Professional Documents
Culture Documents
1, JANUARY 2012
307
I. INTRODUCTION
OWER electronics has deeply changed the use of induction
motors in automotive or automation applications, giving
them the capability of fast torque response and, consequently, a
full control of the drive speed.
When the induction motors are used for applications at high
speed, it is desirable to retain the maximum torque capability
in the field-weakening region. Several papers about this issue
were presented [1][4]. According to these field-weakening algorithms, the optimal flux value of the motor should be updated
by means of lookup tables or explicit expressions containing the
motor parameters and quantities, such as the motor speed, the
motor currents, the dc-link voltage, and the requested torque.
However, the performance of these algorithms is strictly related
to the accuracy by which the parameters are known. A further
problem is represented by the variable value of the leakage
and magnetizing inductances, to which the rotor-flux-oriented
scheme is particularly sensitive [5]. In addition, the drive performance in the high-speed range may depend on the correct determination of the base speed, which is the function of the actual
dc-link voltage and the overload capability. As a consequence,
new methods for compensating the parameter variations and
Manuscript received December 17, 2010; revised February 24, 2011 and
April 22, 2011; accepted May 1, 2011. Date of current version December 16,
2011. Recommended for publication by Associate Editor R. M. Kennel.
The authors are with the Department of Electrical Engineering, University
of Bologna, 40136 Bologna, Italy (e-mail: michele.mengoni3@unibo.it; luca.
zarri2@unibo.it; angelo.tani@unibo.it; giovanni.serra@unibo.it; domenico.
casadei@unibo.it).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TPEL.2011.2156810
308
Fig. 1.
M2
.
Ls Lr
(2)
MENGONI et al.: COMPARISON OF FOUR ROBUST CONTROL SCHEMES FOR FIELD-WEAKENING OPERATION OF INDUCTION MOTORS
clarified. The signals starting from a little circle (O) are user
set points, such as speed reference m , ref , or the maximum
absolute ratings, such as Vs, m ax and Is, m ax . The signals starting
from a little triangle () come from somewhere else in the
control scheme, although the wirings are not shown to keep the
scheme as simple as possible.
The control scheme of Fig. 1 is implemented in a reference
frame that is synchronous with the stator flux vector. The main
control variables are the stator flux magnitude s and the qcomponent of the stator current isq .
In Fig. 1, the speed is adjusted by the PI regulator (a), which
generates the request of torque-producing current isq ,req . The
current reference is tracked on its turn by the PI regulator (d).
Due to the action of the saturation block (g), isq ,ref is limited
in such a way that the stator current magnitude cannot exceed
Is,m ax in Regions I and II. In this case, the maximum value for
isq ,ref depends on the current isd used for the generation of the
flux. The greater is isd , the lower is isq ,m ax . In Region III, the PI
regulator (e) further decreases isq ,m ax until the angle between
the stator and rotor flux vectors is 45 , i.e., the maximum torque
condition is satisfied.
The stator flux command is generated by the PI regulator (b)
on the basis of the voltage request. If this request is greater than
the available voltage, the field-weakening algorithm reduces the
flux; otherwise, the flux is increased, but not beyond its rated
value.
Finally, the switch (s) can create a temporary voltage margin
to enable a fast reaction of the current controller, in order to
improve the transient behavior. If the requested voltage is greater
than the available voltage, i.e., the flux is being decreased, the
switch (s) is closed and the angle s of the reference frame
is modified by adding a small quantity s proportional to
the speed error. As a consequence, this small rotation of the
reference frame, applied to the stator voltage, has the effect of
improving the torque production to the detriment of the flux,
especially, at the beginning of the speed transient [15].
Although this last algorithm has the aim of improving the
behavior of the motor during the speed transients in the fieldweakening speed range, actually, it is not essential for the fieldweakening operation. Hence, for the sake of simplicity, the effects related to the switch (s) have not been considered in this
paper.
It is worth noting that this control scheme does not control
the d-component of the stator current directly. For this reason, if
the response of PI (b) is very fast, the d-component of the motor
may reach a very high peak during the magnetization transient.
It is possible to come to this conclusion by expressing isd as
follows:
isd =
1
Ls
M
r d
s
Lr
(3)
309
To prevent this occurrence, a widely used remedy is to increase the stator flux set point slowly up to the rated value
during the motor startup. Although this solution is very common, it is not the best one, because the optimal slope of the
ramp depends on the motor parameters, and it is not completely
integrated in the normal control scheme.
The solution proposed in Fig. 1, not presented in [15], is to
insert a variable upper bound on the stator flux in block (f). This
bound should be s, rated during the steady-state operation of
the machine, but during the magnetizing transient s, ref should
not overcome the limit value s, lim given by
s,lim =
M
r d + Ls Is,m ax .
Lr
(4)
Under the assumption that the rotor flux varies more slowly than
the other quantities, if s is lower than s, lim , then the current
isd is lower than Is, m ax whatever fast the response of PI (b) is
and, in particular, during the magnetization transient.
Although (4) requires the knowledge of the leakage inductance Ls , the estimation of the rotor flux is not necessary. In
fact, it is possible to find an alternative formulation of s, lim by
solving (3) for r d and substituting its expression in (4). It turns
out that
s,lim = s + Ls (Is,m ax isd ) .
(5)
(6)
2
2
The signal isq , m ax is equal to Is,m
ax isd,ref in Regions I and
II, whereas in Region III, it
decreases until the absolute value of
the vsd is equal to (Vs,m ax / 2). As explained in Section II, this
condition means that, under the assumption that the maximum
310
Fig. 2.
Fig. 3.
MENGONI et al.: COMPARISON OF FOUR ROBUST CONTROL SCHEMES FOR FIELD-WEAKENING OPERATION OF INDUCTION MOTORS
Fig. 4.
311
sd
.
Ls
(7)
The limitation block (b), which works as in Scheme C, ensures the respect of the constraint on the maximum current in
Region II and the maximum torque capability in Region III. To
satisfy the condition = 45 , sq has to be equal to sd ,
whereas the overcoming of the maximum current is prevented
by ensuring that the absolute value of sq , ref is lower than the
quantity sq , available .
The stator flux regulator behaves as a proportional controller,
with some additional terms compensating the stator back EMF
and the voltage drop caused by the stator resistance. The equations of the stator flux regulator can be expressed as follows:
(10)
(11)
D. Control Scheme D
The block diagram of the control Scheme D is shown in Fig. 4.
In this rotor-flux-oriented control scheme, the main control variables are the components of the stator flux vector instead of the
stator current components.
To understand the control principle, it is useful to recall the
main motor equations written in terms of stator flux components
in a rotor-flux-oriented reference frame [20]:
M
Lr dr
+ r =
sd
Rr dt
Ls
T =
3
M
p
r sq .
2 Ls Lr
(8)
(9)
As can be seen, (8) and (9) are quite similar to the corresponding
equations of the traditional field-oriented control based on dq
stator current components. In fact, the rotor flux depends only
on sd , whereas the motor torque is proportional to sq .
According to (9), the torque demand is transformed by the
speed regulator (a) in the request of the q-component of the
stator flux.
sd,ref sd
d
sq ,ref sq
+
q
vsd,req = Rs isd r sq +
312
sd,ref sd
d
(12)
sq ,ref sq
q
(13)
sq ,m ax
.
Ls
(14)
Here, r, m ax is the angular frequency of the rotor flux corresponding to the maximum torque, which is expressed by
r,m ax = r + (signsq )
Rr sq ,m ax |sq |
.
Lr
sd
(15)
(16)
where sd, rated is the rated flux, and sd, lim is given as
sd,lim = Ls (Is,m ax isd ) + sd .
(17)
MENGONI et al.: COMPARISON OF FOUR ROBUST CONTROL SCHEMES FOR FIELD-WEAKENING OPERATION OF INDUCTION MOTORS
313
TABLE I
MOTOR PARAMETERS
TABLE II
REGULATOR PARAMETERS
shows the speed response (at the top) and the corresponding
phase current waveform (at the bottom).
The two intermediate traces of each figure show the waveforms of the main control variables of each control scheme,
i.e., the stator flux and the current isq for Scheme A, the stator
current components for Schemes B and C, and the stator flux
components for Scheme D.
In Figs. 58, the extension of Regions II and III is also represented.
Finally, in Fig. 9, we compare the spectral content of the motor
currents for the four control schemes under a typical operating
condition. The motor torque is 80% of the rated torque, and
the motor speed is 90% of the base speed. As can be seen, the
harmonic content of the currents resulting from Schemes A and
D appears to be higher than that of Schemes B and C.
314
Fig. 9. Experimental results. Spectra of the phase current for the four control
schemes when the motor speed is 90% of the base speed and the motor torque
is 80% of the rated torque. The spectra are normalized with respects to the
fundamental component of the current.
MENGONI et al.: COMPARISON OF FOUR ROBUST CONTROL SCHEMES FOR FIELD-WEAKENING OPERATION OF INDUCTION MOTORS
315
Fig. 14. Behavior of Scheme A during some speed steps from 0% to 300%
of the base speed (500 ms/div). From top to bottom: actual angular speed
(2000 r/min/div), stator flux magnitude (0.25 Wb/div), q-component of the
stator current (20 A/div), and phase current (20 A/div).
316
Fig. 18. Behavior of Scheme A with detuned regulators during some speed
steps from 0% to 150% of the base speed (500 ms/div). From top to bottom: actual angular speed (500 r/min/div), stator flux magnitude (0.25 Wb/div),
q-component of the stator current (20 A/div), and phase current (20 A/div).
Fig. 19. Behavior of Scheme B with detuned regulators during some speed
steps from 0% to 150% of the base speed (500 ms/div). From top to bottom: angular speed (500 r/min/div), d- and q-components of the stator current
(10 A/div), and phase current (20 A/div).
MENGONI et al.: COMPARISON OF FOUR ROBUST CONTROL SCHEMES FOR FIELD-WEAKENING OPERATION OF INDUCTION MOTORS
Fig. 20. Behavior of Scheme C with detuned regulators during some speed
steps from 0% to 150% of the base speed (500 ms/div). From top to bottom: angular speed (500 r/min/div), d- and q-components of the stator current
(10 A/div), and phase current (20 A/div).
317
Fig. 23. Behavior of Scheme A above the base speed after a step in the load
torque (1 s/div). From top to bottom: angular speed (500 r/min/div), stator flux
(0.25 Wb/div), q-component of the stator current (10 A/div), and phase current
(20 A/div).
(18)
Fig. 21. Behavior of Scheme D with detuned regulators during some speed
steps from 0% to 150% of the base speed (500 ms/div). From top to bottom: angular speed (500 r/min/div), d- and q-components of the stator flux (0.25 Wb/div),
and phase current (20 A/div).
Fig. 22. Behavior of Scheme B with detuned regulators (low gain) during a
speed step change from 0% to 700% of the base speed (500 ms/div). From top
to bottom: angular speed (500 r/min/div), d-component of the stator current
(20 A/div), q-component of the stator current (20 A/div), and phase current
(20 A/div).
318
Fig. 24. Behavior of Scheme B above the base speed after a step variation in
the load torque (1 s/div). From top to bottom: angular speed (2000 r/min/div),
d-component of the stator current (10 A/div), q-component of the stator current
(10 A/div), and phase current (20 A/div).
Fig. 25. Behavior of Scheme C above the base speed after a step in the
load torque (1 s/div). From top to bottom: angular speed (500 r/min/div),
d-component of the stator current (10 A/div), q-component of the stator current
(10 A/div), and phase current (20 A/div).
Fig. 26. Behavior of Scheme D above the base speed after a step in the
load torque (1 s/div). From top to bottom: angular speed (500 r/min/div),
d-component of the stator flux (0.25 Wb/div), q-component of the stator flux
(0.25 Wb/div), and phase current (20 A/div).
The problem of the start-up currents may rise when the drive
is turned ON. If the magnetization transient is not specifically
managed, the control system tries to establish the rated flux level
as quickly as possible. If the currents are directly controlled, the
risk of overcurrents is averted. On the contrary, if the main control variables are fluxes, there is not a direct control of the current
amplitudes, and it is necessary to adopt some countermeasures
to avoid overcurrents in Schemes A and D.
In Figs. 2730, we show the behavior of the four control
schemes during the magnetization transient. As can be seen, the
behavior is acceptable for all of them. Schemes A and D inject
into the motor a dc current and show the shortest magnetization transients, but require an additional section of the control
scheme.
MENGONI et al.: COMPARISON OF FOUR ROBUST CONTROL SCHEMES FOR FIELD-WEAKENING OPERATION OF INDUCTION MOTORS
319
TABLE III
COMPARISON OF THE FOUR CONTROL SCHEMES
VI. CONCLUSION
Four control schemes that feature a robust field-weakening
algorithm have been compared. Although the performance is
very much alike, each control scheme presents some advantages
and some disadvantages regarding the complexity of tuning, the
quality of the load currents, the robustness against the parameter uncertainties, and the operation stability, as summarized in
Table III.
The results cannot be generalized, since they depend on the
specific DSP, inverter, and motor used to carry out the experimental tests. Nevertheless, they suggest some practical rules
that can be useful to select which control scheme is the most
suitable for an application.
The control Scheme A should be preferred when the robustness to variations of the motor parameters could be crucial for
the drive performance. Control Schemes B and C should be preferred for a specific application when the quality of the motor
currents plays a key role, or just because the industrial know-how
is mainly related to traditional field-oriented control schemes.
Finally, control Scheme D is preferable when the application
requires a fast torque response in the field-weakening region or
the tuning of the regulators has to be as simple as possible.
REFERENCES
[1] X. Xu and D. W. Novotny, Selection of the flux reference for induction
machine drives in the field-weakening region, IEEE Trans. Ind. Appl.,
vol. 28, no. 6, pp. 13531358, Nov./Dec. 1992.
[2] S. H. Kim and S. K. Sul, Maximum torque control of an induction
machine in the field-weakening region, IEEE Trans. Ind. Appl., vol. 31,
no. 4, pp. 787794, Jul./Aug. 1995.
[3] G. Griva, F. Profumo, M. Abrate, A. Tenconi, and D. Berruti, Wide speed
range DTC drive performance with new flux-weakening control, in Proc.
Conf. Rec. Power Electron. Spec. Conf., May 1722, 1998, Fukuoka,
Japan, pp. 15991604.
[4] D. Casadei, F. Profumo, G. Serra, A. Tani, and L. Zarri, Performance
analysis of a speed-sensorless induction motor drive based on a constantswitching-frequency DTC scheme, IEEE Trans. Ind. Appl., vol. 39, no. 2,
pp. 476484, Mar./Apr. 2003.
[5] H. Grotstollen and J. Wiesing, Torque capability and control of a saturated
induction motor over a wide range of flux weakening, IEEE Trans. Ind.
Electron., vol. 42, no. 4, pp. 374381, Aug. 1995.
[6] R. J. Kerkman, T. M. Rowan, and D. Leggate, Indirect field-oriented
control of an induction motor in the field-weakening region, IEEE Trans.
Ind. Appl., vol. 28, no. 4, pp. 850857, Jul./Aug. 1992.
320
[7] E. Levi and M. Wang, A speed estimator for high performance sensorless
control of induction motors in the field-weakening region, IEEE Trans.
Power Electron., vol. 17, no. 3, pp. 365378, May 2002.
[8] M. S. Zaky, M. M. Khater, S. S. Shokralla, and H. A. Yasin, Widespeed-range estimation with online parameter identification schemes of
sensorless induction motor drives, IEEE Trans. Ind. Electron., vol. 56,
no. 5, pp. 16991707, May 2009.
[9] M. S. Huang, Improved field-weakening control for IFO induction motor, IEEE Trans. Aeronaut. Electric Syst., vol. 39, no. 2, pp. 647658,
Apr. 2003.
[10] M. Ho Shin, D. S. Hyun, and S. B. Cho, Maximum torque control of
stator-flux-oriented induction machine drive in the field-weakening region, IEEE Trans. Ind. Appl., vol. 38, no. 1, pp. 117121, Jan./Feb.
2002.
[11] D. Casadei, G. Serra, A. Tani, and L. Zarri, A Robust method for fluxweakening operation of DTC induction motor drive with on-line estimation of the break-down torque, in Proc. Eur. Conf. Power Electron. Appl.,
Dresden, Germany, Sep. 1114, 2005, pp. 110.
[12] R. Bojoi, P. Guglielmi, and G. Pellegrino, Sensorless stator field-oriented
control for low cost induction motor drives with wide field-weakening
range, in Proc. Ind. Appl. Soc., Edmonton, Canada, Oct. 59, 2008,
pp. 17.
[13] T. S. Kwon, M. H. Shin, and D. S. Hyun, Speed sensorless stator fluxoriented control of induction motor in the field-weakening region using Luenberger observer, IEEE Trans. Power Electron., vol. 20, no. 4,
pp. 864869, Jul. 2005.
[14] A. Biinte, H. Grotstollen, and P. Krafka, Field weakening of induction
motors in a very wide region with regard to parameter uncertainties, in
Proc. Power Electron. Spec. Conf., vol. 1, Jun. 2627, 1996, pp. 944950.
[15] H. Abu-Rub, H. Schmirgel, and J. Holtz, Sensorless control of induction motors for maximum steady-state torque and fast dynamics at field
weakening, presented at Annu. Meeting Industry Applications Society,
Tampa, FL, Oct. 812, 2006.
[16] S. H. Kim and S. K. Sul, Voltage control strategy for maximum torque
operation of an induction machine in the field-weakening region, IEEE
Trans. Ind. Electron., vol. 44, no. 4, pp. 512518, Aug. 1997.
[17] L. Harnefors, K. Pietilainen, and L. Gertmar, Torque-maximizing fieldweakening control: Design, analysis, and parameter selection, IEEE
Trans. Ind. Electron., vol. 48, no. 1, pp. 117122, Feb. 2001.
[18] F. Briz, A. Diez, M. W. Degner, and R. D. Lorenz, Current and flux
regulation in field-weakening operation, IEEE Trans. Ind. Appl., vol. 37,
no. 1, pp. 4250, Jan./Feb. 2001.
[19] D. Casadei, M. Mengoni, G. Serra, A. Tani, and L. Zarri, A control
scheme with energy saving and dc-link overvoltage rejection for induction
motor drives of electric vehicles, IEEE Trans. Ind. Appl., vol. 46, no. 4,
pp. 14361446, Jul./Aug. 2010.
[20] M. Mengoni, L. Zarri, A. Tani, G. Serra, and D. Casadei, Stator flux
vector control of induction motor drives in the field-weakening region,
IEEE Trans. Power Electron., vol. 23, no. 2, pp. 941949, Mar. 2008.
[21] P.-Y. Lin and Y.-S. Lai, Novel voltage trajectory control for fieldweakening operation of induction motor drives, in Proc. Eur. Conf. Cognitive Ergonom., San Jose, CA, Sep. 2024, 2009, pp. 15401546.
[22] D. Casadei, M. Mengoni, G. Serra, A. Tani, and L. Zarri, Field-weakening
control schemes for high-speed drives based on induction motors: A comparison, in Proc. 39th IEEE Annu. Power Electron. Spec. Conf., Rhodes,
Greece, Jun. 1519, 2008, pp. 21592166.
[23] M. Hinkkanen, L. Harnefors, and J. Luomi, Reduced-order flux observers with stator-resistance adaptation for speed-sensorless induction
motor drives, IEEE Trans. Power Electron., vol. 25, no. 5, pp. 1173
1183, May 2010.
[24] F. R. Salmasi, T. A. Najafabadi, and P. J. Maralani, An adaptive flux
observer with online estimation of dc-link voltage and rotor resistance
for VSI-based induction motors, IEEE Trans. Power Electron., vol. 25,
no. 5, pp. 13101319, May 2010.
Michele Mengoni was born in Forl`, Italy, in
1981. He received the M.Sc. and Ph.D. degrees
(Hons.) in electrical engineering from the University of Bologna, Bologna, Italy, in 2006 and 2010,
respectively.
He is currently a Fellow Researcher at the Department of Electrical Engineering, University of
Bologna. His research interests include sensorless
control of induction motors, multiphase drives, and
ac/ac matrix converters.
Domenico Casadei (SM04) received the M.Sc. degree (Hons.) in electrical engineering from the University of Bologna, Bologna, Italy, in 1974.
He joined the Department of Electrical Engineering, University of Bologna, in 1975, as Research
Assistant Professor. He is currently a Full Professor of electrical drives. His scientific work is related
to electrical machines and drives and power electronics. He is author and coauthor of more than 200
scientific papers, published in technical journals and
conference proceedings. His current research interests include vector control of ac drives and diagnosis of electrical machines.
He has been involved in several research projects with the industry in the same
research areas.
Dr. Casadei is a senior member of the IEEE Industrial Electronics Society, a
member of the IEEE Power Electronics Society, and a member of the European
Power Electronics Society.