. 196–207, optimization’, Inf. Sci. (Ny)., vol. 180, no. 18,
2012. pp. 3444–3464, 2010.
[28] H. A. Hejazi, M. A. Hejazi, G. B. [38] M. Q. Duong, T. D. Pham, T. T. Nguyen,
Gharehpetian, and M. Abedi, ‘Distributed A. T. Doan, and H. Van Tran, ‘Determination of
generation site and size allocation through a optimal location and sizing of solar photovoltaic
techno economical multi-objective differential distribution generation units in radial
evolution algorithm’, in PECon2010 - 2010 distribution systems’, Energies, vol. 12, no. 1,
IEEE International Conference on Power and 2019.
Energy, 2010, pp. 874–879. [39] Q. Niu, L. Zhang, and K. Li, ‘A
[29] W. S. Tan, M. Y. Hassan, M. S. Majid, biogeography-based optimization algorithm
and H. Abdul Rahman, ‘Optimal distributed with mutation strategies for model parameter
renewable generation planning: A review of estimation of solar and fuel cells’, Energy
different approaches’, Renew. Sustain. Energy Convers. Manag., vol. 86, pp. 1173–1185,
Rev., vol. 18, pp. 626–645, 2013. 2014.
[30] M. Mosbah, S. Arif, R. D. Mohammedi, S. [40] K. D. and D. Kalyanmoy, Multi-Objective
Hamid Oudjana, and R. Zine, ‘Multiobjective Optimization Using Evolutionary Algorithms.
Optimization for Many Wind Based DG John Wiley & Sons, Inc., Hoboken, New
Position and Sizes in Practical Distribution Jersey, 2001.
Network Using Modified Differential Evolution [41] M. Takahashi and H. Kita, ‘A crossover
Technique’, in Proceedings - PAIS 2018: operator using independent component
International Conference on Pattern Analysis analysis for real-coded genetic algorithms’, in
and Intelligent Systems, 2018. Proceedings of the IEEE Conference on
[31] R. Bawazir and M. Wazir, ‘The Different Evolutionary Computation, ICEC, 2001, vol. 1,
Performance of UPFC and STATCOM Based pp. 643–649.
on Opti- mal Location for Enhancing Voltage
Stability’, 2014, p. 10.
[32] R. O. B. M. W. Mustafa, ‘Optimal
Location of FACTS for ATC Enhancement’, Int.
J. Eng. Adv. Technol., vol. 3, no. 1, pp. 80–85,
2013.
[33] M. Mosbah, A. Hellal, R. D. Mohammedi,
and S. Arif, ‘Genetic Algorithms based Optimal
Load Shedding with Transient Stability
Constraints’, 2014 Int. Conf. Electr. Sci.

Vol. 6, Issue 4, December-2021| ISSN: 2543-3792- EISSN: 2676-1548 205