920 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 30, NO.

4, JULY / AUGUST 1994

Wide-Speed Operation of Interior

Permanent Magnet Synchronous Motors
with High-Performance Current Regulator
Shigeo Morimoto, Member, IEEE, Masayuki Sanada, Member, IEEE, and Yoji Takeda,
Member, IEEE

Abstract-Interior permanent magnet synchronous motors can [4].In the flux-weakening region, the terminal voltage is
be applied to applications requiring wide-speed operation. The nearly the maximum available voltage of the inverter, and
current vector control algorithm of an interior permanent mag-
net synchronous (IPM) motor for constant power operation over
as a result, the commanded voltage vector sometimes
the base speed is proposed. As the available voltage controlling exceeds the maximum available voltage in transient oper-
the armature current vector is small in the flux-weakening ations. In this case, the d- and q-axis current regulators
constant power region, the current vector sometimes becomes are saturated and affect each other. As a result, the
uncontrollable in transient operations because of the current responses of current, torque, and speed become worse.
regulator saturation. The high-performance current regulator is
also proposed to improve the current responses in the flux- Variable speed control system of the IPM motor is
weakening region, which includes the decoupling current con- proposed in this paper for wide-speed operation over the
troller and the voltage command compensator. The control base speed. The effects of the armature reaction and the
performances are confirmed by several drive tests with respect cross-coupling effects on the control performances are
to the prototype IPM motor. examined. The high-performance current regulator with
the decoupling of the d and q axis and voltage command
compensation is also proposed. The proposed control
I. INTRODUCTION algorithms are implemented in the fully digital speed

P ERMANENT magnet synchronous motors (PM mo-

tors) fed by PWM inverters are widely used for
industrial applications, especially servo drive applications,
control system based on a digital signal processor (DSP),
and several drive tests are carried out with respect to the
prototype IPM motor.
in which constant torque operation is desired. In traction
and spindle drives, on the other hand, constant power PRINCIPLE
11. CONTROL OF CURRENT
VECTOR
operation is desired. In the past, dc motors and induction A. Current and Voltage Constraints
motors have generally been applied to these applications. In the d-q coordinates which rotate synchronously with
These motors are controlled by the field-weakening con- an electrical angular velocity w , the voltage equation and
trol method in the constant power region. The direct the torque equation of IPM motors are expressed as
control of magnet flux is not available in PM motors. The follows:
airgap flux, however, can be weakened by the demagnetiz-
ing effect due to the d a i s armature reaction [ll-[4]. This
technique is called flux weakening.
Interior permanent magnet synchronous (IPM) motors
are suitable for constant power operation by flux-weaken- T + P,(Ld - Lq)idiq
= Pn4uiq
ing control because the IPM motor has a relatively large where
armature inductance, and as a result, enough flw- . .
Id7 'q d- and q-axis components of armature cur-
weakening effect due to the d-axis armature reaction is
rent,
expected [5],[6]. The effects of the d- and q-axis armature
' d , q' d- and q-axis components of terminal volt-
reactions on the performance of current control are domi-
nant, especially at high speeds and in transient operations age,
40 = m4j7
4f maximum flux linkage due to permanent
Paper IPCSD 94-26, approved by the Industrial Drives Committee of magnet per phase,
the IEEE Industry Applications Society for presentation at the IAS
Power Conversion Conference (PCC-Yokohama '93). Manuscript re- R armature resistance,
leased for publication April 5, 1994. Ld, Lq d- and q-axis components of armature self-
The authors are with the Department of Electrical and Electronic inductances,
Systems, University of Osaka Prefecture, 1-1 Gakuen-cho, Sakai 593,
Japan. P = d/dt,
lEEE Log Number 9402624. pn number of pole pairs.

0094-9994/94$04,00 0 1994 IEEE

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY BOMBAY. Downloaded on June 30,2021 at 17:17:18 UTC from IEEE Xplore. Restrictions apply.
 MORIMOTO et al.: WIDE-SPEED OPERATION OF IPM MOTORS 921

Considering the voltage and current constraints, the ar- Maximum

mature current Za and the terminal voltage V , are limited Torque-per-Amp
as follows: Trajectory

Current-Limit

v, = -
4 IV,,. (4)

The maximum current I,, is a continuous armature cur-

rent rating in continuous operation or a maximum avail-
able current of the inverter in limited short-time opera-
tion. The maximum voltage Vu, is the maximum available
output voltage of the inverter depending on the dc-link
voltage. The critical condition of (3) (i.e., Z, = l a m is
) -10 0
given by the current-limit circle in the id-i, plane as
Fig. 1. Current-limit circle, voltage-limit ellipse, and maximum torque-
shown in Fig. 1. per-amp trajectory in &i plane.
B. Control in Constant Torque Region
As the IPM motor has a saliency (Ld < L , ) and, as a The critical condition of (8) (i.e., V , = V,,) is given by the
result, the reluctance torque [the second term in (2)] is voltage-limit ellipse in the id-iq plane as shown in Fig. 1.
available, the armature current vector is controlled in The voltage-limit ellipse becomes small as the speed in-
order to produce the maximum torque per armature cur- creases, and as a result, the current vector producing T,,
rent ampere [7], [8]. From (21, the relationship between id cannot satisfy the voltage constraint above the base speed
and i , for the maximum torque-per-amp control is de-
wbase-
rived as The d- and q-axis components of armature current are
controlled in order to keep V , equal to V,, in the
flux-weakening constant power region. The relationship
between id and i, of the flux-weakening control is derived
from (8) and V, = E,:
This relationship is shown as the maximum torque-per-
amp trajectory in Fig. 1. The maximum torque T,, is
produced when 1, = 4- =Z, considering the cur-
rent constraint. The current vector (ids, is,> producing
this maximum torque T,, is a cross point of the maxi-
mum torque-per-amp trajectory and the current-limit cir- where li,l I V,,/(wL,), i.e., lUdol I V,,.
cle, which corresponds to point A in Fig. 1. The d- and If the current vector is controlled according to (9), the
q-axis currents i d , and i, are derived from (5) and resultant terminal voltage V , including the resistance
4- = :,z, drop is always kept within V,, in steady state. Consider-
ing the current constraint (31, the current vector produc-
ing maximum torque at a given speed is given by the cross
point of the current-limit circle and the voltage-limit
ellipse.

D.Transition of Control Mode

The q-axis current command i: is decided from the
C. Control in Constant Power Region speed error (0: - 0,) through the proportional-integral
In order to simplify the control algorithm, the voltage compensator G&s)as shown in Fig. 2 (asterisks designate
constraint is defined by (8) instead of (4): commanded values throughout this paper). The d-axis
current command i$ is decided by (5) in the maximum
torque-per-amp region or by (9) in the flux-weakening
region according to il; and w ( = P, CO,). The transition of
where both control modes is determined according to the
flowchart in Fig. 2. Below the base speed ubase, the
maximum torque-per-amp control is selected because the
voltage V, is always less than V,,. Above the overexcita-
tion threshold speed wc, at which the magnet-generated

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY BOMBAY. Downloaded on June 30,2021 at 17:17:18 UTC from IEEE Xplore. Restrictions apply.
 922 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS,VOL. 30, NO. 4, JULY / AUGUST 1994

I,------------------- -I-,

e- mbosc

U
r---
I
- - - - : decoupling feedforward compensation
Fig. 3. Block diagram of decoupling current controllers.

compensators Gcd(s)and G,,(s) and the decoupling feed-

forward compensation:

Flux-Weakening Maximum (10)

Control Torque-per-Amp
Control
The d- and q-axis current control loops can be linearized
by the above decoupling current control.
Fig. 2. Control block diagram of current vector including transition
flow of control modes.
B. Voltage Command Compensation
In the flux-weakening region, V, is kept equal to V,,
back-EMF w+u ( = P,, w, equals V,,, the flux-weaken- and the terminal voltage V , is nearly the limited voltage
ing control is selected because the maximum torque-per- 5,; as a result, the voltage command vector U * ( U $ , U : )
amp trajectory is outside the voltage-limit ellipse (see Fig. sometimes exceeds the available output voltage of the
1). In the speed region from wbase to w,, the control inverter in transient operations. In this case, the d- and
mode is determined by the calculated voltage V,,, where q-axis current regulators are saturated and the d- and
V,, is calculated by substituting i: and izl given by ( 5 ) in q-axis current controllers affect each other. Therefore, the
(8). If V,, s V,,, then the maximum torque-per-amp con- responses of currents become worse, and sometimes the
trol is selected, or else the flux-weakening control is actual currents cannot follow the commanded currents.
selected. The voltage command compensator compensates the
and u t so that id is controlled in preference to i, based
on the algorithm shown in Fig. 4 [9]. When the voltage
111. HIGH-PERFORMANCE REGULATOR
CURRENT
command vector U * initially lies outside the voltage-limit
A. Decoupling Current Control circle in the Lid-u, plane, which sometimes occurs at the
step change of i: and/or i:, U* is forced back down on
The d - and q-axis currents cannot be controlled inde- the voltage-limit circle (U: in Fig. 5). For example, the
pendently by u d and uq because of the cross-coupling compensated voltage command vector in Fig. 5 is
effects such as wLdid and wL,iq as shown in (1). The determined from UT
through the pass A N -+ BY in Fig. 4
cross-coupling effects of the IPM motor are dominant and U$ is determined from U; through the pass AN .+
because the IPM motor has relatively large inductance. BN.
These effects increase as the speed increases. Therefore,
it seems that the current responses as well as torque IV. EXPERIMENTAL
RESULTS
response are affected by these effects in the high-speed
flux-weakening region. The cross-coupling effects can be A. Laboratory IPM Motor Driue System
cancelled by the feedforward compensation as shown in The block diagram of the laboratory IPM motor drive
Fig. 3. The voltage commands are decided by the PI system is shown in Fig. 6, in which one digital signal

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY BOMBAY. Downloaded on June 30,2021 at 17:17:18 UTC from IEEE Xplore. Restrictions apply.
 MORIMOTO et al.: WIDE-SPEED OPERATION OF IPM MOTORS 923

Vdc= Vd

v q c = vq

Vdc=

v q c = vqo

Vdc , V q c

Fig. 4. Flowchart of voltage command compensation algorithm.

processor (TMS320C25) is used to execute the foregoing

control algorithm. The d- and q-axis current commands
I
are determined by the speed control processor every 400 I

O
ps, in which the algorithm described in Section I1 is
executed. The three phase voltage commands are deter- Fig. 5. Diagram illustrating action of voltage command compensation
mined by the current control processor every 100 p s , in (U: -+ L'c*1, U; -+ &).
which the algorithm described in Section I11 and the
vector rotator ( d , q % U , U , w ) are executed. The specifi-
cations of the laboratory IPM motor drive system are IPM motor, the constant power operation can be achieved
shown in Table I. In the experimentation, the maximum from 1200 to over 3000 rpm.
voltage V,, and the base speed @base are set to 50 V and
1200 rpm, respectively. This is due to the constraint of the C. Transient Characteristics
experimental equipment such as the torque meter and the
load equipment. Therefore, K, and wbase are three or The effects of the proposed voltage command compen-
four times larger in the practical applications. sation in the flux-weakening region are shown in Fig. 8.
The d- and q-axis currents controlled by the conventional
current regulator cannot follow the current commands
B. Steady-State Characteristics
because of the current regulator saturation. The current
The experimental torque versus speed characteristics responses as well as the speed response are greatly im-
taking both of the constraints given by (3) and (4) into proved by the proposed voltage command compensation.
account are shown in Fig. 7. In the conventional id = 0 As the control performance of the d-axis current is very
control method, in which the d-axis current is always kept important in the flux-weakening region, the proposed
at zero, the terminal voltage reaches the limited voltage voltage command compensator is very useful in improving
V,, at 1000 rpm. The torque decreases extremely above the performance of the current control as well as the
1000 rpm and becomes zero at the overexcitation thresh- flux-weakening control.
old speed w, ( = 2200 rpm). In the proposed control Fig. 9 shows the step responses of speed which includes
method, the terminal voltage reaches V,, at 1200 rpm transition from the maximum torque-per-amp control
( = base speed mbase). The maximum torque in the maxi- mode to the flux-weakening control mode. The current
mum torque-per-amp control is about 40% larger than vector is controlled to produce the maximum torque T,,
that in the i, = 0 control below the base speed (i.e., in the considering the current constraint [point A in Fig. 9(b)],
constant torque region). Above the base speed, the oper- and the motor is accelerated by the maximum torque
ating limits are greatly expanded by the flux-weakening below the base speed. The smooth transition between the
control. In the experimental drive system of the prototype two control modes occurs when the rotor speed exceeds

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY BOMBAY. Downloaded on June 30,2021 at 17:17:18 UTC from IEEE Xplore. Restrictions apply.
 924 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 30, NO. 4, JULY /AUGUST 1994

Fig. 6. Block diagram of laboratory IPM motor drive system.

TABLE I ........................................................................
.. .. .. .. .. .. .. .. .. ..
OF LABORATORY
SPECIF'ICATIONS IPM MOTORDRIVESYSTEM

number of pole pairs pn 2

. .... . ... . ... . ..: I .4 ;. . .. .. .;.
armature resistance R 0.57Q ... I
.
. . .. .. .
f

magnet flux-linkage (I, O.IO8Wb

d-axis inductance Ld 8.72mH
q-axis inductance L, 22.8mH
maximum voltage vam 50v
maximum current I,, 8.66A
base meed abase 1200rpm

... ... ... .. .. .. . . .. ..

.. . . . . . .. . .
... .. .. ... ... ... ...
...................................................................... ...
Time t (50ms/div)
(a)
I I I I I
-0- : id=O control method .......................................................................

3 .- -+: Proposed control method

(Maximum torque-per-amp control
plus Flux -w eake n ing Control)
-

!
n " " . . _ ' ' '
. .. .. .. .'. .. .. .. .~. .
E "
. .
'
.
'
. .
' '
. ......................
. . . ....:.
' ~

.....:.
. . .
:....
.. .
. . .
g.2-
h n
0
3
z -7
c-"
1- !

I
I ..... ............... ......

bose
. . . . . .
....................................................................
. .
I
4000 Time t (50ms/div)
OO 1000 2000 3000 (b)
Rotorspeed U, (rpm) Fig. 8. Step responses of speed with and without voltage command
compensation (w: : 2200 + 2600 rpm). (a) Without voltage command
Fig. 7. Torque versus speed characteristics. compensation. (b) With voltage command compensation.

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY BOMBAY. Downloaded on June 30,2021 at 17:17:18 UTC from IEEE Xplore. Restrictions apply.
 MORIMOTO et al.: WIDE-SPEED OPERATION OF IPM MOTORS 925

Transition of control modes Maximum

I Torque-per-Amp
Maximum
Torque-per-Amp Flux-Weakening Trajectory
‘9 Control Control
....................................................................

., . .
.% i ....................................... I. .......
.. .. . \ .
d
d .
.
.
.
.
.
.
.
.
.
..
.
..
.
..
.
2 ; j i. ‘ \ t !
c 0- . . . . . . .
....... ............... .\,,\\................
:
..... :....... i .......
...... :
.......

....... : _ ..... .......

.% :
xi ‘increasing load i
‘:I ....................................
O - 1
. . ( 1.- . ...............................................
d-axis current id (2Ndiv)
. . . . . . . . . . .. . . . . . .. . . . . .
Fig. 10. Current vector trajectory for ramp load disturbance.
. .
. . . . . . . . . . . . . . . . .. . . . .

Time t (200ms/div) speed in this paper. The speed control system was estab-
(a)
lished based on DSP, and the drive tests were carried out
with respect to the prototype IPM motor. The results of
this investigation are summarized as follows.
1) The maximum available torque can be produced by
the maximum torque-per-amp control below the base
speed, and constant power operation can be achieved by
the flux-weakening control above the base speed. As a
result, the operating limits are greatly expanded, and the
maximum available output power is obtained at any speed
considering both current and voltage constraints.
2) The current regulators are sometimes saturated in
transient operations, especially in the flux-weakening re-
;. 10 gion. The decoupling current control and the voltage
............................................
d-axis current id (2Ndiv) command compensation are very useful in improving the
performance of the current vector control and the flux-
(b) weakening control.
Fig. 9. Step responses of speed from 0 to 2500 rpm. (a) Transient 3) The ability of the proposed drive system has been
responses of rotor speed and currents. (b) Current vector trajectory. verified by several experimental results with respect to the
laboratory IPM motor drive, which include the speed
the base speed. Above the base speed, the current vector responses, the current vector responses, and the transition
is controlled to follow the cross point of the current-limit between two control modes.
circle and the voltage-limit ellipse, and as a result, the
motor is accelerated by the maximum available torque at REFERENCES
the operated speed. The current vector moves along the B. Sneyers, D. W. Novotny, and T. A. Lipo, “Field-weakening in
buried permanent magnet ac motor drives,” ZEEE Trans. Znd.
voltage-limit ellipse as the rotor speed reaches the com- Appl., vol. IA-21, pp. 398-407, Mar./Apr. 1985.
manded speed. T. M. Jahns, “Flux-weakening regime operation of an interior
The current vector trajectory for the ramp load distur- permanent-magnet synchronous motor drive,” IEEE Trans. Ind.
Appl., vol. IA-23, pp. 681-689, July/Aug. 1987.
bance at w: = 1800 rpm is shown in Fig. 10. The current B. K. Bose, “A high-performance inverter-fed drive system of an
vector moves along the maximum torque-per-amp trajec- interior permanent magnet synchronous machine,” IEEE Trans.
tory, and smoothly shifts to the voltage-limit ellipse as the Ind. Appl., vol. 24, pp. 987-997, Nov./Dec. 1988.
S. R. Macminn and T. M. Jahns, “Control techniques for improved
load increases. In this case, the actual motor speed is kept high-speed performance of interior PM synchronous motor drives,”
equal to the commanded speed of 1800 rpm by the speed ZEEE Trans. Ind. Appl., vol. 27, pp. 997-1004, Sept./Oct. 1991.
controller, and as a result, the motor torque follows the R. F. Schiferl and T. A. Lipo, “Power capability of salient pole
permanent magnet synchronous motors in variable speed drive
load torque-time ramp. applications,” ZEEE Trans. Ind. Appl., vol. 26, pp. 115-123,
Jan./Feb. 1990.
V. CONCLUSIONS S. Morimoto, Y. Takeda, T. Hirasa, and K. Taniguchi, “Expansion
of operating limits for permanent magnet motor by current vector
The variable speed drive system of the IPM motor was control considering inverter capacity,” ZEEE Trans. Ind. Appl., vol.
proposed for constant power operation over the base 26, pp. 866-871, Sept./Oct. 1990.

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY BOMBAY. Downloaded on June 30,2021 at 17:17:18 UTC from IEEE Xplore. Restrictions apply.
 926 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 30, NO. 4, JULY / AUGUST 1994

[7] T. M. Jahns, G. B. Kliman, and T. W. Neumann, “Interior perma- Masayuki Sanada (M94) was bom in Japan on
nent-magnet synchronous motors for adjustable-speed drives,” IEEE June 1, 1966. He received the B.E., M.E., and
Trans. Ind. Appl., vol. IA-22, pp. 738-747, July/Aug. 1986. Ph.D. degrees from the University of Osaka
[8] S . Morimoto, K. Hatanaka, Y. Tong, Y. Takeda, and T. Hirasa, Prefecture, Japan, in 1989, 1991, and 1994, re-
“High performance servo drive system of salient pole permanent spectively.
magnet synchronous motor,” in Proc. IEEE IAS Annu. Meet., 1991, He joined the Department of Electrical and
pp. 463-468. Electronic Systems, University of Osaka Prefec-
[9] S . Morimoto, Y. Takeda, and T. Hirasa, “Flux-weakening control ture, in 1994, where he is presently a Research
method for surface permanent magnet synchronous motors,” in Associate. His main areas of research interest
P ~ o c I. P E C - T O ~ ~ O ’ W1990,
, pp. 942-949. are linear motors for direct drive applications
and their control systems.
Dr. Sanada is a member of the Institute of Electrical Engineers of
Japan and the Japan Society for Power Electronics.

Shigeo Morimoto (M’93) was bom in Japan on

June 28, 1959. He received the B.E., M.E., and Yoji Takeda (M’93) was bom in Osaka, Japan,
Ph.D. degrees from the University of Osaka on November 10, 1943. He received the B.E.,
Prefecture, Japan, in 1982, 1984, and 1990, re- M.E., and Ph.D. degrees from the University of
spectively. Osaka Prefecture, Japan, in 1966, 1968, and
He joined Mitsubishi Electric Corporation, 1977, respectively.
Tokyo, Japan, in 1984. Since 1988 he has been Since 1968 he has been with the Department
with the Department of Electrical and Elec- of Electrical and Electronic Systems, University
tronic Systems, University of Osaka Prefecture, of Osaka Prefecture, where he is presently a
where he is presently an Associate Professor. Professor. His main areas of research interest
He has been engaged in research of inverter are permanent magnet synchronous motors, lin-
systems and ac drive systems. ear motors, and their control systems.
Dr. Morimoto is a member of the Institute of Electrical Engineers of Dr. Takeda is a m8ember of the Institute of Electrical Engineers of
Japan, the Society of Instrumental and Control Engineers of Japan, and Japan, the Institute ol Systems, Control, and Information Engineers, and
the Japan Society for Power Electronics. the Japan Society for Power Electronics.

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY BOMBAY. Downloaded on June 30,2021 at 17:17:18 UTC from IEEE Xplore. Restrictions apply.