IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 25, NO.

2, MARCHiAPRIL 1989 265

Modeling, Simulation, and Analysis of

Permanent-Magnet Motor Drives ,
Part I: The Permanent-Magnet
Synchronous Motor Drive

Abstract-Two types of permanent-magnet ac motor drives are power density when compared to the IM or the wound-rotor
available in the drives industry. These are the permanent-magnet SM, which makes it preferable for certain high-performance
synchronous-motor (PMSM) drive with a sinusoidal flux distribution, applications like robotics and aerospace actuators. Although
and the brushless dc motor (BDCM) drive with a trapezoidal flux
distribution. The application of vector control to the PMSM and present applications tend to be in the fractional to 30-hp range,
complete modeling, simulation, and analysis of the drive system are the feasibility of using the PMSM for up to 125 hp has recently
given. State-space models of the motor and speed controller and real-time [3] been evaluated.
models of the inverter switches and vector controller are included. Most of the earlier research on the PMSM concentrated on
Performance differences due to the use of pulsewidth-modulation (PWM) its operation from busbar voltages [4], [5]. Damper windings
and hysteresis current controllers are also examined. Particular attention
is paid to the motor torque pulsations and speed response. Some were used to run the machine up to speed on induction motor
experimental verification of the drive performance is also given. action with the machine pulling into synchronism by a
combination of the reluctance and synchronous motor torques
I. INTRODUCTION provided by the magnet. During the startup, the magnet exerts
HE PERMANENT-magnet synchronous motor (PMSM) a braking torque that opposes the induction-motor-type torque
T has numerous advantages over other machines that are
provided by the damper windings. The torque provided by the
conventionally used for ac servo drives. The stator current of damper windings must therefore overcome this magnet brak-
ing torque, in addition to the load and friction, to run the motor
an induction motor (IM) contains magnetizing as well as
torque-producing components. The use of the permanent up successfully. The well-established d, q model of the
magnet in the rotor of the PMSM makes it unnecessary to wound rotor synchronous machine is easily adapted to study
supply magnetizing current through the stator for constant air- the performance of a permanent-magnet synchronous motor
with damper windings. More recently [6]-[ 121, the possibility
gap flux; the stator current need only be torque-producing.
Hence for the same output, the PMSM will operate at a higher of using the PMSM for servo drives has been examined. It is
recognized that with rotor position feedback, the motor can be
power factor (because of the absence of magnetizing current)
held in synchronism with the inverter at all times, and the rotor
and will be more efficient than the IM. The conventional
wound-rotor synchronous machine (SM), on the other hand, cage [lo], [ll] is not needed to run the machine up to the
must have dc excitation on the motor, which is often supplied commanded speed.
by brushes and slip rings. This implies rotor losses and regular Vector control is normally used in ac machines to convert
brush maintenance, which implies downtime. Note that the them, performancewise, into equivalent separately excited dc
key reason for the development of the PMSM [l] was to machines which have highly desirable control characteristics.
The application of vector control to the PMSM and its drive
remove the foregoing disadvantages of the SM by replacing its
system simulation are given in this part of the two-part paper.
field coil, dc power supply, and slip rings with a permanent
For high-performance servo drives, hysteresis [3] or PWM
magnet. The PMSM, therefore, has a sinusoidal induced EMF
and requires sinusoidal currents to produce constant torque current controllers are used to ensure that the actual currents
just like the SM. Current research [2] in the design of the flowing into the motor are as close as possible to the sinusoidal
references. This paper evaluates the use of hysteresis as well
PMSM indicates that it has a higher-torque-to-inertia ratio and
as PWM current controllers when used in the PMSM drive.
Paper IPCSD 88-22, approved by the Industrial Drives Committee of the Complementary switching of the inverter power devices is
IEEE Industry Applications Society for presentation at the 1987 Industry considered undesirable and not implemented. In this evalua-
Applications Society Annual Meeting, Atlanta, GA, October 19-23. Manu- tion, the entire nonlinear drive system is simulated, including
script released for publication July 12, 1988.
P. Pillay is with the Department of Electrical and Electronic Engineering, the nonlinear d, q axis equations of the motor, state-space
University of Newcastle-upon-Tyne, Merz Court, Newcastle-upon-Tyne, model of the speed controller [ 121, and real-time models of the
England NE1 7RU. inverter switches and vector controller. Although the switches
R. Krishnan is with the Electrical Engineering Department, Virginia
Polytechnic Institute and State University, Blacksburg, VA 24061. are assumed to be ideal, the simulation software developed is
IEEE Log Number 8825302. flexible enough to accommodate their turn-on and turn-off

OO93-9994/89/03OO-0265$01.OO 0 1989 IEEE

266 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 25, NO. 2, MARCHIAPRIL 1989

times. Every instant of a power switch opening or closing is and the equation for the motor dynamics is
simulated to predict accurately the stator current changes and
resulting torque pulsations. The effects of changing the Te= TL + Bu, + JPW,. (6)
magnitude of the hysteresis windows on the motor torque P is the number of pole pairs, TL is the load torque, B is the
pulsations, the inverter switching frequency, and drive per- damping coefficient, W, is the rotor speed, and J is the moment
formance are evaluated. In addition, a comparison between the of inertia. The inverter frequency is related to the rotor speed
PWM and hysteresis current controllers, based on the previous as follows:
criteria, is also made. Finally, both the small and large signal
behavior of the drive are evaluated. These results can be used w, = Pw,. (7)
to evaluate the drive dynamics for application considerations
[ 131. Key results are verified experimentally.
The machine model is nonlinear as it contains product terms
The paper is organized as follows. Section I1 presents the such as speed with id and iq. Note that U,, iq, and id are state
mathematical model of the PMSM. Using this model, vector variables.
control of the PMSM is developed in Section 111. The For dynamic simulation, the equations of the PMSM
operation of the current controllers used in high-performance presented in (1)-(6) must be expressed in state-space form as
servo drives are discussed in Sections IV and V while the shown in (8)-(10):
structure of the entire drive system is explained in Section VI. pid= (Ud- Rid+ w,Lqiq)/Ld (8)
Sections VI1 and VI11 have the results and conclusion,
respectively. piq = ( U q - Riq- WsLdid- w,hOj)/Lq (9)
11. MACHINE
MODEL PW, = ( T ,- TL - Bw,)/J. (10)
The stator of the PMSM and the wound rotor SM are The d, q variables are obtained from a, b, c variables through
similar. The permanent magnets used in the PMSM are of a the Park transform defined below:
modern rare-earth variety with high resistivity, so induced

1
currents in the rotor are negligible. In addition, there is no cos (e) cos (e-za/3) cos (e+2a/3)
difference between the back EMF produced by a permanent
magnet and that produced by an excited coil. Hence the [:;] [ = 2/3 sin (0) sin (0 - 2 d 3 )
1/2 1 /2
sin (0 + 2 d 3 )
1 /2

["1
mathematical model of a PMSM is similar to that of the wound
rotor SM. The following assumptions are made in the
derivation. * (11)
1) Saturation is neglected although it can be taken into
account by parameter changes.
The a, b, c variables are obtained from the d, q variables
2) The induced EMF is sinusoidal.
through the inverse of the Park transform defined below:
3) Eddy currents and hysteresis losses are negligible.
4) There are no field current dynamics.
5) There is no cage on the rotor.
With these assumptions, the stator d, q equations of the
PMSM in the rotor reference frame are [lo]-[12]:
cos ( 0 - 2 d 3 )
cos (6+ 2 d 3 )
sin (6)
sin ( 0 - 2 d 3 )
sin (0 + 2 d 3 )
:] [i;]
1 *

(12)
Note that these transformations apply equally well to currents
and flux linkages. The total input power to the machine in
terms of the a, 6, c variables is
where
power = vaia+ Ubib + u,i,
A, = Lqiq (3)
while in d, q variables,
and
power = 3 (Udid + uqiq)/2 (14)
for a balanced system.
ud and uq are the d, q axis voltages, id and iq are the d , q axis
stator currents, Ld and L , are the d , q axis inductances, hd and 111. VECTOR
CONTROL
OF A PMSM
A, are the d, q axis stator flux linkages, while R and w, are the Equations (3) and (4) are represented in a phasor diagram
stator resistance and inverter frequency, respectively. X,J is shown in Fig. 1 . The rotor flux linkage revolves at rotor speed
the flux linkage due to the rotor magnets linking the stator. or and is positioned away from a stationary reference by the
The electric torque is rotor angular position, given by

Te= 3 P [haJiq+ (Ld - Lq)idiq]/2, (5) Or= [

J
U, dt (15)
PILLAY AND KRISHNAN: PERMANENT-MAGNET MOTOR DRIVES, PART I 267

Fig. 2. Inverter circuit of PMSM.

)r phase A

Fig. 1. Phasor diagram of vector controller.

where t is the time. If id is forced to be zero, then

hd=haf (16)
and

T,= 3Phafiq/2. (17)

Since the magnetic flux linkage is a constant, the torque is
directly proportional to the q axis current. This is represented
as
NI I I

T, = Ktiq (18)
Fig. 3. Hysteresis current controller.
where

K , = 3PXaf/2. (19) TABLE I

_.

The torque equation is similar to that of a separately excited dc i,* 1, T1 T4 VO”

motor, and this completes the transformation of a PMSM to an
equivalent separately excited dc motor. Similarly, the intro- 20 i,, 5 (i: - Ai) on off + VJ2
duction of a negative id will weaken the airgap flux as seen r O io z (i: + Ai) off off - V,/2 ( 0 4 on)
from ( 4 ) . <O i, 2 (i: + Ai) off on - VdJ2
<O io 5 (i: - Ai) off off + V J 2 ( 0 1 on)
IV. HYSTERESIS
CURRENT
CONTROLLER
The power circuit that drives the PMSM is shown in Fig. 2 . through the machine winding cannot go to zero instantane-
It is assumed that a reasonably well-filtered dc supply is ously, the freewheeling diode across its complementary
available. The six switches T l - T 6 are used to control the transistor, in this case T 4 , begins to conduct the phase A
three stator phase currents. The control strategy is as follows. current. When this occurs, the voltage of phase A switches
The actual values of i,, and ib that are flowing into the motor from + v&/2 to - v & / 2 ,where the midpoint of the dc supply
are measured. From this i, can be constructed; this removes v d c is taken as the reference. The opposite occurs when T4
the need for an additional current sensor. The actual and switches from “on” to “off.” A similar procedure exists in
reference values are compared and error signals generated. In the other phases. The reason that this is called a hysteresis
making the comparison between the actual currents and the controller is that the phase voltage switches to keep the phase
reference values, the scheme in Fig. 3 is used. Fig. 3 shows currents within the hysteresis bands. The phase currents are,
the reference value i,*. In addition, two other curves consisting therefore, approximately sinusoidal: the smaller the hysteresis
of i,* + Ai and i,* - A i are shown. Ai defines the hysteresis bands, the more closely do the phase currents represent sine
bands. The hysteresis property allows the actual value of i,, to waves. Small hysteresis bands, however, imply a high
exceed or be less than the reference value by Ai. The logic is switching frequency, which is a practical limitation on the
shown in Table I. Similar logic applies to the other two phases. power device switching capability. Increased switching also
Note that complementary switching of the power devices is implies increased inverter losses.
considered undesirable and, therefore, is not used.
Whenever TI is “on,” i, increases positively using either V. PWM CURRENT
CONTROLLER
the B or C phases as a return path. As soon as T1 switches A second method used to generate the required stator
from an “on” to an “off” position, and since the current currents is to use a pulsewidth-modulated (PWM) current
268 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 25, NO. 2, MARCHIAPRIL 1989

the inverse Park transform to generate the a, b, c stator

reference currents. Rotor position feedback is needed to

v v
shwtooth
generate these currents. The hysteresis or PWM current
controller attempts to force the actual motor currents to equal
the commanded values at all times. Current feedback is
required for the hysteresis or PWM current controllers to
achieve this. Current control is implemented by the appropri-
ate firing of power devices T l - T 6 as discussed previously.
Both position and speed feedback can be obtained from a
resolver/signal processor combination.
When greater than rated speed is commanded, the machine
then operates in the constant power or flux-weakening mode.
Fig. 4. Pulsewidth-modulated current controller. Here, the airgap flux is weakened by applying a direct axis
current in opposition to the rotor magnet flux. The torque-
controller. The actual values of the three stator currents are speed profile of the drive is as shown in the block labeled FW.
measured and compared to the reference currents. Thus error The output of the block is unity up to rated speed and decreases
currents are generated. These error currents are compared to a hyperbolically with speed between the rated and maximum
sawtooth-shaped triangular wave as shown in Fig. 4 . speeds to ensure constant output power. When the output of
If the current error signal is positive and larger than the FW is unity, then A,,, = X,fand id* = 0. If the output of FW
sawtooth, the voltage is switched positively, while if the is less than unity, then a lower reference torque is demanded.
current error signal is positive and smaller than the sawtooth, In addition, A,,, is less than Xaf so that a negative id* is
the voltage is switched negatively. Note that it is unnecessary commanded to buck the magnet flux. The speed controller is
to use complementary switching to achieve this voltage designed at rated speed but is required to operate properly up
profile. to the maximum speed.
For example, if T1 is conducting, U,, is equal to + V d 2 ,
where Vdc is the dc supply voltage and the reference is taken VII. RESULTS
as the midpoint of the supply. By switching T1 off, the Digital computer simulations of the entire drive system
freewheeling diode across T4 immediately starts conducting to shown in Fig. 5 are presented in this section. The state-space
maintain the current flow through the motor inductance. This models of the PMSM and speed controller and switching logic
automatically forces U,, to equal to - Vd,/2 even though T4 is of the current controllers are included in the simulation. Every
not yet conducting. This is called a PWM current controller instant of a power device switching on or off is modeled. The
because of the pulsewidth modulation of the voltage. speed controller is a particular case of the pseudo-derivative
The advantage of the PWM current controller over hys- feedback controller discussed in [ 121. Large and small signal
teresis is that the switching frequency is preset, and it is, transients are considered, and in addition, comparisons be-
therefore, easy to ensure that the inverter switching capability tween the PWM and hysteresis current controllers are made.
is not exceeded. In the hysteresis controller the switching Fig. 6 shows that key transient results when a PMSM (see
frequency depends on the value of the hysteresis window, and the Appendix for the parameters) is started up from standstill
the actual switching frequency demanded from the inverter is to a speed of 1750 r/min. A 2-kHz PWM current controller is
unknown. A trial-and-error procedure must be adopted to used. The speed is underdamped in the design used here. The
ensure that the inverter switching frequency is not exceeded. linear manner in which the speed increases is possible because
The advantage of the hysteresis over the PWM controller is of vector control. During the startup period, the commanded
that, from a control point of view, there is no transportation torque equals the maximum capability of the motor. This
delay or system lag. In the PWM controller this does exist with ensures that the machine runs up in the shortest time possible.
the average lag being equal to half the period of the PWM. From Fig. 6 it is clear that the machine runs up before
However, if this lag is less than about one-tenth the stator time completing a full current cycle. The phase voltage switches
constant of the machine, it has a negligible effect on the drive continuously in an effort to force the actual current to equal the
performance. companded value. The close tracking of the commanded
current by the actual is evident, except for the initial rise time
VI. DRIVESYSTEM due to the stator time constant.
The machine, speed and position feedback, speed and At 0.025 s, a load of 1 pu is applied to the motor. This
current controllers, and inverter constitute the PMSM drive causes a small decrease in the speed as shown in Fig. 6. This
system as shown in Fig. 5. All reference or commanded values decrease is barely perceptible and is less than the speed
are superscripted with a "*" in the diagram. The error overshoot that occurred during the startup. The motor electric
between the reference and actual speeds is operated upon by torque increases to 1 pu to satisfy the load torque require-
the speed controller to generate the torque reference. In the ments.
constant airgap flux mode of operation where id* = 0, the Fig. 7 shows the corresponding curves when a hysteresis
torque reference is divided by the motor torque constant to instead of a PWM current controller is used. It is clear that the
give the reference quadrature axis current. This goes through large signal speed transient is the same as when the PWM
PILLAY AND KRISHNAN: PERMANENT-MAGNET MOTOR DRIVES, PART I 269

TI,TZ,T3,T4,TS,T6

Fig. 5 . Schematic of PMSM speed servo drive system.

FICTUFIL
5
0

(a)

; . I
a-

-
5
E?
I

(b)
9 I I I

QCTURL

9 I I I 7
'0.00 o'.ozs 0.05
TIME ( S I
(C) (c)
Fig. 6. Transients when fed from PWM CSI. (a) Speed. @) Torque. (c) Fig. 7 . Transients when fed from hysteresis CSI. (a) Speed. @) Torque. (c)
Current. Current.
270 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 25, NO. 2. MARCHlAPRIL 1989

8- I I I
I

5
a
'8
(a)

-8 q q 1

3 .
a-
5
e? I
v I I I
'0.m 0.025 0.05
(b) TIME( S I
(b)
Fig. 9. Transients for 0.1-pu load. (a) Speed. (b) Torque.
0 FICTUQL

0.05
TIME(S1
(C)
-
L O

Fig. 8. Transients when fed from hysteresis CSI. (a) Speed. (b) Torque. (c)
5
U
Current. '8
current controller is used. However, the actual electric torque
in Fig. 7 shows larger instantaneous pulsations than those in
P
-. , -- I I
x
I
REF
Fig. 6 . These larger pulsations are produced by the current 0 FICTUHL
waveform in Fig. 7. The voltage is used to force the currents -8
3 .--
to remain within the hysteresis bands. Although the oscilla- a-
tions in the current and, consequently, torque are larger with 5, -
the hysteresis current controller, the average value of the E?- I I I
torque is the same with both current controllers, thus
producing the same large-signal dynamics. Lower current
oscillations, and hence torque pulsations, are produced by P I I
x
I
REF
decreasing the size of the hysteresis bands as shown in Fig. 8. 0 QCTUFIL
This is achieved at a higher switching frequency of the
I1
inverter. The linear manner in which the speed increases
during the initial startup and subsequent underdamped re-
sponse are evident.
I I I
It is therefore clear that the speed transient is similar 1 i

irrespective of whether a PWM or hysteresis current controller

is used. However, if the hysteresis bands are so large as to
produce large-magnitude and low-frequency torque pulsa-
tions, then significant speed pulsations would occur.
The torque and current responses when a load torque of 0.1
pu is applied are shown in Fig. 9. These are scaled-down increase in the current as shown in Fig. 10. Up to now
versions of the curves corresponding to the load torque of 1 transient results have been presented. Steady-state results are
pu. This indicates that when vector control is used, the large presented next.
and small signal responses are very similar. From the results presented earlier, it is clear that the motor
Fig. 10 shows the speed, torque, and current for a 0.1-pu torque pulsations increase as a function of the hysteresis
increase in the speed of the machine after it has run up. When window size. This trend is plotted in Fig. 11 in pu. There is a
the speed command is input, a pulse of torque is demanded to linear relationship between the magnitude of the motor torque
increase the actual speed of the motor. This is provided by an pulsations and the window size. This result was obtained by
PILLAY AND KRISHNAN: PERMANENT-MAGNET MOTOR DRIVES, PART I 27 1

0.16..

2.0.12.
v

4.3
4
m
0.08.
P
0
a
W
0
0.04

0.04 0.08 0.12 0.16

Window size (pu) Fig. 13. Back EMF of PMSM. X axis: 5 ms/div. Y axis: speed (pu)
Fig. 1 1 . Torque pulsations versus window size.

Fig. 14. Measured speed of PMSM. X axis: 5 ms/div. Y axis: speed (pu).

the PWM switching frequency should be chosen on the basis

of the torque bandwidth and inverter switching capability
1 rather than on the resulting torque pulsations.
0.1 0.2 0.3 0.4 0.5

Hysteresis windov size (pu)

Experimental Verification
Fig. 12. Inverter frequency versus window size. Fig. 13 shows the back EMF of the PMSM, thus verifying
that it is sinusoidal. To test the model developed, the machine
varying the hysteresis window size and determining the was started up and its speed measured as shown in Fig. 14.
corresponding torque pulsations. If the window size is The theoretical prediction of this speed is given in Fig. 15. The
expressed in pu with the rms motor stator current as the base, theoretical prediction and the practical measurement compare
then the pu motor torque pulsations are approximately equal to favorably, thus verifying that the model and computer
this window size. In other words, multiplication of the pu simulation program used in this investigation is valid for the
window size by the rated torque of the motor gives the actual machine used.
value of the motor torque pulsations in Nm.
From the previous results, it is also clear that the inverter VIII. CONCLUSION
switching frequency increases as the hysteresis window size This paper has presented the modeling, simulation, and
decreases. This relationship is shown in Fig. 12. As the analysis of a vector-controlled PMSM drive. A systematic
hysteresis window size decreases, the required inverter derivation of the vector controller for the PMSM was given.
switching frequency increases in a nonlinear manner. A The performance differences due to both PWM and hysteresis
tenfold decrease in the window size results in approximately a current controllers were examined.
sevenfold increase in the inverter switching frequency. Since vector control transforms the PMSM to an equivalent
It was found that the torque pulsations due to the 2-kHz separately excited dc machine, the transfer function between
PWM current controller shown in Fig. 6 were extremely the electric torque and current is linear. The results therefore
small, and their effect on the speed was not even noticeable. indicate that the small- and large-signal responses are very
Changing the PWM switching frequency did not affect the similar. This result is not true for ac machines that are not
torque pulsations as much as varying the window size in the under vector control since then the model is nonlinear. The
hysteresis current controller, Hence the results indicate that large- and small-signal speed response is the same whether
272 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 25, NO. 2 , MARCHiAPRIL 1989

1:

I I
0 6.m 6.40 0.50
TIME( SI *lo-’
Fig. 15. Predicted speed of PMSM

PWM or hysteresis current controllers are used. This is 4 Angle between stator phase A and the d axis, rad.
because, even though the torque pulsations may be different * Superscript indicating reference value.
due to the use of different current controllers, the average
value, which determined the overall speed response, is the APPENDIX
same. MOTORPARAMETERS
The larger the hysteresis bands, the lower the inverter
switching frequency and the larger are the motor torque We have the following:
pulsations. A PWM switching frequency of 2 kHz produces
approximately the same torque pulsations as a hysteresis R=1.4 Q
current controller switching at 3.8 kHz. From this point of Ld=6.6 mH
view the PWM is better. However, the PWM can have a delay
of up to one period corresponding to its carrier frequency. Lq=5.8 mH
This may have implications in very high speed applications. J=0.00176 kgm2
However, if this delay is less than one-tenth of the stator time
constant, it has been found in practice that it has a negligible B = 0.00038818 Nmlradls
effect on the drive performance. A,= 0.1546 Vlradls
The motor torque pulsations are related linearly to the size
of the hysteresis bands. If the window size is expressed in pu 6 poles.
with the rms value of the motor rated current as the base, then
ACKNOWLEDGMENT
the pu torque pulsations are approximately equal to this pu
window size. The authors acknowledge Industrial Drives, Kollmorgen
Corporation, of Radford, VA, for the donation of the
NOMENCLATURE permanent-magnet synchronous motor drive that was used in
Damping constant, N * mlradls. this paper.
a, b, and c phase currents, A.
d and q axis stator currents, A . REFERENCES
Moment of inertia, kg-m2. [l] D. P. M. Cahill and B. Adkins, “The permanent magnet synchronous
stator d, q inductances, H. motor,” Proc. Inst. Elec. Eng., vol. 109, part A, no. 48, pp. 483-
491, Dec. 1962.
Derivative operator. [2] R. Krishnan and A. J. Beutler, “Performance and design of an axial
Number of pole pairs. field permanent magnet synchronous motor servo drive,” in Proc.
Stator resistance, Q . IEEE Ind. Appl. Soc. Annu. Meeting, 1985, pp. 634-640.
[3] E. Richter, T. J. E. Miller, T. W. Neumann, and T. L. Hudson, “The
Electric torque, N m. ferrite PM ac motor-A technical and economic assessment,” IEEE
Load torque, N*m. Trans. Ind. Appl., vol. IA-21, no. 4, pp. 644-650, May/June 1985.
d and q axis stator voltages, V. [4] V. B. Honsinger, “Permanent magnet machines: Asynchronous
operation,” IEEE Trans. PowerApp. Sysr., vol. PAS-99, no. 4, pp.
DC bus voltage, V. 1503-1509, July/Aug. 1980.
Rotor speed, radls. [5] T. J. E. Miller, “Transient performance of permanent magnet
Synchronous speed, radls. machines,” in Proc. Ind. Appl. Soc. Annu. Meeting, 1981, pp. 500-
502.
Mutual flux linkage between rotor and stator [6] W. Leonard, Control of Electrical Drives. New York: Springer-
due to magnet, Wb-turn. Verlag, 1984.
Stator d and q axis flux linkage, Wbvturn. [7] G . Pfaff, A. Weschta, and A. Wick, “Design and experimental results
of a brushless ac servo drive,” in Proc. IEEE Ind. Appl. Soc. Annu.
Air gap flux linkage, Wb aturn. Meeting, 1982, pp. 692-697.
Angle between stator phase A and the q axis, rad [8] R. Krishnan, “Analysis of electronically controlled motor drives,”
PILLAY AND KRISHNAN: PERMANENT-MAGNET MOTOR DRIVES. PART I 273

class notes, Virginia Polytechnic Inst. and State Univ., Blacksburg, Societies of the IEEE and a member of the Industrial Drives Committee of the
1986. Industry Applications Society. He is also an Associate Member of the
M. Lajoie-Mazenc, C. Villanueva, and J. Hector, “Study and Institution of Electrical Engineers, England, and a member of the Greek
implementation of hysteresis controlled inverter on a permanent magnet honor society, Phi-Kappa-Phi.
synchronous machine,” IEEE Trans. Ind. Appl., vol. IA-21, no. 2,
pp. 408-413, Mar./Apr. 1985.
P. Enjeti, J. F. Lindsay, and M. H. Rashid, “Stability and dynamic
performance of variable speed permanent magnet synchronous mo-
tors,” in Proc. IECON, 1985, pp. 749-754.
-, “Parameter estimation and dynamic performance of permanent
magnet synchronous motors,” in Proc. IEEE Ind. Appl. SOC. Annu.
Meeting, 1985, pp. 627-633.
P. Pillay and R. Krishnan, “Control characteristics and speed
controller design for a high-performance permanent magnet synchro-
nous motor drive,” in Proc. IEEE 1987 Power Electronics Special- Ramu Krishnan (S’81-M’82) received the B.E.,
ists’ Cony., 1987, pp. 598-606. M.E., and Ph.D. degrees in electrical engineering.
, “Application characteristics of permanent magnet synchronous He taught for seven years in India. He was Staff
and brushless dc motors for servo drives,” in Proc. Ind. Appl. Soc. Engineer and Principal Investigator of ac servo
Annu. Meeting, 1987, pp. 380-390. drive projects at Gould Research Center, Rolling
Meadows, IL, between 1982 and 1985. Since
September 1985 he has been an Associate Professor
Pragasen Pillay (S’84-M’87) received the B.Eng, in the Electrical Engineering Department at Vir-
M.Sc(Eng), and Ph.D. degrees, all in electrical ginia Polytechnic Institute and State University,
engineering. The Ph.D. degree was obtained at the Blacksburg. His teaching and research interests are
Virginia Polytechnic Institute and State University,
Blacksburg, while funded by a Fulbright scholar-
-.
in high-uerformance vector-controlled variable-
speed drives, switched-reluctance motor drives, electrical machine design,
ship. and static power conversion. He has published more than 50 papers on these
He is currently with the Department of Electrical topics. He has developed a graduate program in electric motor drives and
and Electronic Engineering, University of New- machine design at Virginia Polytechnic.
castle-upon-Tyne, England, UK. His research inter- Dr. Krishnan is a recipient of four IEEE-IAS awards for his papers, both
ests are in the modeling, control, and design of presented and published. He has been Associate Editor of the IEEE
electric motor drive systems. TRANSACTIONS ON INDUSTRIAL ELECTRONICS since June 1987. He is a
Dr. Pillay is a past recipient of an IEEE prize award paper. He is a member member of the IAS Machine Tools, Robotics, and Factory Automation
of the Industry Applications, Industrial Electronics, and Power Engineering Committee.