16th international conference on Sciences and Techniques of Automatic control STA'2015-PID3819-DEM

& computer engineering - STA'2015, Monastir, Tunisia, December 21-23, 2015

Performance Analysis of FOC and DTC for

PMSM Drives using SVPWM Technique
M. Abassi, A. Khlaief, O. Saadaoui, A. Chaari and M. Boussak, Senior Member, IEEE
post

Abstract—the aim of this article is to present the (PMSM) was made possible by the introduction of new
performance of field oriented control (FOC) and direct torque magnets [1], [2], [3]. Two types of controls were used with
control (DTC) of a permanent magnet synchronous motor these motors, which are the field oriented control (FOC) and
(PMSM) using the technique of space vector pulse width the direct torque control (DTC)[4], [5], [6].
modulation (SVPWM). The FOC technique provides a decoupling of the flux and
First, after developing a mathematical model of PMSM we
torque similar to that of a DC motor, where the control
present the principle of space vector modulation technique to
use in commands. Secondly, we developed a field oriented strategy is to guide the park plan to cancel a component of
control and direct torque control strategy. Finally, we proposed the flux to obtain a simplified expression of torque for better
a failover system between direct torque control and field control.
oriented control in the event of default or loss of information. The vector control requires knowledge of the position, which
Keywords— direct torque control (DTC); field oriented typically imposes a mechanical sensor for determining the
control (FOC); permanent magnet synchronous motor (PMSM); speed [7], [8].
space vector pulse width modulation (SVPWM). The DTC technique allows a direct control of the flux and
the electromagnetic torque of PMSM; hence the strategy of
NOMENCLATURE this command is to choose a voltage vector to apply to the
inverter from the error of the flux and the direction of
Vd,Vq stator voltage components in the rotor frame in the rotation of the flux in the Concordia plan. With this control
"d" and "q" axes, respectively strategy, the position is estimated from the stator variables.
id, iq stator current components in the rotor frame in the "d" [9], [10], [11].
and "q" axes, respectively This paper presents a simplified study for these two control
Ld, Lq stator inductance components in the rotor frame in the strategies using the technique of space vector modulation
"d" and "q" axes SVPWM for generating a reference voltage vector. These
ĭf "d" component of the flux induced by the permanent two types of control with a decision system to develop,
magnets of the rotor in the stator phases allows the elaborate of a fault tolerant control of the position
Ke back-emf coefficient constant sensor or the control algorithm [12], [13], [14].
Kt torque constant
Rs resistance of the stator windings II. MODELING OF PERMANENT MAGNET SYNCHRONOUS
J the rotor inertia MOTOR
B viscous friction coefficient The representation of the permanent magnet
N number of the pole pairs of the motor synchronous motor in the stationary and rotating frames is
VDC DC bus voltage shown in the figure below.
I. INTRODUCTION

T HE PERMANENT magnet synchronous motors (PMSM)

are widely used in many industrial applications; such as
machine tools, robotics, aeronautics, automotive..., thanks to
the benefits they offer namely, low maintenance, high torque
to inertia ratio and high performance. This development

M. Abassi, A. Khlaief, O. Saadaoui and A. Chaari was with research

unit on Control, Monitoring and Safety of Systems (C3S) at High National
Engineering School of Tunis (ENSIT), 5 Avenue Taha Hussein – BP 56,
Bab Manara 1008 Tunis, Tunisia – Fax (216) 71 391 166 (e-mail:
[email protected], [email protected], [email protected],
[email protected]).
M. Boussak is with Laboratoire des Sciences de l’Information et des
Systèmes (LSIS), UMR CNRS 7296, Ecole Centrale de Marseille (ECM),
Technopôle de Château Gombert, 13451 Marseille Cedex 20, France (e- Figure 1. Representation of the synchronous motor permanent magnet.
mail: [email protected]).

978-1-4673-9234-1/15/$31.00 ©2015 IEEE 228

Electric and magnetic equations permanent magnet Matrix notation of the inverter output voltages (Va , Vb, Vc) in
synchronous motor in dq frame, are expressed as follows. accordance with the state of switches is given by the
following relationship:
Φ d = Ld .id + Φ f (1)
ªVa º ª 2 − 1 − 1º ª C1 º
Φ q = Lq .iq ( 2) «Vb » = Vdc « − 1 2 − 1» «C 2 » (8)
« » 3 « »« »
¬«Vc ¼» «¬ − 1 − 1 2 »¼ «¬C3 »¼
d
Vd = Rs .id − ω.Lq .iq + Ld . id (3)
dt
d The voltage vector in the inverter output is of the following
Vq = Rs iq + Lq . iq + ω.Ld id + ω.Φ d (4) form:
dt
2π 4π
2 j j
The equivalent circuit of the synchronous motor permanent V s = Vα + jVβ = (Va + Vb e 3 + Vc e 3 ) (9)
magnet is as follows: 3

By exploiting (8) and (9) we can set Table. I which shows

the eight voltage vectors to be applied to the motor at any
time; six of which are active and two draws.
As indicated in Fig. 3, the distribution of voltage vectors in
the Į-ȕ frame forms a hexagon where the amplitude of the
Figure 2. The equivalent circuit of the PMSM in the d-q frame. reference voltage vector is limited by a circle of radius
0.707Vdc .
The expression of the electromagnetic torque of PMSM.

(( )
Te = P Ld − Lq .id .iq + Φ f .iq ) (5)

The mathematical model of PMSM in the d-q frame is

given by the following form:

ªdid º ª R L º ª1 º
« dt » « − s ω d 0 0» 0 0»
Ld Lq «
« di » « »ªid º « Ld »ªv º
« q» « Ld R φf »« » 1 d
« dt » = « −ω − s − 0»«iq » « 0 0 »« » (6)
L L Lq +« Lq »«vq »
« dω » « q q »«ω» « »
« » « P2 2 »« » « P «T »
( )
« dt » « Ld − Lq iq
P
φf −
f
0»¬θ ¼ « 0 0 − »¬ l ¼
dθ
« » «J J J J»
»
¬« dt ¼» ¬« 0 0 1 0¼» ¬« 0 0 0 ¼»

III. SVPWM TECHNIQUE

This technique is based on two principles; the reference
signal is sampled at regular time intervals and the average
value of the voltages is equal to the value of the reference
voltage at each sampling instant. Figure 3. The distribution of votage vector in the Į-ȕ frame
So this technique is to approach three instantaneous voltages
through action on the control signals of the switches as The SVPWM consists in approaching two voltages
possible. The vector control voltage is generally calculated instantaneous by switch signals control [6]. As indicated by
and approximated over modulation by means of voltage Fig.3, the objective is to find the good combination of switch
vector. control. In order to reconstitute by stage the shape of the
The switches of each branch of the three-phase inverter sinusoidal signals, for each period of commutation. For one
are complementary; it is the same for related signals. period of commutation Tc, the reference voltage vector Vref is
in a sector formed by Vi and Vi+1, as Fig.3 indicates it. So
­C 4 = 1 − C1 that the average value on Tc to be equal to Vref, we can use
° (7 ) the following relation:
®C5 = 1 − C 2
°C = 1 − C
¯ 6 3 Tc = tiVi + ti +1Vi +1 (10)

229
 IV. FIELD ORIENTED CONTROL 4
iq
id

Current id and iq (A)

A. Principle of command 3
The principle of field-oriented control of a PMSM by
rotor flux orientation is to guide the flux along the rotor axis 2
(d), this strategy is to maintain the axis (d) constantly aligned
with the flux vector of the magnet. The reference for the 1
current (id) is maintained at zero. The reference for the
current (iq) is determined by the intermediary of one speed 0
control loops (IP).
The diagram of FOC control applied to PMSM provided -1
0 0.05 0.1 0.15 0.2
with a three-phase inverter with PWM is given in Fig. 4. Time (s)
Figure 7. Quadrature current iq and id

4 ia
ib
ic

Stator current (A)

2

θ 0
³

-2
Figure 4. Structure of field oriented control with space vector modulation.

B. Simulation results -4
0 0.05 0.1 0.15 0.2
The mathematical model of PMSM is established under Time (s)
Simulink and simulation results of FOC with space vector
modulation applied to motor are presented. The parameters Figure 8. The stator currents
of the PMSM are given in appendix. The reference torque
signal is given by: (0, 4) N.m at (0, 0.08) s, respectively.
2.5
2500
w
w ref
2000 2
Speed (rpm)

Tem (N .m)

1500 1.5

1000
1
500
0.5
0
0 0.05 0.1 0.15 0.2
Time (s) 0
0 50 100 150 200
Figure 5. The response speed
Speed (rad/s)

25
Electromagnetic torque (m.N)

Figure 9. PMSM speed-torque characteristics

20

15
The simulation results are obtained with a speed step of 2000
10 rpm; Fig. 5 shows the evolution of the rotor speed before and
5 after applying a torque load at 0.08 s, Fig. 6 illustrates the
response of the electromagnetic torque which remark low
0
torque oscillation due to the space vector modulation
-5
technique, Fig. 7 shows that the current iq is also proportional
-10 to electromagnetic torque also the current id is maintained at
0 0.05 0.1 0.15 0.2
Time (s) zero, Fig. 8 shows the stator current response depending on
the evolution of the load and the Fig. 9 illustrates the
Figure 6. The response of the electromagnetic torque

230
behavior of the mechanical characteristic of the motor 2500
without load.
2000
V. DIRECT TORQUE CONTROL

Speed (rpm)
1500
A. Principle of command
The diagram of DTC control applied to PMSM 1000
provided with a three-phase inverter with space vector
modulation proposed in Fig. 10, two Proportional-Integral 500
(PI) controllers are used to determine a reference voltage
vector in the Į-ȕ plane. 0
0 0.05 0.1 0.15 0.2
Time (s)
Figure 12. The speed response

4 ia
ib
ic
2

Stator current (A)

0

-2

-4
0 0.05 0.1 0.15 0.2
Figure 10. Structure of direct torque control with space vector modulation Time (s)
Figure 13. The stator current
The flux and torque estimator in the Į-ȕ plane.
Φ s (α , β ) = ³ (Vs(α , β ) − Rs .is (α , β ) ).dt (10)
4 iq
Te = N p (Φα .iβ − Φ β .iα ) (11)
id
The proposed DTC technique requires the elimination
Current id and iq (A)

3
hysteresis controllers and the switching table of conventional
DTC, with this method the inverter is operating with a fixed 2
frequency due to the vector modulation applied to the vectors
outputs. 1

B. Simulation results 0
Some simulations were carried out for the proposed DTC
Space Vector Modulation technique by the simulation -1
0 0.05 0.1 0.15 0.2
software Matlab Simulink. The motor is fed by an ideal Time (s)
voltage source. The reference flux signal is kept constant at
0.2 Web. Figure 14. Quadrature current id and iq
15 Te ref
Electromagnetic torque (m.N)

Te 2.5
10
2

5
Tem (N.m)

1.5

0
1

-5 0.5

-10 0
0 0.05 0.1 0.15 0.2 0 50 100 150 200
Speed (Rad/s)
Time (s)
Figure 11. The response of the électromagnetic torque
Figure 15. PMSM speed-torque characteristics

231
 The simulation results obtained with reference torque APPENDIX
signal are given by: (0, 4) N.m at (0, 0.08) s, respectively.
Fig. 12 shows the evolution of the rotor speed before and TABLE I. SWITCHING SEQUENCE FOR EACH VECTOR SPACE
after applying a torque load at 0.08 s, Fig. 13 shows that
current iq is also proportional to electromagnetic torque, Fig.
_ ANGLE C1 C2 C3 VI
Vs = Vα + jVβ (°)
11 shows the stator current response depending on the
evolution of the reference torque and Fig. 9 illustrates the 0 0 0 0 0 V0
behavior of the mechanical characteristic of the motor
without load. 0.81Vdc 0 1 0 0 V1

0.81Vdc (0.5 + j 0.86 ) π

1 1
0 V2
VI. TRANSITION BETWEEN THE DTC AND FOC CONTROLS 3
The results of the validation of the two control strategies
0.81Vdc (− 0.5 + j 0.86) 2π 0 V3
for a permanent magnet synchronous motor 0 1
(field-oriented control and direct torque control) are much 3
closer to each other.
If a fault arises at one of the control algorithms and to ensure −0.81Vdc π V4
0 1 1
continuity of operation and without losing system
performance, we propose a principle diagram (shown in 0.81Vdc (− 0.5 − j 0.86) 4π V5
0 0 1
Fig.16) that can detect the fault and thereafter set transition 3
from FOC to DTC and vice versa.
0.81Vdc (0.5 − j 0.86) 3π V6
1 0 1
2

0 0 1 1 1 V7
TABLE II. SWITCHING SEQUENCE FOR EACH VECTOR SPACE

Ld =Lq 53.24 mH
ĭf 0.249 Wb
Ke 0.565 V.s.rad-1
Kt 0.9149 Nm/A
Rs 2ȍ
J 0.0036 Kg.m2
B 0.0011 Nm.s.rad-1
N 3
VDC 540 V

Figure 16. Structure of a default failover system.

REFERENCES
[1] O. Saadaoui, A. Khlaief, A. Chaari, M. Boussak, “A new approach
VII. CONCLUSION rotor speed estimation for PMSM on sliding mode observer,” Journal
of Automation & Systems Engineering (JASE), vol. 9, pp. 66-78,
During this study, we presented two control strategies 2015.
for permanent magnet synchronous motors, using the space [2] A. Khlaief, “Contribution à la commande vectorielle sans capteur
vector modulation technique, to know the field-oriented mécanique des machines synchrones à aimants permanents (MSAP), ”
control and direct torque control. Doctoral thesis, Aix Marseille university, High School of Sciences
and Techniques of Tunis, July. 2012.
The field-oriented control requires knowledge of the position, [3] F. Grouz, “Diagnostic et isolation des défauts d’un actionneur
which typically requires a mechanical sensor for determining synchrone à aimants permanents, ” Doctoral thesis, National
the speed; however, the direct torque control is based on the engineering school of Gabes, march 2013.
estimate of the flux and the electromagnetic torque through [4] A. Kronberg, “Design and simulation of field oriented control and
irect torque control for a permanent magnet synchronous motor with
magnitudes power of PMSM. positive saliency,” Thesis for M.S. degree, Uppsala university, Maj
The performances obtained by these commands are similar 2012.
strategies including speed, torque, current... and they are used [5] M. Sebba, S. Hassaine, S. Moreau, A. Chaker, “ Analyse et synthèse
as a means to develop a system of decision that operates in d’une structure de contrôle vectoriel simplifiée associée au couple de
charge appliquée au moteur synchrone à aimant permanents, ” Acta
degraded motor (fault of the speed sensor). Therefore, future Electrotechnica, vol. 50, no. 2, 2009.
work will focus on experimental validation using test bench [6] G. Chandaka and G. Prasanth, “Direct torque control and field
based on the DS1104 card from dSpace to verify the oriented control of PMSM using SVPWM Technique”, IJARSE, vol.
proposed method. No.3, Issue No.11, Novembre 2014.

232
[7] A. Khlaief, M. Abassi, M. Boussak, M. Gossa, “ DSP based SVPWM
technique for field oriented speed control of permanent magnet
synchronous motor drive,” International Conference STA’2009.
[8] A. Khlaief, M. Boussak, M. Jemli, M. Gossa, “Field-oriented control
of interior permanent magnet synchronous motor with a space vector
modulation” Conference international STA’2008, Tunisia.
[9] M. E. Haque, L. Zhong, and M. F. Rahman, “A Sensorless Initial
Rotor Position Estimation Scheme for a Direct Torque Controlled
Interior Permanent Magnet Synchronous Motor Drive” IEEE Trans.
Power Electron., vol. 18, no. 6, pp. 1376–1383, Nov. 2003,
[10] M. Abassi, A. Khlaeif, M. Jemli, M. Boussak, M. Gossa: “Direct
Torque Control of permanent magnet synchronous motor drive,”
International conference STA’2008, Tunisia.
[11] L. Tang, L. Zhong, M. F. Rahman, Y. Hu, “A Novel Direct Torque
Control for Interior Permanent Magnet Synchronous Machine Drive
System with Low Ripple in Torque and Flux-A Speed Sensorless
Approach” in Conf. Rec IEEE-IAS Annual meeting, vol. 1, pp.104–
111, 2002.
[12] M.S. Merzoug and F. Naceri “Comparison of field-oriented control
and direct torque control for permanent magnet synchronous motor
(PMSM)”, IJECEECE vol:2, No: 9, 2008.
[13] C. Chakraborty, V. Verma, “Speed and Current Sensor Fault Detection
and Isolation Technique for Induction Motor Drive Using Axes
Transformation”, IEEE Trans. Ind. Electron., vol. 62, no. 3, Mar., pp.
1943–1954, Mar. 2015.
[14] S. Benaicha, “Contribution à la commande tolérante aux défauts d’un
système à motorisation asynchrone : Apport de l’intelligence
artificielle pour l’aide à la supervision et à la décision”, Doctoral
thesis, Batna university, Faculty of engineering Sciences, march
2010.

233