1

Abstract-- In this paper, the problem of efficiency optimization

in vector controlled interior permanent magnet synchronous mo-
tor drives is investigated. A loss model controller is introduced
that determines the optimal d-axis component of the stator cur-
rent that minimizes power losses. For the implementation of the
suggested controller, the knowledge of the loss model is not re-
quired, since an experimental procedure is followed to determine
its parameters. Furthermore, it is shown that the loss model of the
interior permanent magnet motor can be used as a basis for deriv-
ing loss minimization conditions for surface permanent magnet
synchronous motors and synchronous reluctance motors as well.
Experimental results of an interior permanent magnet motor are
presented to validate the effectiveness of the proposed method and
demonstrate the operational improvements.

Index Terms-- Permanent magnet motors, losses, optimization
methods, optimal control, variable speed drives.
NOMENCLATURE
R
s
Stator resistance.
L
md
, L
mq
d- and q-axis magnetizing inductances.
L
ls
Stator leakage inductance.

e
Supply frequency.

r
Motor speed.
V
s
Stator voltage.
V
m
Air-gap voltage.

m
Air-gap magnetic flux.
I
s
Stator current.
I
d
, I
q
d- and q-axis components of stator current.
I
m
Magnetizing current.
I
md
, I
mq
d-and q-axis components of magnetizing current.
f
I Equivalent excitation current of the permanent
magnet.
T
e
Electromagnetic torque.
P
l
Total power losses.
P
Cu
Copper losses.
P
Fe
Iron losses.
P
str
Stray losses.
P
m
Mechanical losses.
c
Fe
Iron loss coefficient.
c
str
Stray loss coefficient.
c
m
Mechanical loss coefficient.

C. Mademlis, I. Kioskeridis and N. Margaris are with the Department of
Electrical and Computer Engineering, Aristotle University of Thessaloniki,
Thessaloniki GR-54 124, Greece (e-mail: mademlis@eng.auth.gr).
I. INTRODUCTION
ERMANENT Magnet (PM) synchronous motor adjustable
speed drives offer significant advantages over induction
motor drives in a wide variety of industrial applications (i.e.
high power density, high efficiency, improved dynamic per-
formance and reliability) [1]. Since vector control in PM syn-
chronous motors provides fast dynamic response with a less
complex and non-parameter dependent controller, PM motor
drives can be an attractive alternative choice [2].
Improvement of PM motor efficiency is a most important
priority, especially in cases where drives are powered by a
battery source. Therefore, significant efforts are taken to im-
prove their efficiency. Since there are a great variety of PM
motor configurations, the efforts are mainly focused on the
search for the optimum rotor structure [3]-[7]. However, effi-
ciency can also be improved by intervening in the motor op-
eration principle with automatic control techniques.
Several control methods have been proposed in order to re-
duce the loss of PM motor drives and improve their perform-
ance. The copper loss can be minimized by the maximum
torque-per-ampere current control [8]. In surface PM motor
drives, maximum torque-per-ampere current ratio is attained
by keeping the d-axis component of the stator current equal to
zero (I
d
= 0) [9], [10]. Since the I
d
=0 control prevents the
demagnetization of the PM, it is often employed in interior PM
motor drives. However, the I
d
current, that provides maximum
torque-per-ampere current ratio in interior PM motor drives, is
a function of the I
q
current and opposes the excitation field of
the PM [8]-[11].
Several attempts to minimize both copper and iron losses
have been recently presented [12]-[14]. However, the pro-
posed loss minimization conditions are complex and can only
be implemented using off-line made look-up tables. Therefore,
a number of costly and time-consuming measurements are re-
quired. A control method, described in [15], improves effi-
ciency of PM motors and is implemented using a voltage
source inverter. The efficiency improvement is attained with
appropriate control of stator voltage in order to keep power
factor equal to unity. However, although the real to apparent
power ratio (kW/kVA) of the PM motor is maximized, power
losses are not minimized. An adaptive search controller for
interior PM motors was developed in [16]. Finally, an ap-
proach that specifies the optimal d-axis current for minimizing
interior PM motor losses was presented in [17].
Optimal Efficiency Control Strategy for Interior
Permanent Magnet Synchronous Motor Drives
Christos Mademlis, Member, IEEE, Iordanis Kioskeridis, and Nikos Margaris, Member, IEEE

P
2
This paper presents an optimal efficiency method for vector
controlled interior PM synchronous motor drives. The loss
minimization is accomplished with a loss model controller that
is based on the optimal d-axis current condition, as presented
in [17]. For the implementation of the proposed controller, an
experimental procedure is followed to adjust its parameters
without requiring the knowledge of the exact motor model.
Additionally, a further investigation of the loss model of the
interior PM motor is conducted and it is shown that this model
can be used as a basis for deriving loss minimization expres-
sions for surface PM synchronous motors and synchronous
reluctance motors (SynRMs).
The contents of the paper are organized as follows. In Sec-
tion II the basic equations of an interior PM synchronous mo-
tor drive are given and the motor loss model is presented. The
loss minimization condition is derived in Section III. The im-
plementation of the optimal interior PM motor drive with
search controller (SC) and loss model controller (LMC) is
described in Section IV. The experimental results are pre-
sented in Section V. In Section VI, comparisons between the
performance of the optimum control, conventional I
d
=0 con-
trol and maximum torque-per-ampere current control are pre-
sented. Loss minimization conditions for surface PM synchro-
nous motors and SynRM, derived from the interior PM loss
model, are presented in Section VII. Finally, conclusions are
drawn in Section VIII.
II. BASIC EQUATIONS LOSS MODEL
Fig. 1 shows the d- and q-axis equivalent circuits of the in-
terior PM synchronous motor in the synchronously rotating
reference frame [18]. The equivalent circuits are given in the
per-unit system and the effects of iron and stray losses are ig-
nored. The phasor diagram in the synchronously rotating d-q
reference frame is illustrated in Fig. 2. In the figure, the I
d
cur-
rent is negative (demagnetizing current) and results in field
weakening. The d- and q-axis components of the magnetizing
current are given respectively by


md f d
= + I I I (1)
and

mq q
= I I (2)

Since the magnetic permeability of the PM is close to air,
the interior PM synchronous motor presents inverse saliency.
Consequently, the q-axis inductance of the interior PM motor
exceeds the d-axis inductance [9]


mq md
L L > (3)

The electromagnetic torque of the motor is given by [9]

( 1)
e md f q md d q
T L I I L I I = (4)

where is the saliency ratio

1
mq
md
L
L
= > (5)


Fig. 1. Per-unit equivalent circuits of interior PM synchronous motor.

The main losses of the permanent magnet synchronous mo-
tor are the following:
Copper Losses: These are due to current flow through the
stator windings and are given by


( )
2 2 2
Cu s s s d q
P r I r I I = = + (6)

Iron Losses: These are due to hysteresis and eddy currents
and are given by the following empirical formula [9], [19]


2 2 2 2 2
( )
Fe Fe e m Fe e md md mq
P c c L I I

= = + (7)

where =1.5 1.6. Substituting (1) and (2) in (7), iron losses
are given by


2 2 2 2
( )
Fe Fe e md f d q
P c L I I I

( = + +

(8)

Stray Losses: These arise on the copper and iron of the mo-
tor and are given by [20]

( )
2 2 2 2 2
str str e s str e d q
P c I c I I = = + (9)

Mechanical Losses: These are due to friction and windage
losses and are proportional to the square of rotor speed [18]


2
m m e
P c = (10)



Fig. 2. Phasor diagram of interior PM synchronous motor.
3



Fig. 3. Power loss versus Id current in a 3.4-kW interior PM motor drive
(experimental results).

Harmonic Losses: These are additional losses due to non-
sinusoidal stator voltage supplied to the synchronous motor.
The harmonic voltages increase the iron losses and the har-
monic currents increase the stator copper losses [1].

The mechanical losses are independent from the electrical
variables; hence they are not controlled by flux weakening.
Additionally, harmonic losses are not directly controlled by
flux weakening. However, these losses are indirectly con-
trolled by the decrease of harmonic voltages, because of flux
weakening.
The losses that can be minimized by flux weakening, ex-
pressed in d-q axis components, are


2 2 2 2 2
( ) b ( )
l Cu Fe str d q f d q
P P P P a I I I I I ( = + + = + + + +

(11)

where

2
s str e
a r c = + (12)
and

2
Fe e md
b c L

= (13)

III. LOSS MINIMIZATION CONDITION
The loss minimization condition at steady state (T
e
and
e
constant) with respect to I
d
is given by


,
0
e e
l
d
T
P
I

= (14)

Using (11), condition (14) is satisfied when


2
( ) ( ) 0
q
d f q
d
I
I a b bI a b I
I

+ + + + = (15)

Since the electromagnetic torque is kept constant, it is deduced
that
0
e
e
d
T
I

= (16)

From (4) and (16), we obtain


( 1)
( 1)
q q
d f d
I I
I I I

=

(17)

Substituting (17) in (15) yields


2 2 2
2
( 1)( )
2
( 1) 0
f q
d d f
bI I a b
a b b
I I I
a b a b

+ +
+
=
+ +

(18)
The solution of (18) is as follows:


1,2
2
2
2
(2 )
d d d q
a b a b a b
I G G I
a b a b a b
+ + + (
= +
(
+ + +

(19)
where

2( 1)
f
d
I
G

(20)

Substituting (19) in (4) and after some algebraic operations,
the electromagnetic torque for the two solutions of (18) is re-
spectively given by


2( )
e md f q
a b
T X I I
a b
+
=
+


2
2 2
2 2
( )( )( 1)
1 1 4
( )
q
f
I
a b a b
I a b

(
+ +
( +
+ (

(21)

From (21) it is concluded that, in motoring operation (I
q
>0)
the solution
1
d
I results in negative electromagnetic torque and
in braking operation (I
q
<0) in positive electromagnetic torque.
Therefore, solution
1
d
I is rejected. On the contrary, the second
solution
2
d
I is acceptable. Moreover, the second derivative of
P
l
with respect to I
d



2 2 2
2
2 2
,
( )( 1)
2( ) 6
( 1)
e e
l
q
d
T
f d
P a b
a b I
I
I I

+
= + +
(

(22)

is always positive. Therefore, the loss minimization condition
for the interior PM motor is given by


2
2
2
(2 )
opt
d d d q
a b a b a b
I G G I
a b a b a b
+ + + (
= +
(
+ + +

(23)

The existence of the loss minimum is experimentally veri-
fied. Fig. 3 shows the variation of power loss (electrical input
power minus mechanical output power) versus d-axis current,
measured in a 3.4-kW interior PM synchronous motor. The
parameters of the motor are given in Table I. They were meas-
ured based on the approach presented in [21] and using the
LMC parameter values that are experimentally obtained as
described in Section V. For the points noted by asterisk, the
loss minimization is achieved and these operating points sat-
isfy the optimal condition (23).
Note that the loss minimum curves are smooth and flat
around the minimum.
4


Fig. 4. Variation of the optimal Id current versus speed for various load
torques (simulation results).



Fig. 5. Optimal (Iq, Id) trajectories for various speeds and load torques (simu-
lation results).



Fig. 4 illustrates the optimal I
d
current versus speed for
various load torques and Fig. 5 illustrates a family of optimal
( ,
q d
I I ) trajectories for various speeds and load torques. In
both figures, the results are obtained by simulations. The inter-
section of each ( ,
q d
I I ) trajectory with a load torque curve
corresponds to an operating point (noted by asterisk) for that
particular load torque, in which loss minimization is accom-
plished.
IV. IMPLEMENTATION OF THE LOSS MINIMIZATION CONDITION
The block diagram of the optimal interior PM motor drive
is shown in Fig. 6. The loss minimization can be accomplished
with a search controller (SC) or a loss model controller
(LMC). The SC measures the input power to the drive and


Fig. 6. Optimal vector controlled interior PM synchronous motor drive.




Fig. 7. Performance of the optimal PM motor drive with SC (experimental
results).

adjusts the I
d
current searching for the minimum input power.
The LMC measures the speed and the I
q
current and specifies
the optimal I
d
current through the loss minimization condition
(23).
Although the use of SC could be the obvious way for loss
minimization, experiments prove that the drive performance is
not satisfactory. As shown in Fig. 7, the drive does not reach a
steady state, causing undesirable torque disturbances. Also
during transient operation, the SC should remain disabled and
is setting I
d
=0 [16]. Generally the SC approach has several
disadvantages and such performance is expected from the rele-
vant literature [14], [22]. On the contrary, the LMC offers
superior performance; hence it is preferable for the implemen-
tation of the optimal controller.
Substituting (12) and (13) in (23), the equation of the LMC
that specifies the optimal I
d
current, is given by


2
2 2 2 2 2 2
2 2 1 3 4
2 2 2 2 2 2
2 2 2
1 1 1
1 1 1
opt
e e e
d d d q
e e e
T T T
I G G I
T T T


( + + +
= +
(
+ + +

(24)
where

2 2 2
1
(2 )
str Fe
e md
s s
c c
T L
r r

= + (25)

5

(a)


(b)

Fig. 8. Performance of the optimal PM motor drive with LMC: (a) Id current
and input power and (b) Iq current and stator voltage (experimental results).


2 2 2
2
str Fe
e md
s s
c c
T L
r r

= + (26)

2 2 2
3
str Fe
e md
s s
c c
T L
r r

= + (27)
and

2 2 2 2
4
str Fe
e md
s s
c c
T L
r r

= + (28)

From (25)-(27), it is concluded that the following relation
holds

2 2 2
1 2 3
2 T T T = (29)

In order to calculate the optimal I
d
current from (24), the
motor speed
e
and the I
q
current are required. Since loss
minimization is accomplished at steady state, the command
signals of the speed and I
q
current can be used. The LMC pa-
rameters can be adjusted experimentally as follows.
1) A three-phase wattmeter is used for measuring the total
input power of the drive.
2) The motor rotates at low speed, about 15%-25% of its
nominal value. Under this condition (
e
<< 1), (24) be-
comes

2 2
opt
d d d q
I G G I + (30)

and the gain G
d
is adjusted so that the minimum wattmeter
indication is obtained.
3) Speed is increased up to 50%-70% of its nominal value and
the motor rotates with no load. In this case, the I
q
current
could be taken as approximately equal to zero and using
(29), condition (24) is reduced to


2 2 2
2 3
2 2
2
( )
2
1
opt
e
d d
e
T T
I G
T

=
+
(31)

The I
d
current is adjusted so that the minimum wattmeter
indication is accomplished. Two measurements for two dif-
ferent speed values give an equation system with unknown
parameters T
2
and T
3
. The solution of the equation system
gives the values of T
2
and T
3
. Parameter T
1
is then calcu-
lated by (29).
4) Speed is maintained equal to the value of the previous step.
Load torque is increased up to 40%-55% of its nominal
value. Parameter T
4
is adjusted so that the minimum watt-
meter indication is accomplished.
5) Steps 2)-4) are repeated until the desired accuracy is ob-
tained.
The q-axis magnetizing inductance L
mq
varies due to satura-
tion [9]. Since the gain G
d
and parameters T
1
, T
3
and T
4
de-
pend on L
mq
, their values are affected by saturation. Therefore,
G
d
is an increasing function of the load torque and Step 2)
must be repeated for medium- and high load torques. Addi-
tionally, the variation of parameters T
1
, T
3
and T
4
are partially
compensated by the increase of G
d
.
The armature resistance r
s
and excitation current
f
I vary
depending on the temperature [1]. The gain G
d
decreases, as
load torque increases, and consequently temperature increases.
On the contrary, due to saturation, G
d
increases as load torque
increases and therefore the temperature variation of G
d
is par-
tially compensated. Additionally, parameters T
1
to T
4
are af-
fected by temperature variations. However, T
1
to T
4
are varied
by temperature in both numerator and denominator of each
fraction of (24), and consequently the variation of I
d
is narrow.
Thus, parameters variation is considered by means of G
d
de-
pendency on I
q
current. Furthermore, successful approximation
of the minimum is possible since loss curves, as shown in Fig.
3, are smooth and flat around the minimum. The above is con-
sistent to the approach presented in [12].
The conclusion is that, in practice, knowledge of the loss
model is not required. Furthermore, no additional feedback
signals from the motor are required, beyond those already used
in the pre-existing control, i.e. I
q
and
e
control signals. There-
fore, the LMC controller does not affect the cost and the com-
plexity of the drive. Moreover, due to experimental adjustment
of the LMC parameters, the LMC minimizes not only the PM
motor losses, but also the whole drive losses.
V. EXPERIMENTAL RESULTS
For the experimental verification of the theoretical results
and the effectiveness of the LMC operation, a 3.4-kW interior
PM synchronous motor was used. The LMC parameters that
have been adjusted experimentally, according to the rules de-
scribed in Section IV, are given in Table II.
Fig. 8 shows the performance of the suggested LMC. Com-
6

(a)


(b)

Fig. 9. LMC response to load torque demand: (a) Id current and speed and
(b) Iq current and power input (experimental results).



paring Fig. 7 to Fig. 8, it is concluded that the optimal control
scheme performs better with an LMC than with an SC. In Fig.
8, the I
d
command of the LMC decreases at a low rate in order
to avoid strong armature current and torque disturbances.
Fig. 9 illustrates the LMC performance to an abrupt torque
demand. It can be seen that at any torque disturbance, the
LMC reacts almost immediately and after equilibrium is estab-
lished, the LMC reaches its new optimal I
d
current value.
VI. COMPARISONS BETWEEN THE OPTIMUM CONTROL AND
OTHER CONTROL METHODS
The I
d
current for producing maximum torque-per-ampere
current is given by [10], [11]


( / )
max
2
2
2( 1) 2( 1)
T I
e s
f f
d q
I I
I I


| |
= +
|

\ .
(32)

Substituting (20) in (32) yields


(a)


(b)

Fig. 10. Comparisons of the LMC against Id = 0 control and maximum
torque-per-ampere current control: (a) power saving and (b) efficiency im-
provement (experimental results).


( / )
max
2 2
T I
e s
d d d q
I G G I = + (33)

Fig. 10(a) provides the power saving of the suggested LMC
method, against the conventional I
d
=0 control and the
maximum torque-per-ampere current control methods, for
various speeds and load torques. Fig. 10(b) provides the ratio
of the optimal efficiency of the suggested LMC method, to the
conventional I
d
=0 control and the maximum torque-per-
ampere current control methods, for various speeds and load
torques.
Finally, the performances of the LMC and the maximum
torque-per-ampere current controller are compared in Fig. 11.
Power losses are reduced by maximum torque-per-ampere
current control, however are minimized by LMC.
VII. LOSS MINIMIZATION CONDITIONS FOR ALL SYNCHRO-
NOUS MOTOR TYPES
It is well known that, the model of an interior PM motor
can be regarded as a general type motor model. The equivalent
circuits and the equations that describe the behavior of a sur-
face PM motor or a SynRM are derived from that of an interior
PM motor if equal inductances in d- and q-axis are obtained
(
mq md
L L = and therefore =1) or rotor excitation is eliminated
7

(a)


(b)

Fig. 11. Comparison of the LMC and maximum torque-per-ampere current
controller performances in a 3.4-kW interior PM motor drive: (a) Id current
and input power and (b) Iq current and stator current (experimental results).

( 0
f
I = ), respectively. Therefore, based on the analysis for
efficiency optimization of the interior PM motor, as presented
in Section III, loss minimization expressions for the surface
PM motor and the SynRM can be derived.
Specifically, condition (18) as rewritten


2 2 2
2
( 1)( )
2
( 1) 0
f q
d d f
bI I a b
a b b
I I I
a b a b

+ +
+
=
+ +

(34)

is modified to obtain the loss minimization expression for each
specific motor.
A. Optimal efficiency locus of Interior PM motor
From (34), it is concluded that the optimal efficiency locus
of an interior PM motor is a hyperbola, as described by


2 2
2
2 2
2
d d q d
a b b a b a b
I G I G
a b a b a b
+ + + | | | |
=
| |
+ + +
\ . \ .
(35)

with center at


(2 )
and 0
d d q
a b
I G I
a b
+ | |
= =
|
+
\ .
(36)

and the slopes of the asymptotes are



(a)



(b)

Fig. 12. Optimal efficiency locus of: (a) interior and surface PM synchronous
motors and (b) synchronous reluctance motors.


2
a b
a b
+

+
(37)

For a given speed value, the intersection of the hyperbola
with a motoring load torque curve corresponds to an operating
point (point A, Fig. 12a), in which interior PM motor optimal
efficiency is attained for that particular load torque. Note that,
the hyperbola that corresponds to points with positive I
d
cur-
rent does not intersect the load torque curve. These are points
that correspond to the first solution
1
d
I of (19) and, as ex-
plained in Section III, solution
1
d
I is rejected.
B. Optimal efficiency locus of Surface PM motor
Substituting saliency ratio equal to unity (=1) in (34), the
loss minimization condition for a surface PM motor is derived
[11], [12], [22]


opt
d f
b
I I
a b
=
+
(38)

8
The optimal efficiency locus of a surface PM motor is a
straight line (Fig. 12a). This can also be derived from the op-
timal efficiency locus of the interior PM motor. As saliency
ratio tends to unity, the slope of the asymptotes increases and
the center of the hyperbola moves to infinity. Thus, for =1
the left part of hyperbola that corresponds to accepted operat-
ing points is reduced to a straight line, as specified by (38).
In surface PM motor, the electromagnetic torque (4) be-
comes [9]



e md f q
T L I I = (39)

For a given speed value, the intersection of the line (38) with a
motoring load torque curve corresponds to an operating point,
in which surface PM motor optimal efficiency is attained for
that particular load torque (point B, Fig. 12a).
C. Optimal efficiency locus of SynRM
Eliminating the rotor excitation ( 0
f
I = ) in (34), the solu-
tion that gives the optimal d-axis current for a SynRM is


1,2
2
d q
a b
I I
a b
+
=
+
(40)

The solutions (40) correspond to straight lines that are parallel
to the asymptotes of the hyperbola of the interior PM optimal
efficiency locus.
In SynRM, the electromagnetic torque (4) becomes [9]

(1 )
e md d q
T L I I = (41)

Since
md mq
L L > , the saliency ratio is less than unity ( <1).
The solution
2
d
I is rejected since in motoring operation (I
q
>0)
results in negative electromagnetic torque and in braking op-
eration in positive electromagnetic torque. Thus, the optimal d-
axis current for a SynRM is given by the first solution [11],
[23]

2
opt
d q
a b
I I
a b
+
=
+
(42)

As for the PM motors, at a given speed value the intersec-
tion of the locus of (42) with a motoring load torque curve
corresponds to an operating point (point C, Fig. 12b), in which
SynRM optimal efficiency is attained for that particular load
torque.
VIII. CONCLUSIONS
This paper has described a method for minimizing the
losses in vector controlled interior PM synchronous motor
drives. An LMC for determining the optimal I
d
current was
presented. The suggested controller uses the command signals
of the speed and I
q
current, and for its implementation the
knowledge of the loss model is not required. The LMC pa-
rameters were determined by following a simple experimental
procedure. Additionally, the controller does not affect signifi-
cantly the cost, complexity and dynamic performance of the
drive. The performance of the loss model controller was com-
pared against the conventional I
d
= 0 control and the maxi-
mum torque-per-ampere current control. Finally, based on the
interior PM motor loss model, the loss minimization condi-
tions for surface PM synchronous motors and synchronous
reluctance motors were also derived.
REFERENCES
[1] J.K. Gieras ans M. Wing, Permanent Magnet Motor Technology: De-
sign and Applications, New York: Marcel Dekker, 1997.
[2] B.K. Bose, Power Electronics and AC Drives, Englewood Cliffs, NJ:
Prentice-Hall, 1986.
[3] P. Viarouge, M. Lajoie-Mazenc, and C. Andrieux, Design and Con-
struction of a Brushless Permanent Magnet Servomotor for Direct-Drive
Application, IEEE Trans. on Industry Applications, vol. IA-23, no. 3,
pp. 526-531, May/June 1987.
[4] G.R. Slemon, On the Design of High-Performance Surface-Mounted
PM Motors, IEEE Trans. on Industry Applications, vol. 30, no. 1, pp.
134-140, Jan./Feb. 1994.
[5] L. Xu, L. Ye, L.Zhen, and A.El-Antably, A New Design Concept of
Permanent Magnet for Flux Weakening Operation, IEEE Trans. on In-
dustry Applications, vol. 31, no. 2, pp. 373-378, March/April 1995.
[6] J. D.La Ree and N. Boules, Magnet Shaping to Reduce Induced Volt-
age Harmonics in PM Machines with Surface Magnets, IEEE Trans.
on Energy Conversion, vol. 6, no. 1, p.p. 155-161, March 1991.
[7] W.L. Soong, D.A. Staton, T.J. Miller, Design of a New Axially-
Laminated Interior Permanent Magnet Motor, IEEE Trans. on Industry
Applications, vol. 31, no. 2, pp. 358-367, March/April 1995.
[8] T.M. Jahns, G.B. Kliman, and T.W. Neumann, Interior Permanent-
Magnet Synchronous Motors for Adjustable-Speed Drives, IEEE
Trans. on Industry Applications, vol. IA-22, no. 4, pp. 738-747,
July/Aug. 1986.
[9] S.A. Nasar, I. Boldea, L.E. Unnewehr, Permanent Magnet, Reluctance
and Self-Synchronous Motors, Florida: CRC Press, Inc., 1993.
[10] S. Morimoto, Y. Takeda, T. Hirasa, and K. Taniguchi, Expansion of
Operating Limits for Permanent Magnet Motor by Current Vector Con-
trol Considering Inverter Capacity, IEEE Trans. Industry Applications,
vol. 26, no. 5, pp. 866-871, Sep./Oct. 1990.
[11] B.J. Chalmers, L. Musaka, and D.F. Gosden, Variable-Frequency Syn-
chronous Motor Drives for Electric Vehicles, IEEE Trans. on Industry
Applications, vol. 32, no. 4, pp. 896-903, July/Aug. 1996.
[12] S. Morimoto, Y. Takeda, and T. Hirasa, Loss Minimization Control of
Permanent Magnet Synchronous Motor Drives, IEEE Trans. on Indus-
trial Electronics, vol. 41, no. 5, pp. 511-517, Oct. 1994.
[13] F.F. Bernal, A. Garca-Gerrada, and R. Faure, Model-Based Loss
Minimization for DC and AC Vector-Controlled Motors Including Core
Saturation, IEEE Trans. on Industry Applications, vol. 36, no. 3, pp.
755-763, May/June 2000.
[14] R.B. Colby and D.W. Novotny, An Efficiency-Optimizing Permanent
Magnet Synchronous Motor Drive, IEEE Trans. on Industry Applica-
tions, vol. 24, no. 3, pp. 462-469, May/June 1988.
[15] Y. Nakamura, T. Kudo, F. Ishibashi, and S. Hibino, High-Efficiency
Drive Due to Power Factor Control of a Permanent Magnet Synchro-
nous Motor, IEEE Trans. on Power Electronics, vol. 10, no. 2, pp.
247-253, March 1995.
[16] S. Vaez and M.A. Rahman, An On-line Loss Minimization Controller
for Interior Permanent Magnet Motor Drives, IEEE Trans. on Energy
Conversion, vol. 14, no. 4, pp. 1435-1440, Dec. 1999.
[17] C. Mademlis and N. Margaris, Loss Minimization in Vector-Controlled
Interior Permanent-Magnet Synchronous Motor Drives, IEEE Trans.
on Industrial Electronics, vol. 49, no. 6, pp. 1344-1347, Dec. 2002.
[18] P.C. Krause, Analysis of Electric Machinery, New York: McGraw Hill,
1986.
[19] V.B. Hosinger, Performance of Polyphase Permanent Magnet Ma-
chines, IEEE Trans. on Power Apparatus and Systems, vol. PAS-99,
no. 4, pp. 1510-1518, July/Aug. 1980.
[20] S.A. Nasar, Handbook of Electric Machines, McGraw-Hill: New York,
1987.
[21] F. Fernndez-Bernal, A. Garca-Cerrada, and R. Faure, Determination
of Parameters in Interior Permanent-Magnet Synchronous Motors With
9
Iron Loss Without Torque Measurement, IEEE Trans. on Industry Ap-
plications, vol. 37, no. 5, pp. 1265-1272, Sept./Oct. 2001.
[22] C. Mademlis, J. Xypteras, and N. Margaris, Loss Minimization in
Surface Permanent-Magnet Synchronous Motor Drives, IEEE Trans.
on Industrial Electronics, vol. 47, no. 1, pp. 115-122, Feb. 2000.
[23] I. Boldea, Reluctance Synchronous Machines and Drives, Clarendon
Press: Oxford, 1996, pp. 26-30.
BIOGRAPHIES
Christos Mademlis was born in Arnea Chalkidikis, Greece, on February 7,
1964. He received the Diploma degree in Electrical Engineering (1st class
Hons.) and the Ph.D. degree in electrical machines from the Aristotle Univer-
sity of Thessaloniki, Greece, in 1987 and 1997, respectively.
Since 1990, he has been with the Electrical Machines Laboratory, Depart-
ment of Electrical and Computer Engineering, Aristotle University of Thessa-
loniki as a Research Associate (1990-2002) and he was recently appointed as
a Lecturer in the same Department. His research interests are in the areas of
electrical machines and drives, especially in machines design and control
optimization.

Iordanis Kioskeridis was born in Thessaloniki, Greece, on January 29, 1965.
He received the Diploma degree in Electrical Engineering and the Ph.D. de-
gree in asynchronous motors loss minimization from Aristotle University of
Thessaloniki, Greece, in 1989 and 1994 respectively.
From 1995 to 2000 he worked as Superintendent Engineer in the Natural
Gas of Greece Project. He is currently working with the Technological Educa-
tional Institute of Thessaloniki and is engaged in teaching power electronics
and electrical machines. His primary research activities include power elec-
tronic converters, control and modeling of adjustable speed drives.

Nikos Margaris was born in Athens, Greece, on February 10, 1949. He re-
ceived the Diploma in physics, the Postgraduate degree in electronics and the
Ph.D. degree in automatic control from the Aristotle University of Thessalo-
niki, Greece, in 1972, 1975 and 1982 respectively.
Since 1977, he has been with the Electrical and Computer Engineering
Department, Aristotle University of Thessaloniki, teaching graduate and
postgraduate courses in electronics, automatic control, power electronics and
circuit theory. From 1992 to 1994 he was the Director of the Electronics and
Computer Division and from 1993 to 1995 the Vice President of the Electri-
cal and Computer Engineering Department.
His current research interests include the loss minimization in variable and
constant speed drives, the study of nonlinear oscillations, the analysis and
design of switch mode dc-dc converters and the robust control theory.