International Journal of Scientific Research in Knowledge, 2(8), pp.

381-394, 2014
Available online at http://www.ijsrpub.com/ijsrk
ISSN: 2322-4541; 2014 IJSRPUB
http://dx.doi.org/10.12983/ijsrk-2014-p0381-0394


381
Full Length Research Paper

Optimal Design of Six-phase Permanent Magnet Synchronous Generator for direct-
drive Small-Scale Wind Turbines

Mohammad Ebrahim Moazzen
*
, Sayyed Asghar Gholamian

Babol University of Technology, Babol, Iran
*Corresponding Author: Email: M.Moazzen@stu.nit.ac.ir, M.E.Moazzen@gmail.com

Received 21 May 2014; Accepted 29 July 2014

Abstract. Permanent Magnet (PM) generators are widely employed in directly coupled wind turbines. Therefore, optimal
design of these generators is quite necessary to improve their overall performance. In this paper, single/multi-objective design
optimization of six-phase permanent magnet synchronous generator (PMSG) for directly coupled wind turbines is carried out.
The studied PM generator in this paper is the radial-flux surface mounted PM generator. The aim of optimal design is to reduce
the cost of PMSG manufacturing and generator losses. For this purpose, the main equations of PM generator design are
extracted. Therefore, the generator design steps are developed and then imperialist competitive algorithm is used to optimize
the design variables of the generator. To find the best design for this generator, the cost and losses, separately, are optimized
and are then optimized simultaneously. Finally, the optimum results are presented. Comparison between optimum results
shows that simultaneous optimization of the cost and losses leads to suitable design for PM wind generator.

Keywords: Permanent Magnet Generator, Wind power, Optimal Design, Imperialist Competitive Algorithm

1. INTRODUCTION

Today, renewable energy is rapidly replacing with the
other traditional sources of electricity generation like
coal, gas and nuclear power. The contribution of wind
energy in electricity generation, as one of the most
important types of renewable energy sources, has
grown swiftly over the past decade. At the end of the
year 2013, global installed wind power capacity is
around 318 GW (Global Wind Report, 2013). In the
last two decades, In order to maximize the energy
capture, minimize costs and improve power quality,
various wind turbine concepts have been developed
(Gundogdu and Komurgoz, 2012). According to the
construction of drive trains, there are two types of
wind turbine systems, namely, the geared drives and
the direct-drive systems (Li et al., 2009).
The some advantages of direct-drive system
compared to the geared drive are as: high overall
efficiency, high reliability and low maintenance due to
omitting the gearbox (Li et al., 2009; Grauers, 1996).
Different types of generators can be used in direct-
drive wind turbine. The generators used in the market
can be divided into electrically excited generators and
permanent magnet generators (Gundogdu and
Komurgoz, 2012). High power density, high
reliability and high efficiency are the advantages of
permanent magnet generators compared to the
electrically excited generators (Widyan, 2006). Over
the last few decades, several studies have concentrated
their research on the PM wind generator, for example:
In (Grauers, 1996), a PM synchronous generator is
compared to an induction generator, which it is shown
that the efficiency of PMSG can be more than
induction generator. In (Chen et al., 2000), outer-rotor
permanent-magnet generator for directly coupled
wind turbines has been studied. It is shown that a
generator with a simple construction can operate with
good and reliable performance over a wide range of
wind speeds. An optimal design procedure for
permanent-magnet (PM) wind generator based on
maximizing the energy yield from the generator is
introduced in (Khan and Pillay, 2005).
In (Li and Chen, 2009), forty-five direct-drive PM
wind power generator systems with the different rated
power level and different rated rotational speed have
been optimally designed for the minimum generator
system cost. Finally, the optimum results have been
compared to each other. In (Duan et al., 2010),
optimal design and finite element analysis of 2 MW
direct drive permanent magnet wind generator has
been proposed. The results indicate that to improve
the EMF and reduce the flux leakage, the best pole-
slot number and the optimal structure of pole shoe can
Moazzen and Gholamian
Optimal Design of Six-phase Permanent Magnet Synchronous Generator for direct-drive Small-Scale Wind Turbines
382
be obtained by the comparison of different models. In
(Eriksson and Bernhoff, 2010), loss evaluation and
optimal design for direct-drive wind power generator
have been proposed. Also, it is expressed that an
optimized generator for a minimum of losses will
have a high overload capability. In (Abbasian and
Hassanpour, 2011), optimal design for minimizing the
payback period of the initial investment on direct-
drive PM wind generator systems, based on genetic
algorithm method, has been proposed. In (Pinilla and
Martinez, 2012), the optimum design of permanent-
magnet direct-drive generator, considering the cost
uncertainty in raw materials, is developed with genetic
algorithm. Design and study of the behavior for a
direct drive PMSG have been proposed in (Mihai et
al., 2012) and the effect of the positioning of the
permanent magnets on the rotor surface, concerning
spatial harmonics of the air gap flux density curve is
investigated. It is shown that the choice of an
increased air gap leads to improved voltage
waveform. Potgieter and Kamper (Potgieter and
Kamper, 2012) have presented a cost-effective
technique to reduce the cogging torque and improve
the voltage quality in PM wind generators. In (Tapia
et al., 2013), a simple analytic optimization algorithm
based on a well-known mathematical approach-
Lagrange multiplier- is used to maximize the apparent
air-gap power transferred under tangential stress
constraint for PM generator. A direct-drive wind
power generator has to operate at very low speeds.
Therefore, in order to produce electricity in a
normal frequency ranges (10-70 Hz or 30-80 Hz),
generator needs to have a large number of poles (Chen
et al., 2000; Khan and Pillay, 2005). In order to use
the direct-drive generator in variable speed operation,
the direct-drive PMSG connected to the grid via back-
to-back power electronic converter, as illustrated in
Figure 1. Different topologies of PM machines are
possible to be used for direct-drive wind turbines but
the radial flux, surface mounted permanent magnet
generator, due to simple structure and reliability, can
be a good choice for this application (Li and Chen,
2009). This machine is shown in Figure 2. This paper
focuses on the optimal design of inner rotor direct
drive PMSG which uses surface magnetic poles and
radial direction magnetic circuit. To this end, the
equations needed for radial flux permanent magnet
synchronous generator design are extracted and then
the imperialist competitive algorithm is used to
optimal design which leads to minimize the cost of
PMSG Manufacturing and generator losses. The
selected permanent magnet material in this paper is
the NdFeB, which its remanent flux density is 1.23
(T) (Mohammadi and Mirsalim, 2014).
In this paper, the main equations needed for radial
flux permanent magnet (RFPM) synchronous
generator design are presented in Section 2. Generator
losses are introduced in Section 3. Wind turbine
modeling and calculating shaft speed are presented in
Section 4. In Section 5, cost estimation for PMSG
manufacturing is presented. The design procedure is
expressed in Section 6. In Section 7, Imperialist
competitive algorithm is introduced. Generator
parameters and design variables are given in Section
8. In Section 9, optimal design is carried out. Finally,
conclusions are given in Section 10.

2. MAIN EQUATIONS OF RFPM

In this section, the determination of the main
dimensions of a RFPM machines is investigated. In
this paper, parallel-sided teeth are rectangular and
stator slots are trapezoidal, as illustrated in Figure 2.

2.1. Main Dimensions

Geometrical parameters are calculated based on the
magnetic equivalent circuit of the machine (Bianchi et
al., 2006). In order to establish an analytical
relationship between the electromagnetic parameters
and geometric dimensions of the machine, the air-gap
apparent power is expressed as a function of the
induced voltage and current on the armature (stator)
winding as follows (Tapia et al., 2013; Gieras, 2010):

Where m is the generator number of phases.

is
the no-load rms voltage induced in one phase of the
stator winding (EMF).

is the stator rms current.

The air gap apparent power of the generator as a
function of main dimensions is given by (Gieras,
2010):

Where

is the fundamental winding factor.

is the is the rotational speed in rev/sec. D is the air-
gap diameter (stator inner diameter). L is the armature
stack length.

is the peak value of the first

harmonic of the air gap magnetic flux density, which
as follows (Li and Chen, 2009; Gieras, 2010):

Where

and

are the peak value of the air

gap magnetic flux density and the pole-shoe arc to
pole-pitch ratio, respectively. In (Gieras, 2010), it has
been assumed that

is equal to

), but we have used

in (2). The range of

values for

is selected between 0.6 and 0.9

(Grauers, 1996).

and

are illustrated in
Figure 3. For the sizing procedure of NdFeB
International Journal of Scientific Research in Knowledge, 2(8), pp. 381-394, 2014
383
Permanent Magnet synchronous machines can
initially be estimated as Bmg= (0.6 - 0.8) Br. Br is
the remanent flux density of the PM (Gieras, 2010). In
(3),

is the peak value of the stator line current

density, which can be expressed as (Gieras, 2010):

Where

is the number of turns per phase.

Finally, the main dimensions of the machine as a
function of the output power as follows (Gieras,
2010):

Where

is the output power of the generator.

is the no-load EMF to phase voltage ratio. In this
paper, this parameter is used to regulate the terminal
voltage.

is the output coefficients and can be

extracted from (2). We have determined the main
dimensions of the machine by selecting a suitable air-
gap diameter to armature stack length ratio (L/D). The
relative apportionment of air-gap diameter and
armature stack length in a RFPM design depends on
the machine application. The suitable range of values
for ratio L/D is typically 0.14 - 0.5 for direct-drive PM
wind generator applications (Khan and Pillay, 2005).




Moazzen and Gholamian
Optimal Design of Six-phase Permanent Magnet Synchronous Generator for direct-drive Small-Scale Wind Turbines
384
2.2. Stator and Rotor Sizing

The stator teeth, the back iron of the stator and rotor
provide the paths for flux. It is notice that to prevent
the saturation of these parts, adequate cross-sectional
area of these parts is provided. Hence, the back iron
thickness and the tooth width are expressed as a
function of the flux densities (Bianchi et al., 2006).
The back iron thickness (Stator or Rotor) can be
calculated as (Grauers, 1996; Li and Chen, 2009;
Bianchi et al., 2006):

Where

is the pole-shoe arc.

is the peak flux

density of the back iron.

is the available length of

the stator core, which can be expressed as :

.
Where

is the lamination stack factor of the stator

core. In this paper,

= 0.97 is assumed (Grauers,

1996). The tooth width can be expressed as (Grauers,
1996; Bianchi et al., 2006):

Where

and

are the slot pitch and the peak

teeth flux density, respectively. Finally, the stator slot
area, neglecting the tooth tip and referring to Figure 2,
can be calculated as (Bianchi et al., 2006):

Where is the number of slots.

and

is the
outer diameter of stator and the tooth height,
respectively. The slot shape and its dimensions are
shown in Figure 4.

Table 4: Generator Parameters


2.3. Magnet Dimensions

The magnet thickness can be expressed as (Li and
Chen, 2009):

Where

is the relative permeability of the PM

materials.

is the effective air gap, which can be

calculated as (Li and Chen, 2009):

)
Where

and is the Carter's coefficient and g is

the mechanical air-gap.

3. GENERATOR LOSSES

The generator losses are investigated in this section.
The losses considered in this paper are the stator iron
losses, the copper losses, ventilation and windage
losses and additional losses. The stator iron losses can
be divided into hysteresis losses and eddy current
losses. The rotor iron losses are neglected.




3.1. Core Losses

The stator iron losses in the separate parts, i.e., teeth
and back iron, estimated by determining the peak flux
densities in the teeth and back iron. Therefore, we can
use the equation (11) for calculate the stator core
losses, which can be expressed as (Li et al., 2009;
Grauers, 1996):

Where

is the core losses in the back iron or

teeth of the stator. f is the rated frequency of the
generator.

and

are the empirical factors for the

hysteresis and eddy current losses.

is the weight
of the core in the back iron or teeth of the stator.

and

, respectively, are the specific hysteresis and

eddy current losses (W/kg) in the core laminations at
frequency

(50 Hz) and flux density

(1.5T).

is the flux density in the back iron or tooth

of the stator. The iron used for the stator is
Surahammar CK-30 (Grauers, 1996). Values of

and

are given in Table 1.

International Journal of Scientific Research in Knowledge, 2(8), pp. 381-394, 2014
385
Table 5: Design variables and their variation boundaries


3.2. Copper Losses

The copper losses can be calculated as:

Where

is the stator winding resistance per

phase, which can be expressed as:

Where

is the copper resistivity (

at ) (Gieras, 2010).

and

are the Number

of turns per phase and the conductor cross section,
respectively.

is the average length of the one turn

per phase, which can be calculated as:

Where

is the end-winding length, which is

referring to Figure 5, can be approximated as (Bianchi
et al., 2006):

Where

, is the slot pitch measured at the half

height of the stator tooth, which can be calculated by
(16). is the coil pitch (in number of slots).

is the
increasing factor. This factor is considered due to the
end-winding overlapping. The range of values for

is 1.6 - 2 (Bianchi et al., 2006).

= 1.8 is assumed
in this paper.

3.3. Windage and Ventilation Losses

The windage and ventilation losses can be calculated
as (Pyrhnen et al., 2008):

Where

is an empirical factor, which depends on

the cooling system.

= 10 is chosen (Pyrhnen et
al., 2008).

and

are the rotor diameter, the

pole pitch and the rotational speed in rev/sec,
respectively.
Moazzen and Gholamian
Optimal Design of Six-phase Permanent Magnet Synchronous Generator for direct-drive Small-Scale Wind Turbines
386


International Journal of Scientific Research in Knowledge, 2(8), pp. 381-394, 2014
387
3.4. Additional Losses

The additional losses are very difficult to calculate
and can be commonly determined as a fraction of the
output power as follows (Pyrhnen et al., 2008):

The range of values for

is 0.0005 0.0015 for

Nonsalient-pole synchronous machine (Pyrhnen et
al., 2008). In this paper,

=0.0015 is chosen.



4. WIND TURBINE MODEL AND SHAFT
SPEED

A wind turbine is a device that converts kinetic energy
from the wind into electric energy. The output power
from the wind turbine is ideally given by (Chen et al.,
2000; Jagau et al., 2012):

Where

is the conversion efficiency with typical

values between 0.3 and 0.45 (Jagau et al., 2012).

is the air density (kg/m
3
). is the radius of the turbine
blade. u is the nominal wind speed. The rated speed of
the PMSG, largely, determines the overall size of the
generator. The nominal shaft speed of the PM wind
generator can be estimated as:

Where

is the angular speed of the turbine shaft

as well as the PM generator shaft. is the generator
efficiency, which is assumed 0.9 (Li and Chen, 2009).
Moazzen and Gholamian
Optimal Design of Six-phase Permanent Magnet Synchronous Generator for direct-drive Small-Scale Wind Turbines
388
is The tip speed ratio, which can be expressed as
(Eriksson and Bernhoff, 2010; Jagau et al., 2012):

The range of values for is reported 57 and 68
(Li and Chen, 2009; Jagau et al., 2012). In this paper,
=6 is chosen. The wind speed is an important
parameter for calculating the nominal shaft speed.
According to the studies and researches previously
done by Iranian Renewable Energy Organization
(SUNA), the range of wind speed in the Nehbandan -
Zabol road of Iran, is 5-10.2 m/s (Iranian Renewable
Energy Organization, 2010). Therefore, our generator
has been designed to operate at the wind speed of 10
m/s. According to nominal wind speed, the shaft
speed is estimated 250 rpm. Therefore, the number of
poles to produce electricity in 50 Hz frequency and
nominal shaft speed is chosen 24.

5. COST ESTIMATION FOR PMSG
MANUFACTURING

In this section, the estimation of the cost of PMSG
manufacturing is described. This cost can be
expressed as (Gieras, 2010):

Where

is the coefficient depending on the

number of manufactured generators per annum. In this
paper

=1 is assumed.

and

are the
cost of winding, the cost of core, the cost of PMs and
the cost of shaft, respectively, which can be expressed
as (Gieras, 2010):

Where

is the coefficient of the cost of

fabrication of coils (insulation, assembly,
impregnation, etc).

is the coefficient taking into

account the cost of the rotor winding.

is the
coefficient taking into account the cost of frame, end
bells and bearings.

is the coefficient of utilization

of electrotechnical steel.

is the stacking factor.

is the coefficient including the cost of stamping,
stacking and other operations.

is the coefficient
accounting for the increase of the cost of PMs due to
complexity of their shape.

is the coefficient
including the cost of magnetization of PMs.

is
the total volume of the steel bar to the volume of the
shaft.

is the coefficient taking into account the cost

of machining of the shaft.

and

are
the weight of the copper, the core, PMs and the shaft,
respectively.

and

, respectively, are the

specific cost (U.S. dollars per kilogram) of the copper,
the core, PMs and the shaft steel. The different
coefficients of k and the different specific cost are
given in Table 2.

6. THE DESIGN PROCEDURE

The design procedure and flowchart is shown in
Figure 6. The main steps of generator design are
presented as:
(1) Initialize design values (

)
(2) Dedicate value to each design variables
(

)
(3) Calculate magnet thickness
(4) Computation of main dimensions (L and D) and
number of turns per phase (

)
(5) Calculate the geometrical dimensions of stator
(slot dimensions) and rotor
(6) Calculate the air-gap flux density and compare
to

. If continue condition is satisfied continue, if

not go to 2.
(7) Calculate the cost of PMSG manufacturing and
generator losses
(8) Calculate the voltage induced (

) and compare
to

. If stop condition is satisfied stop, if not go to

2.

International Journal of Scientific Research in Knowledge, 2(8), pp. 381-394, 2014
389

Fig. 6: Design flowchart of the radial flux, surface mounted PM generator


Start
Is there an empire
with no colonies
Eliminate this empire
Stop condition
satisfied
END
Assimilate colonies
Exchange the positions of that
imperialist and the colony
Is there a colony in an
empire which has lower cost
than that of the imperialist
Compute the total cost of all empires
Imperialistic Competition
Yes
Yes
Yes
No
No
No
Initialize the empires
Revolve some colonies
Unite Similar Empires

Fig. 7: Flowchart of imperialist competitive algorithm

7. IMPERIALIST OPTIMIZATION METHOD

In this paper, the optimization of permanent magnet
synchronous generator has been performed by
imperialist competitive algorithm (ICA). Optimization
of imperialist competitive algorithm has inspired from
a political-social process. Comparing to other
optimization methods, this method is much faster and
has high abilities. Figure 7 shows the proposed
algorithm stages. At first, we create a 1N
var
array of
adjustable input parameters for generators which must
be optimized and then we called it a country. To start
the algorithm, N
country
of the initial countries is created.
N
imp
of the best members of this population (countries
with the lowest value of objective function) are
selected as imperialist. The rest of the N
col
countries,
make the colonial empires, each of them belongs to an
empire. The total power of each empire is modeled as
the total power of the imperialist countries, plus a
percentage of its colonies power average.
Moazzen and Gholamian
Optimal Design of Six-phase Permanent Magnet Synchronous Generator for direct-drive Small-Scale Wind Turbines
390
With the initial state of all the empires, colonial
competitive algorithm begins. Having the cost of all
imperialists, the normalized cost and as a result,
normalized relative power of each imperialist is
calculated and based on the above calculation, the
colonized countries are divided between these
imperialist. In the assimilation (absorption) policy of
the colonization process, the colony continues to
move in the direction approaching the imperialist,
with the amount of x unit and deviation. During the
move, some of the colonies might reach a better
position than the imperialist, which in this case the
colonies and imperialists change the roles and the
algorithm with the imperialist continues in a new
situations. The final stage of the imperialist
competition is the time when we have a single empire
with colonies very close to it. At this time, the
imperialist optimization method by reaching the
optimal peak of the objective function and considering
the stop criterion, is stopped (Atashpaz Gargari,
2008). The values of these parameters are adjusted to
achieve the desired goal in this paper by using
imperialist optimization method, in Table 3 is
obtained.


Fig. 8: The generator losses value at each iteration for the first 70 iterations

8. GENERATOR PARAMETERS AND DESIGN
VARIABLES

The optimal design in this paper is based on a 3500 w
direct-drive PM wind generator. The machine
parameters used in the optimal design are expressed in
Table 4.

and

are chosen 1 mm.

=1.5 mm is
chosen. Design variables and their variation
boundaries are given in Table 5.

International Journal of Scientific Research in Knowledge, 2(8), pp. 381-394, 2014
391

Fig. 9: The generator cost value at each iteration for the first 70 iterations


Fig. 10: The value of F
3
at each iteration for the first 70 iterations

9. OPTIMIZATION OF PM WIND
GENERATOR DESIGN AND RESULTS

In this paper, the optimal design in order to
minimizing the cost of PMSG manufacturing and
generator losses has been done. To find the best
design for PM wind generator, Firstly, the generator
losses are optimized. Figure 8 depicts the generator
losses minimization for the first 70 iterations. The first
objective function as:

F
1
=

(28)
Moazzen and Gholamian
Optimal Design of Six-phase Permanent Magnet Synchronous Generator for direct-drive Small-Scale Wind Turbines
392

The second objective function is to optimize the cost
of PM wind generator, which can be expressed as:

F
2
= C
t
(29)

Figure 9 plots the cost minimization for the first 70
iterations. The third objective function is to optimize
the cost of PM wind generator and generator losses,
simultaneously. This objective function can be
expressed as:

F
3
=

(30)
Where and are minimum values of the PM
wind generator cost and generator losses, respectively.
Figure 10 plots the F
3
minimization for the first 70
iterations. The dimensions of the optimal generator
and optimal values of design variables are listed in
Table 6.

Table 6: Optimal generator parameters


10. CONCLUSION

In this paper, by employing imperialist competitive
algorithm, single/multi-objective design optimizations
of 3.5 kw six-phase direct-drive permanent magnet
wind generator by choosing manufacturing cost and
generator losses and their combination as objective
function were performed. For this purpose, the main
equations of generator design are extracted and then
the cost and losses of generator are investigated. To
find the best design for this generator, firstly,
generator losses and then the cost of generator are
optimized. Finally, the cost and losses of generator
referring to (30), simultaneously, are optimized.
Referring to Table 6, the third objective function can
lead to the best design for the generator, because of
two reasons: First, compared to optimizing the losses,
the losses of generator only 14.51% is increased.
Second, compared to optimizing the cost, the cost of
generator only 5.64% is increased.



International Journal of Scientific Research in Knowledge, 2(8), pp. 381-394, 2014
393
REFERENCES

Abbasian M, Hassanpour A (2011). Optimal Design
of a Direct-Drive Permanent Magnet
Synchronous Generator for Small-Scale Wind
Energy Conversion Systems. Journal of
Magnetics, 16(4): 379-385.
Atashpaz Gargari E (2008). Development of social
optimization algorithm and evaluate its
performance. M.Sc. dissertation, Tehran
University, Tehran, Iran.
Bianchi N, Bolognani S, Frare P (2006). Design
Criteria for High-Efficiency SPM Synchronous
Motors. IEEE Transactions on Energy
Conversion, 21(2): 396-404.
Chen J, Nayar Ch, and Xu L (2000). Design and
Finite-Element Analysis of an Outer-Rotor
Permanent-Magnet Generator for Directly
Coupled Wind Turbines. IEEE Transactions on
Magnetics, 36(5): 3802-3809.
Duan G, Wang H, Guo H, Gu G (2010). Direct Drive
Permanent Magnet Wind Generator Design and
Electromagnetic Field Finite Element Analysis.
IEEE Transactions on Applied
Superconductivity, 20(3): 1883-1887.
Eriksson S, Bernhoff B (2010). Loss evaluation and
design optimisation for direct driven permanent
magnet synchronous generators for wind
power. Journal of Applied Energy, 88: 265
271.
Gieras J. F (2010). Permanent magnet motor
technology: design and applications, 3th ed.
CRC Press, ISBN 978-1-4200-6440-7, USA.
Global Wind Report (2013). Annual Market Update
2013. Global Wind Energy Council, Brussels,
Belgium.
Grauers A (1996). Design of direct-driven permanent-
magnet generators for wind turbines. Ph.D.
dissertation, Chalmers Univ. Technol.,
Goteborg, Sweden.
Grauers A (1996). Efficiency of three wind energy
generator systems. IEEE Transactions on
Energy Conversion, 11(3): 650-657.
Gundogdu T, Komurgoz G (2012). Technological and
economical analysis of salient pole and
permanent magnet synchronous machines
designed for wind turbines. Journal of
Magnetism and Magnetic Materials, 324: 2679
2686.
Iranian Renewable Energy Organization (SUNA)
(2010). Tehran, Iran.
Jagau H, Khan M, Barendse P (2012). Design of a
Sustainable Wind Generator System Using
Redundant Material. IEEE Transactions on
Industry Applications, 23(6): 1827-1837.
Khan A, Pillay P (2005). Design of a PM Wind,
Optimised for Energy Capture over a Wide
Operating Range. IEEE Conference on
Electrical Machines and Drives, Spain, 1501-
1506.
Li H, Chen Z (2009). Design optimization and site
matching of direct-drive permanent magnet
wind power generator systems. Journal of
Renewable Energy, 34: 11751184.
Li H, Chen Z, Polinder H (2009). Optimization of
Multibrid permanent-magnet wind generator
systems. IEEE Transactions on Energy
Conversion, 24(1): 8292.
Mihai A, Benelghali S, Simion Al, Outbib R,
Livadaru. L (2012). Design and FEM analysis
of five-phase Permanent Magnet Generators for
Gearless Small-Scale Wind Turbines," IEEE
Conference on Electrical Machines,
Marseille,150-156.
Mohammadi S, Mirsalim M (2014). Design
optimization of double-sided permanent-magnet
radial-flux eddy-current couplers. Journal of
Electric power systems research, 108: 282292.
Pinilla M, Martinez S (2012). Optimal design of
permanent-magnet direct-drive generator for
wind energy considering the cost uncertainty in
raw materials. Journal of Renewable Energy,
41: 267276.
Potgieter J, Kamper M (2012). Torque and voltage
quality in design optimization of low-cost non-
overlap single layer winding permanent magnet
wind generator. IEEE Transactions on Industry
Application, 59(5): 2147-2156.
Pyrhnen J, Jokinen T, Hrabovcov V (2008). Design
of Rotating Electrical Machines, 1th ed. John
Wiley & Sons, Ltd, ISBN 978-0-470-69516-6,
Wiltshire.
Tapia J, Pyrhnen J, Puranen J, Lindh P (2013).
Optimal Design of Large Permanent Magnet
Synchronous Generators. IEEE Transactions on
Magnetics, 49(1): 642-650.
Widyan M (2006). Design, optimization, construction
and test of rare-earth permanent-magnet
electrical machines with new topology for wind
energy applications. Ph.D. dissertation, Berlin
Univ. Technol., Berlin, Germany.

Moazzen and Gholamian
Optimal Design of Six-phase Permanent Magnet Synchronous Generator for direct-drive Small-Scale Wind Turbines
394




Mohammad Ebrahim Moazzen was born in Mazandaran, Iran, in 1989. He is
currently working towards the M.Sc. degree in Babol University of Technology, Babol,
Iran.








Sayyed Asghar Gholamian was born in Mazandaran, Iran. He received the PhD degree in
electrical engineering from K.N. Toosi University of Technology, Tehran, Iran in 2008. He is
currently an assistant professor in the department of Electrical Engineering at the Babol
University of Technology, Babol, Iran. His research interests include design, simulation,
modeling and control of electrical machines.