INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH

Saeid Lotfi Trazouei et al., Vol.3, No.2, 2013

Optimal Design of a Hybrid Solar -Wind-Diesel

Power System for Rural Electrification Using
Imperialist Competitive Algorithm

Saeid Lotfi Trazouei*‡, Farid Lotfi Tarazouei**, Mohammad Ghiamy**

*Department of Energy and Environmental Engineering, Science and Research Branch, Islamic Azad University
**Department of Electrical Engineering, Ardabil Branch, Islamic Azad University
s_lotfi@srbiau.ac.ir, fl0tfi@yshoo.com, m.ghiamy@iauardabil.ac.ir

‡
Corresponding Author; Saeid Lotfi Trazouei, Department of Energy and Environmental Engineering, Science and Research
Branch, Islamic Azad University, Tehran, Iran, +989398208778, s_lotfi@srbiau.ac.ir

Received: 29.03.2013 Accepted: 19.05.2013

Abstract- In this paper, optimal design of a stand-alone hybrid solar- wind- diesel power generation system using Imperialist
Competitive Algorithm, Particle swarm optimization and ant colony optimization is presented. The final goal of this paper is
minimization of net present cost of hybrid system for lifetime of project ( here 20 years) considering by reliable supply of load
and loss of power probability (LPSP) reliability index. In order to find out the least expenditure and best combination, the
result of these algorithms compared together. Among these algorithms, the imperialist competitive algorithm is faster and more
accurate than others and has more certain design in comparison to PSO and ACO algorithms. In this paper, first the
mathematical model of various parts of hybrid system is presented. Then the purposed algorithm is used. Finally, simulation
results ( number of PV panels, number of wind turbines, number of battery storages, system total cost ,power diagram of
hybrid power system components and reliability diagram) for solar-wind -diesel systems is presented.
Keywords- Renewable energy, stand- alone power system, reliability, Imperialist competitive algorithm, optimization.

1. Introduction energy requirements of remote areas [4]. In [5] have

developed the Hybrid Optimization by Genetic Algorithms
In recent years, a considerable growth in using (HOGA) program helping to determine the optimal
renewable energy resources has been observed. Specially, configuration of the hybrid PV/diesel system. [6] presented a
solar and wind energy are infinite, site-dependent, non- techno-economic analysis based on solar and wind biased
polluting, and high potential sources for alternative energy months for an autonomous hybrid PV/wind energy system.
production. For remote systems such as radio He has observed that an optimum combination of the hybrid
telecommunications, satellite earth stations, or at sites far PV/wind energy system provides higher system
from a conventional power system, the hybrid energy performances than either of the single systems for the same
systems is referable [1,2]. Such systems are usually equipped system cost for a given battery storage capacity. He has also
with diesel generators to meet the peak load demand during observed that the magnitude of the battery
short periods, when the available energy is deficit [3]. Hybrid
storage capacity has an important bearing on the system
energy systems are best suited to reduce dependence on
performances of single photovoltaic and wind systems. [7]
fossil fuel using available wind speed and solar radiations. A
analyzed solar radiation data of Rafha, K.S.A., to assess the
configuration of a hybrid energy system is shown in Fig. 1. It
techno-economic feasibility of hybrid PV- diesel-battery
includes PV panels and/or wind turbines and/or diesel
ower systems to meet the load requirements of a typical
generator and/or batteries. These energy systems are
remote village Rawdhat Bin Habbas (RBH) with annual
considered as one of the cost effective solutions to meet
electrical energy demand of 15,943 MWh. HOMER software
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
Saeid Lotfi Trazouei et al., Vol.3, No.2, 2013
has been used to perform the techno-economic evaluation. In where PPV is the PV panel power (W) at the maximum
[8], The optimal model of the solar system for a region of power point at day t, Vmpp is the PV panel voltage at the
Turkey is done using the SA algorithm. [9] presents the
analysis and design of a mixed integer linear mathematical maximum power point (V) at day t, Vmpp ,ref is Vmpp at
programming model (time series) to determine the optimal reference operating conditions (V), I mpp is the PV panel
operation and cost optimization for a hybrid energy
generation system consisting of a photovoltaic array, current at the maximum power point (A) at day t, I mpp,ref is
biomass (fuel wood),biogas, small/micro-hydro, a battery I mpp at reference operating conditions (A), I SC ,ref is the
bank and a fossil fuel generator for a remote rural of India.
short circuit current at reference operating conditions (A),
GT is the daily irradiance on a tilted surface (W/m²), Gref is
the irradiance of 1000W/m² at reference operating
conditions, V ,OC and I , SC are the temperature coefficients
for open circuit voltage (V/°C) and short circuit current
(A/°C) respectively, TC ,ref is the PV panel temperature of
25°C at reference operating conditions and TC corresponds
to the PV panel operating temperature (°C) at day t and
which can be expressed as follows [18] :
NOCT  20
TC  t   Ta  t   .GT (4)
Fig. 1. Schematic of a hybrid solar- wind- diesel power 800
system[2]
where Ta is the ambient temperature (°C) of the site under
The main goal of this paper is the size optimization of consideration at hour t and NOCT (Normal Operating Cell
hybrid solar- wind- diesel power system regarding to the Temperature) is defined as the cell temperature when the PV
LPSP reliability index to provide reliable time with panel operates under 800W/m² of solar irradiation and 20°C
minimum cost using an evolutionary algorithm based on of ambient temperature, NOCT is usually between 42°C and
socio-political behaviour (ICA). In order to find out the least 46°C [19,2].
expenditure and best combination, the result of this algorithm
compared by the PSO algorithm and ACO algorithm. Also The PV panels are connected in series to form strings,
the simulation results for solar-wind -diesel systems with where the number of panels to be connected in series N pv, s
reliability index are presented
is determined by the selected DC bus voltage ( Vbus ) as
follows [2]:
2. Mathematical model
VBus
N PV,S  (5)
The models used in this paper for optimal design of V pv , nom
hybrid systems is based on a sampled average daily weather
data, which has been recorded by Ardabil meteorological Where VPV ,nom is the nominal PV panel voltage. Then
organization. This data includes the ambient temperature, N PV , s is not subject to the optimization, whereas the number
solar irradiation and wind speed [10].
of parallel strings N pv, p is the design variable that needs
optimization.
2.1. PV Array Modeling

The output power from a PV panel can be calculated by 2.2. Wind Turbine Modeling
an analytical model which defines the current-voltage
relationships based on the electrical characteristics of the PV The relationship between output from wind turbine and
panel. This model includes the effects of radiation level and wind speed is as follows [11]:
panel temperature on the output power. With a maximum
power point tracker (MPPT), the output power from a PV  v  vei
panel is given as [10]:  v  v PR Vei  V  Vr
 t    PR
r eo
PWT Vr  V  Vei (6)
PPV  Vmpp .I mpp (1)  0 other wise

Vmpp  Vmpp,ref  V ,OC (TC  TC ,ref ) (2) 
Where V is sampling wind speed; vei is cut-in wind
I mpp  I mpp,ref  I SC ,ref (GT / Gref )  I , SC (TC  TC ,ref )
speed; veo is cut-out wind speed; vr is rated wind speed; PR
(3)
is rated wind turbine power.[11] Using the wind speed at a
reference height hr(m) from the database, the velocity at a

404
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
Saeid Lotfi Trazouei et al., Vol.3, No.2, 2013
specific hub height (h(m)) for the location is estimated on an The number of batteries to be connected in series is
hourly basis throughout the specified period through the therefore not subject to the optimization but is a
following expression.[2] straightforward calculation, whereas the number of parallel
 battery strings, each consisting of Nbat ,s batteries connected
 h
V  t   Vr  t    (7) in series, is a design variable that needs optimization [2].
 hr 
Fig. 2 shows typical wind turbine characteristics 2.4. Power Reliability Model Based on LPSP Index

Because of the intermittent solar radiation and wind

speed characteristics, which highly influence the resulting
energy production, power reliability analysis has been
considered as an important step in any system design
process. A reliable electrical power system means a system
has sufficient power to feed the load demand during a certain
period or, in other words, has a small loss of power supply
probability (LPSP). LPSP is defined as the probability that
an insufficient power supply results when the hybrid system
is unable to satisfy the load demand. It is a feasible measure
Fig. 2. Wind power versus wind speed. of the system performance for an assumed or known load
distribution. A LPSP of 0 means the load will always be
satisfied; and an LPSP of 1 means that the load will never be
2.3. Battery Bank modeling
satisfied. Loss of power supply probability (LPSP) is a
statistical parameter; its calculation is not only focused on
For the charging process and discharging process of the the abundant or bad resource period. Therefore, in a bad
battery bank, the state of charge (SOC) can be calculated resource year, the system will suffer from a higher
from the following equation: probability of losing power. There are two approaches for
 P t   the application of LPSP in designing a stand-alone hybrid
SOC  t  t   SOC  t   bat  B  t (8) system. The first one is based on chronological simulation.
 Vbus  This approach is computationally burdensome and requires
the availability of data spanning a certain period of time. The
where  bat is equal to the round-trip efficiency in the second approach uses probabilistic techniques to incorporate
charging process and is equal to the 100% in the discharging the fluctuating nature of the resource and the load, thus
process [12], Vbus is the DC bus voltage (V) and t is the eliminating the need for time-series data. Considering the
daily time step is set equal to 1day. For longevity of the energy accumulation effect of the battery, to present the
battery bank, the maximum charging rate ( SOCmax ) is given system working conditions more precisely, the chronological
method is employed in this research. The objective function,
as the upper limit, it is equal to the total nominal capacity of LPSP, from time 0 to T can then be described by
the battery bank ( Cn (Ah)) which is related to the total
∑
number of batteries ( Nbat ), the number of batteries ∑
(12)
connected in series ( Nbat ,s ) and the nominal capacity of each
Where LPS(t) can be calculated as follow:
battery ( Cn (Ah)), as follows
LPS (t )  Pload .t  ( Pp .t  SOC (t  1)
(9)  (SOCmin ))inv
(13)
The lower limit that the state of charge of the battery
Where Δt is the step of time used for the calculations (in
bank cannot exceed at the time of discharging ( SOCmin ) may
this study Δt =1 Day). During that time, the power produced
be expressed as follows by the hybrid system are assumed constant. So, the power is
SOCmin  1  DOD  .SOCmax (10) numerically equal to the energy within this time step [13,14].

The batteries are connected in series to give the desired 2.5. System Operation Strategies
nominal DC operating voltage and are connected in
parallel to yield a desired Ah system storage capacity. Then, The power generated by the hybrid system and the
the number of batteries connected in series depends on the amount of energy stored are time dependent. So, the
DC bus voltage and the nominal voltage of each individual input/output power of the battery bank, is controlled by the
battery Vbat ,nom , it is calculated as follows following equation []:

P  Pre (i)  Pl (i) (14)

(11)
Where
405
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
Saeid Lotfi Trazouei et al., Vol.3, No.2, 2013

Pload 3. Size Optimization Algorithm

Pl (i )  (15)
inv
3.1. Imperialist Competitive Algorithm
where Pload is the power demanded by the load and inv is
the DC/AC converter efficiency specified by the Optimal sizing algorithm the proposed algorithm starts
manufacturer. Depending on the P value, the following with an initial population (countries in the world). Some of
cases apply: the best countries in the population are selected to be the
imperialists and the rest form the colonies of these
If P  0 , the remaining power will be used to charge imperialists. All the colonies of initial population are divided
the batteries. If the batteries are completely charged, the among the mentioned imperialists based on their power. The
exceed power is dumped. power of an empire which is the counterpart of the fitness
If P  0 : the remaining power will be given by the value in GA, is inversely proportional to its cost. After
batteries or by the diesel generator, depending on the dividing all colonies among imperialists, these colonies start
dispatch strategy: moving toward their relevant imperialist country. The total
power of an empire depends on both the power of the
if the batteries are able to give P , then: batteries imperialist country and the power of its colonies. We will
discharge and diesel generator turns off model this fact by defining the total power of an empire by
the power of imperialist country plus a percentage of mean
Pb (i)  P (16)
power of its colonies .Then the imperialistic competition
if the batteries are not able to give P , the diesel begins among all the empires. Any empire that is not able to
generator turns on, and the batteries will neither be charged succeed in this competition and can’t increase its power (or
nor discharged. Then the diesel generator output power at least prevent decreasing its power) will be eliminated from
Pd (i) is expressed as follows the competition. The imperialistic competition will gradually
result in an increase in the power of powerful empires and a
Pd (i)  P (17) decrease in the power of weaker ones. Weak empires will
lose their power and ultimately they will collapse. The
The flowchart of this strategies shown in figure (3). movement of colonies toward their relevant imperialists
along with competition among empires and also the collapse
mechanism will hopefully cause all the countries to converge
to a state in which there exist just one empire in the world
and all the other countries are colonies of that empire. In this
ideal new world colonies, have the same position and power
as the imperialist [15].

3.2. Particle Swarm Optimization

Particle Swarm Optimization (PSO) is a relatively novel

approach for global stochastic optimization. It simulates a
swarm, or flock, of beings wandering in the parameter space
and is based on a "social interaction" metaphor in which the
path of the beings in the parameter space is controlled
according to a swarm- or flock-like set of rules [22].

3.3. Ant Colony Optimization

Ant colony algorithms are based on the behavior of

social insects with an exceptional ability to find the shortest
paths from the nest to the food sources using a chemical
substance called pheromone. [23].

4. Objective Function and Constraints

The Annualized Cost (AC) of component i is defined as

[16]:
AC ( x)  N  {[CC  RC  K (ir , L, y )]
Fig. 3. The flowchart of system operation strategies  CRF (ir , R)  CM } (18)

406
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
Saeid Lotfi Trazouei et al., Vol.3, No.2, 2013
where, N may be number (unit) or capacity (kW or kg), CC N PV  N PV , P  N PV , S is the total number of PV panels
is capital cost ($/unit), RC is cost of each replacement
($/unit), and O&MC is annual operation and maintenance and NWT is the total number of wind turbines.
cost ($/unit-yr) of the component, and R is the useful lifetime
of the project (here, 20 years). ir is the real interest rate (here, PB is the input/ output power of the battery bank,
6%) which is a function of nominal interest rate (irnominal) and Pb  0 during the charging process of the battery and Pb  0
annual inflation rate (fr), defined by in the discharging process .[2]
(ir , no min al  f r ) Then, the model of the sizing optimization of the
ir  (19) standalone hybrid PV-Wind-diesel system is expressed as
1  fr
follows
lso, CRF and K are capital recovery factor and single
payment present worth [20], respectively, which are defined Min STC(x) = minx{AC(x)+Cfuel} (26)
as follows: where x  {N PV , P , NWT , N Bat , P } (27)
(1  ir ) R  1 PP  t  PL  t  (28)
CRF (iR , R)  (20)
(1  ir ) ir
R
SOCmin  t   SOC  t   SOCmax  t  (29)
y

 (1  i ) LPSPmin  LPSP  LPSPmax

1 (30)
K (ir , L, y )  n L
(21)
n 1 r 0  N PV , P  N PV , P max (31)
where, L and y are useful lifetime and number of
replacements of the component during useful lifetime of the 0  NWT  NWT max (32)
project, respectively. Number of replacements of each
component is a simple function of useful lifetimes of the 0  N Bat , P  N Bat , P max (33)
component and the project [16].
The following equation shows the fuel consumption cost
for 1 h of running of the diesel generator [2]

C fuel  Pr fuel .  A.PD  B.PRD  (22)

where Pr fuel is the fuel price ($/l), A = 0.246 l/kW and
B= 0.0845 l/kW are the fuel curve coefficients.
The minimization of the objective function is subject to
the constraints that the power produced by the system is
larger than to the power demanded by the load and the state
of charge of the battery bank is limited between SOCmin and
SOCmax and LPSP is limited between LPSPmin and LPSPmax
as follows
 Pp  Pl
 SOCmin  SOC  SOCmax (23)
 LPSPmin  LPSP  LPSPmax
where PP is the power produced by the system, it is
calculated as follows
Pp  PRE  PD  Pb (24)
where PRE is the power produced by the renewable
resources as follows

PRE  N PV .Pmpp  NWT .PWT (25)

Pmpp and PWT are respectively the power at the maximum

power point of the PV panel (W) and the output power of the
wind turbine (W), N PV and NWT are respectively the total
number of PV panel and wind turbine and

Fig. 4. The flowchart for designing proposed hybrid system

407
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
Saeid Lotfi Trazouei et al., Vol.3, No.2, 2013
The goal of optimization is to find an optimal solution in
terms of the variables of the problem. The algorithm form an
array of variable values to be optimized. In ACO
terminology, this array is called "ant", In PSO terminology is
called "particle" but here the term" country" is used for this
array. In a N var - dimensional (in this paper Nvar  4 )
optimization problem, a country is a 1 Nvar array. this array
is defined by
Country  [ NPV ; NWT ; Nbat ; Ninv ] (34)

The variable values in the country are represented as

Fig. 5. Daily mean values of ambient temperature
floating point numbers. The cost of a country is found by
evaluating the cost function f at the variables (NPV ,NWT, Nbat,
Ninv).then
Cost  f (country)  f ( NPV ; NWT ; Nbat ; Ninv ) (35)

To start the optimization algorithm the initial population

of size N pop (here N pop  30 ) is generate. Then select N imp
of most powerful countries to form the empires. The
remaining Ncol of the population will be the colonies each of
which belongs to an empire. The formation of the N imp and
Ncol is explained completely in [15].Then for each initial
Fig. 6. Daily mean values of solar irradiation
population (in this algorithm countries) operation strategy is
checked for the overall performance during the year.
Populations that able to supply the load selected by the Wind speed measured at a height of 40 meters and has
algorithm. The algorithm continues to find the best been shown in the figure 4.
combination by minimum cost and compliance with the
relevant Constraints of the objective function is carried out.
This process is done until the number of iterations for the
algorithm to be finished and it is shown in figure (4). Finally
the best system with the minimum cost is obtained. In this
paper outputs are the optimum number of the components of
hybrid system ensuring that the system total cost is
minimized subject to the constraint that the load demand is
completely supplied. In this paper, number of population
(Npop for the ICA) is set to 30 and the number of iterations
(decades for the ICA) is set to 100.

5. Results and Discussion Fig. 7. Daily mean values of wind speed

Daily distribution of power demand for a year is shown
The proposed method was used for design a hybrid
in Figure(8)
system to provide variable power (15 kw peak load) for a
region of northwest of Iran (in Ardebil province). 12-month 16000

series of wind speed, solar radiation and ambient temperature 14000

in daily average form which recorded by the meteorological 12000

organization and shown in Figures 4 to 7 [21].

10000

8000

6000

4000

2000
0 50 100 150 200 250 300 350 400

Fig. 8. Daily mean values of demanded power during a year

A diesel generator has been used in this application, has
a rated power of 15 kW. The system components

408
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
Saeid Lotfi Trazouei et al., Vol.3, No.2, 2013
4
characteristics and initial investment, maintenance and 6.55
x 10

installation costs, are listed in Table 1 to 3 [2]. ACO

6.5 PSO
ICA
Table 1. PV panels specifications 6.45

VOC(V) 33.2 6.4

ISC(A) 8.85
6.35
Vmax(V) 26.6

Best Cost
Imax(A) 7.9 6.3

NCOT(oC) 47.9 6.25

Capital cost($/kw) 400

6.2
Replacement cost($) 400
Maintenance cost ($/year) 4 6.15

Table 2. Wind turbine specifications 6.1

6.05
Rated power (KW) 5 0 10 20 30 40 50 60 70 80 90 100
Decade (Iteration)
vr (m/s) 11
vci(m/s) 2.5
vCO(m/s) 24
Fig. 9. Convergence of optimization algorithms
Capital cost($) 7500 According to Figure (9) and Table (4), the ICA
Replacement cost($) 7500
Maintenance cost ($/year) 75
algorithm is faster and more accurate than other algorithms
and has more certain design in comparison to PSO and ACO
Table 3. Battery specifications algorithms.
4
Nominal capacity (Ah) 230 3
x 10

Voltage (v) 12 Pre

DOD (%) 80 Pl
Pp
Efficiency (%) 85 2.5

Capital cost($) 225

Replacement cost($) 225
2
Maintenance cost ($/year) 2
The PV panel lifetime is 20 years. The diesel generator
[W]
p

lifetime is 15000 h. The fuel price is 0.9 $/l. The capital cost
P e,P &P

1.5
l

and the maintenance cost of the diesel generator are

r

respectively 3000 $ and 3 $/h . 1

The DC bus voltage is chosen to be equal to 48 V.

Hence, the number of PV panels and batteries connected in 0.5

series are determined by fulfilling the DC bus voltage. The

expected battery lifetime has been set to 3 years with proper 0

maintenance, then Ybat  6 resulted.

0 50 100 150 200 250 300 350 400
Number of Days

Fig 10. Hybrid system power(Pp), renewable sources

Results of the optimal design of hybrid systems with the
power(Pre) and load demand (Pl)
overall system costs are shown in Table 4.
4
Table 4. Optimal results of different hybrid system 2.5
x 10

Optimization algorithm 2
ICA ACO PSO
optimal results
1.5

NPV 31 34 34
NWT 4 4 4 1
P [W]

NBat 16 15 15
b

Ninv 10 8 8 0.5

Total cost ($) 60670 61600 61730

0
Finally the cost of various combinations with diesel
generator has been shown in Figure 6 : -0.5

-1
0 50 100 150 200 250 300 350 400
Number of Days

Fig. 11. Input/output of battery bank during a year

409
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
Saeid Lotfi Trazouei et al., Vol.3, No.2, 2013
power providing. Also, the batteries SOC and system LPSP
14000
in hybrid system is shown.
12000 Figure (9) , shows the power produced by renewable
sources ( Pre ), the total power produced by the hybrid system
10000
( Pp ) and the demand power load solar - wind – diesel
8000 system. Figure (11), represents the input/ output power of the
P [W]

battery bank and determine how to charge and discharge

d

6000 them over the years for the solar - wind – diesel system.
Pb  0 is indicates the battery charging and the input power
4000
and Pb  0 represents the battery discharging and output
2000 battery power. Figure (12), are show the diesel usage over
the year. In this design the diesel generator will supply the
0
0 50 100 150 200 250 300 350 400
load, when the renewable sources and the battery banks can’t
Number of Days supply the load. According to figure it is noted that the diesel
generator does not work with the rated power. Its maximum
Fig. 12. Diesel generator power during a year output power in the hybrid systems is 13 KW, which
operated 1 days per year for hybrid solar- wind -diesel
110 systems. Also, for 256 days the diesel generator is off which
represents the fuel saving. Figure (13) shows that the state of
100 charge of the battery bank never exceeds the permissible
90 maximum value of SOCmax (100% of SOC), and never be
80
below the permissible minimum value of SOCmin (20% of
SOC). Figure (14) illustrate the reliability index (LPSP) of
70
the hybrid system during the year. If LPSP = 0, the designed
SOC %

60 system will be reliable supply the load, and if LPSP = 1, the

designed system, will not be able to supply the load. From
50
the above figures concluded that, the 15 kW load, with
40 system can be fully supported. In each time, the batteries and
the renewable resources are unable to supply the load, the
30
diesel generator turns on and provides the load.
20

10 6. Conclusion
0 50 100 150 200 250 300 350 400
Number of Days
In this paper the optimal design of a stand-alone hybrid
Fig. 13. State of charge of battery bank system using Imperialist Competitive Algorithm, Particle
swarm optimization and ant colony optimization is
100
presented. This design, was implemented considering the
reliable load providing and the LPSP reliability index, for
90
wind- solar- diesel system. The results show that the
80 Imperialist Competitive Algorithm is faster and more
70
accurate than other algorithms and has more certain design in
comparison to PSO and ACO algorithms. The costs were
60
obtained for Imperialist Competitive Algorithm, Particle
LPSP %

50 Swarm Optimization and Ant Colony optimization are

40
60670 $, 61730 $ and 61600 $, respectively. The reliability
index for the optimal combination of these systems are 3.7%,
30
4.4% and 4.1%, respectively which these indices are in the
20 normal range. The cost of fuel consumption using the hybrid
10 solar - wind- diesel system, is 34700 $/yr, while in the diesel
system is 7282.8 $/yr. Finally, the cost saving in fuel
0
0 50 100 150 200 250 300 350 400 consumption using the hybrid.
Number of Days

Fig. 14. Hybrid system reliability index (LPSP) References

Figures (9) to (13) represent the share of renewable
[1] Lotfi, Tarazuie, S. , Ghiamy, M , Kazemi Kargar, h.:'
sources, diesel generator and battery supply for reliable
Modeling of hybrid solar- wind- diesel energy system in

410
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
Saeid Lotfi Trazouei et al., Vol.3, No.2, 2013
rural areas', 2nd Iranian conference on renewable energy [13] B. Ould Bilal, V. Sambou, P.A. Ndiaye, C.M.F.
and distributed generation 2012, Tehran.Iran,P(138) Kébé, M. Ndongo.:'Optimal design of a hybrid solar-
wind-battery system using the minimization of the
[2] Belfkira, R.,Zhang L, Barakat, G:' Optimal Sizing study
annualized cost system and the minimization of the loss
of hybrid wind/PV/diesel power generation system ' Solar
of power supply probability (LPSP)'. Renewable
Energy ., 2011,85,100-110.
Energy.2010, 35,(10),pp. 2388-2390.
[3] Borowy, B.S., Salameh, Z.M.:' Methodology for
[14] Hongxing Yang, Wei Zhou, Lin Lu, Zhaohong
optimally sizing the combination of a battery bank and
Fang.:' Optimal sizing method for stand-alone hybrid
PV array in a wind/PV hybrid system'. IEEE Transactions
solar-wind system with LPSP technology by using
on Energy Conversion., 1996., 11 (2).,pp.367-375
genetic algorithm'. Solar Energy. 2008, 82, (4), pp. 354-
[4] Deshmukh, M.K., Deshmukh, S. S.: ' Modeling of hybrid 367.
renewable energy systems ' . Renewable & Sustainable
[15] Esmaeil, Atashpaz-Gargari, Caro Lucas,"
Energy Reviews . 2008 , 12,(1),pp. 235-249.
Imperialist Competitive Algorithm: An Algorithm for
[5] Dufo-Lopez, R., Bernal-Agustin, J.L.: ' Design and Optimization Inspired by Imperialistic Competition"
control strategies of PV-diesel systems using genetic 2007 IEEE Congress on Evolutionary Computation (CEC
algorithms '. Solar Energy , 2005,79, (1), pp.33-46. 2007).
[6] Celik, A.N.:' Optimization and techno-economic analysis [16] A. Kashefi Kaviani, G.H. Riahy, SH.M.
of autonomous photovoltaic-wind-hybrid energy systems Kouhsari.:'Optimal design of a reliable hydrogen-based
in comparison to single photovoltaic and wind systems'. stand-alone wind/PV generating system, considering
Energy Conversion and Management. 2002, 43 (18), component outages'. Renewable Energy, 2009, 34, (11),
pp.2453-2468. pp. 2380-2390.
[7] S.M. Shaahid , I. El-Amin. :'Techno-economic evaluation [17] R. Luna-Rubio, M. Trejo-Perea, D. Vargas-
of off-grid hybrid photovoltaic-diesel-battery power Vázquez, G.J. Ríos-Moreno.:' Optimal sizing of
systems for rural electrification in Saudi Arabia-A way renewable hybrids energy systems: A review of
forward for sustainable development' . Renewable and methodologies.' Solar Energy,2012, 86,(4),pp. 1077-
Sustainable Energy Reviews, 2009, 13,(3),pp. 625-633. 1088.
[8] Orhan Ekren, Banu Y. Ekren.:'Size optimization of a [18] Markvar, T.:' Solar Electricity', second ed. J. Wiley
PV/wind hybrid energy conversion system with battery & Sons. , 2000.
storage using simulated annealing' . Applied Energy.
[19] Yang, H., Lu, L., Zhou, W.: ' A novel optimization
2010, 87, (2), pp. 592-598.
sizing model for hybrid solar-wind power generation
[9] Ajai Gupta, R.P. Saini, M.P. Sharma.:' Modeling of system' . Solar Energy . 2007,81,pp. 76-84.
hybrid energy system-Part I: Problem formulation and
[20] Shahirinia AH, Tafreshi SMM, Gastaj AH,
model development'. Renewable Energy. 2010, 35, (2),
Moghaddomjoo AR. Optimal sizing of hybrid power
pp. 520-535.
system using genetic algorithm. In: International
[10] Belfkira, R., Barakat, G., Nichita, C.:' Sizing conference on future power systems 2005; 16-18
optimization of a stand-alone hybrid power supply unit: November 2005.
wind/PV system with battery storage'. International
[21] Meteorology Organization of Iran, Ardabil Province
Review of Electrical Engineering (IREE), 2008, 3 (5).
branch. Daily mean values of solar radiation, wind speed
[11] Yang Qi, Zhang Jianhua, Liu Zifa , Xia Shu and Li and temperature,2010.
Weiguo.:' A New Methodology for Optimizing the Size
[22] Kennedy,J, and Eberhart, R. C ."Particle swarm
of Hybrid PV/wind System'.IEEE conf 2008.
Optimization. "Proceedings of IREE International
[12] Koutroulis, E., Kolokotsa, D., Potirakis, A., Conference on Neural Networks,Piscalaway, NJ, 1995.
Kalaitzakis, K.:' Methodology for optimal sizing of stand-
[23] Dorigo, M. and Stutzle, T. (2004). Ant Colony
alone photovoltaic/wind generator systems using genetic
Optimization, Cambridge, The MIT Press.
algorithms'. Solar Energy, 2006. 80, (9),pp. 1072-1088.

411