Proceedings of the 3rd International Conference on WeAB.

1
Systems and Control, Algiers, Algeria,
October 29-31, 2013

Real-Time Implementation in DSPACE of Intelligent Vector Control

of Induction Motor
Souad RAFA, Najib ESSOUNBOULI, Abdelkader LARABI, Malik MANCEUR, Abdelaziz
HAMZAOUI, , Linda BARAZANE.

Abstract— The induction motor has a multi-variable model, overcome the assumption on the machine parameters
nonlinear and strongly-coupled. Due to parameters variation knowledge, an adaptive fuzzy system is exploited to
during operation of induction motor, the decoupling approximate the unknown dynamics, which allows
implemented by field oriented control is destroyed. For that, a considering the advantages of direct and indirect adaptive
novel method named as Fuzzy vector control for induction
motor is proposed. This technique is used to replace the block of
schemes. Nevertheless, this structure requires calculation
Field oriented control with a new block control its role is to time and measuring currents which increases the
maintain the decoupling and overcome the problem of implementation costs.
robustness with respect to the parametric variations. The Another approach applies in the control field using sliding
simulation results show that the two control schemes provide in mode control to deal to the uncertainties; this approach is
their basic configuration, comparable performances regarding powerful to control nonlinear and uncertain systems [10-12].
the decoupling. However, the choice of the Fuzzy vector control
appears advantageous from the point of view robustness. The
However, it is difficult to obtain the bound of lumped
fuzzy vector control scheme is successfully implemented in real- uncertainty in advance for practical applications of SMC.
time using a digital signal processor board DS 1104. The Besides, the disadvantage exists in the application of SMC
efficacy of this technique is verified as well as experimentally at system is the undesired chattering phenomena which raises
different dynamic operating conditions such as parameter imperfection for practical switching device using.
variations and the change of speed. The fuzzy vector control is To overcome all these disadvantages, we propose in his
found to be a robust control for application in an induction paper, a new approach based on fuzzy logic to control the
motor.
induction motor with keeping the classical controllers speed.
Indeed, this control retains ownership of the rotor flux
I. INTRODUCTION orientation without using of the park transformation, which
The vector control method is very effective approach for allows us to eliminate the PI and current sensors and make its
high-performance control of an induction motor [1-4]. By implementation cheaper.
using vector control, we achieve decoupled control of the For this control we impose the rotor flux and we use a
rotor flux and the torque, with a fast torque response. Usually single loop speed to control the motor, the control voltages
in most industries, the motor control issues are traditionally and the stator pulsation are deduced from the fuzzy rules
handled by fixed-gain proportional-integral (PI) and taking account parametric variations.
proportional integral derivative (PID) controllers. This control strategy has been performed on a typical
However, this structure cannot deal with parametric induction motor with 1.5 kW rating, with parameter reported
variations and external disturbances. Hence, it is necessary to in the APPENDIX. The organization of the paper is as
synthesize a robust controller to reject their effect. Many follows: in section II, the vector control theory and the
nonlinear control methods have been developed for the propose approach using the fuzzy logic are presented. In
induction motor drive system, such as variable structure section III, the simulation results are presented. In section
control [5-7], adaptive and robust control [8-9], sliding mode IV, the implementation of the propose approach is shown.
control (SMC) [10-12], backstepping [13-15] and so on. Finally, concluding remarks are given in section VII.
For example in [14], the authors propose a backstepping II. THE MODEL OF THE FUZZY VECTOR CONTROL
fuzzy adaptive controller, where they use backstepping to
control the speed and the current in the same loop. To The mathematical model of the induction motor in the
reference frame linked to the rotating field is obtained by
*Resrach supported by Agence Universitaire de la Francophonie (AUF). considering the components of voltage (Vds, Vqs) as control
S. RAFA is with Centre de Développement des Energies Renouvelables, variables and (Ids, Iqs, dr, qr, ) as state variables. This
Algiers, 16000 Algeria (corresponding author to provide phone:
model is presented as follows [16-19]:
+213550247147; fax: +21321901560; e-mail: s.rafa@ cder.dz).
M. Manceur, N. Essounbouli and A. Hamzaoui are with University of
Reims Champagne Ardenne, CReSTIC, IUT of Troyes, Troyes, France.
A. Larabi and L. Barazane are with the University of Science and
Technology Houari Boumediene, Laboratory of the Industrial Electrical
System, Algiers, Algeria.

978-1-4799-0275-0/13/$31.00 ©2013 IEEE

Proceedings of the 3rd International Conference on WeAB.1

dI ds k 1 speed regulation and fast dynamic response.

 I ds  s I qs   dr  pk qr  V
dt Tr Ls ds However, this control requires two current sensors, which
dI qs k 1 make it more costly. Moreover, it’s highly parameter
 I qs  s I ds   qr  pk dr  V variation sensitive due to the presence of the rotor time
dt Tr Ls qs (1)
constant. To solve these problems, we propose to replace the
d ds Lm 1 block of FOC with a new simple model that encompassing
 I ds   dr  (s  p) qr
dt Tr Tr all the uncertainties of the motor, using the type-1 fuzzy
d qs Lm 1 logic. The type-1 Fuzzy logic is applied to the vector control
 I qs   qs  (s  p) dr
dt Tr Tr by replacing the block of FOC by a fuzzy vector control
d pLm block taking account of the parametric variation [20].
  dr I qs   qs I dr   Cr
dt JLr J
The dynamic behavior of an induction motor is complex
due to the coupling effect between the stator and rotor
phases. To render this dynamic similar to the DC motor a
vector control is required. This technique is used to vary the
speed of an induction motor over a wide range. It was
initially developed by Blaschke [1]. The main objective of
the vector control or field oriented control of induction
motors consists in the control of the torque and the flux
independently. To achieve field orientation, the flux (d axis) Figure 2. The new fuzzy vector control.
component of stator current Ids is aligned in the direction of
This new model (Figure. 3) has a type-1 fuzzy control
the rotor flux and the torque component of current Iqs is
algorithm which updates the values of the stator voltage and
aligned in the direction perpendicular to it, at this condition
the synchronous pulse. This new block has four inputs and
[1-4].
three outputs. The inputs are the electromagnetic torque, the
 dr  r and  qr  0 (2) derivative of electromagnetic torque, the rotor flux and rotor
With vector control the dynamic equations of stator current speed, with each of these inputs corresponding to a fuzzy
components and rotor flux are given by (1): variable. The output is the fuzzy control decision [20].
dI ds k 1
 I ds  s I qs   r  V
dt Tr Ls ds
dI qs 1
 I qs  s I ds  pk r   V
dt Ls qs (3)
d r Lm 1
 I ds   r
dt Tr Tr
d pLm C
  r .I qs  r Figure 3. Block diagram of the fuzzy control replaces the vector control.
dt JLr J
The vector control block is shown in Fig 1. 1

0.8
Membership

0.6

0.4

0.2

0
-6 -4 -2 0 2 4 6
Torque Fuzzy sets

0.8

Figure 1. Block diagram of the vector control.

Membership

0.6

0.4

Essentially, the vector control is reformulated to resemble 0.2

the control of a DC motor. This control offers a number of 0

0.2 0.25 0.3 0.35 0.4
benefits including speed control over a wide range, precise Flux Fuzzy sets

978-1-4799-0275-0/13/$31.00 ©2013 IEEE

Proceedings of the 3rd International Conference on WeAB.1

~
~ dC ~ ~ ~ ~ ~
1

0.8 With Ce , e ,  r ,  and Vds ,Vqs ,  s are fuzzy

dt
Membership

0.6
variables corresponding to the input variables
dC
0.4

0.2 Ce , e ,  r ,  and to the control variables Vds ,Vqs , s

0
dt
-300 -200 -100 0 100
Speed Fuzzy sets
200 300
respectively.
In this case, the numbers of fuzzy set of control variables
~ ~ ~
1
Vds ,Vqs ,  s are equal to M1*M3*M4, M1*M2*M3*M4 and
0.8

M1*M3*M4 respectively.
Membership

0.6

M1, M2, M3, M4 are the number of fuzzy set of the input
0.4
~
~ dCe ~ ~
variables Ce , ,  r ,  respectively.
0.2

0
-6000 -4000 -2000 0 2000
Derivative Torque Fuzzy sets
4000 6000 dt
The rules are then presented as follows:
Figure 4. Input fuzzy sets. 
~ ~ ~
 
Vkds1,...,k 4  : IF Ce ,  r ,  is FCek1 , Fkr3 , Fk 4 THEN
The expression of the vector control is written as follows: Vds is FVdsk1, k 3, k 4 
y  f x  (4)
~
 ~ dCe ~ ~ 
Where: Vkqs1,...,k 4  :  Ce ,
 IF ,  r ,   is
t  dt 
d r
y  Vds ,Vqs , s  ; x   Ce , e ,  r ,
 dC 
t

 dt dt
,

FCe , FdCeFr , F  THEN Vqs is FVqs
k1 k2 k3 k4 k1, k 2, k 3, k 4 

Considering that the flux of reference r is constant, this  

ks1,...,k 4  : IF Ce ,  r ,  is FCek1 , Fkr3 , Fk 4  THEN
~ ~ ~
means that:
 s is Fk1, k 3, k 4 
d r s
0 The defuzzification is calculated according to the method
dt (5) of the weighted average:
The new input variable is given as follows:
    fi
f i  k1 k 2 k 3 k 4
( k 1,k 2,k 3,k 4 )
 
.T  xk11 ,  xk22 ,  xk33 ,  xk44
(9)
 
t
 dC  k1 k 2 k 3 k 4 .T  xk11,  xk22 ,  xk33 ,  xk44
x   Ce , e ,  r ,   (6)
 dt  f i : represents one of the output variables, i = 1,…., 4.
( k 1, k 2, k 3, k 4 )
The system of equations which characterize the model is fi is the representative value of the fuzzy sets.
given by:
 
III. SIMULATION RESULTS
R L L  L C L L  L  C
2 2
R
f1 x   Vds  s  r  r s r 2 m e 3  s r m  e In order to verify the effectiveness of the proposed
Lm Lm p r Lm r approach, some simulation studies were carried out using
pLs
f 2 x   Vqs 
L L  L R  C L L  L 1 dCe (7)
r  s r s r e  s r m
 2
 MATLAB/SIMULINK. The induction motor used in the
simulation is a 2 pole machine with a 1.5 kW rated power.
Lm Lm p  r Lm p  r dt The test concerns a no-load starting of the motor with a
R C reference speed Ωref =150rad/sec and reference rotor flux
f 3x   s  r e2  p ɸr=0.3 Wb.
p r
In Figure. 4, the application of the load torque 1Nm causes
However, it remains to conceive the fuzzy representation a step change in the electromagnetic torque response without
of the system; this representation gathers all uncertainties of any effects on the flux responses, which are maintained
the system defined by: constant, due to the decoupled control introduced by the
proposed approach.
~
Vds 


~ ~ ~ ~
f1 Ce ,  r , 
~
 

 200

~  ~
150

 ~  ~ dC ~ ~  
X: 0.4084

y  Vqs   f ~
~ x    f 2  Ce , e ,  r ,   
Speed(rad/s)

Y : 150
100

(8)
dt
50

 ~    ~ ~ ~ ~ 

 
0

  s  f 3 Ce ,  r ,  
 
-50
0 1 2 3 4
Time(s)

978-1-4799-0275-0/13/$31.00 ©2013 IEEE

Proceedings of the 3rd International Conference on WeAB.1

4
0.4
3

2 0.35
Torque(N.m)

Rotor flux (wb)

1

0 0.3
-1

-2 0.25
0 1 2 3 4
Time(s)

0.2
0.4
Rotor flux d axis (wb)

3 3.2 3.4 3.6 3.8 4

Time (s)
b) Fuzzy vector control
0.2
Figure 6. The steady-state performance of the two control techniques.

0
The robustness of the fuzzy vector control is confirmed
0 1 2 3 4 with the variation of the rotor resistance and the moment of
Time (s)
inertia in terms of rise time and overshoot of the speed.
0.1
0.5
Rotor flux q axis (wb)

0.05 Fuzzy vector control

0.4 vector control

The rise time (s)

0
0.3
-0.05
0.2
-0.1
0 1 2 3 4
0.1
Time (s)

Figure 5. The performance of the fuzzy vector control. 0

0 20 40 60 80
Variation of Rr (%)

Evaluation of the steady-state performance for a load of 1 0.14

Fuzzy vector control
Nm has been performed. In this simulation, VC only 0.12
Vector control
The rise time (s)

incorporates proportional integral current controllers. The 0.1

comparison between the two control techniques shows that 0.08
the torque and rotor flux ripples are more significant in the
0.06
case of the vector control. These results are shown in Figure.
5. 0.04
0 20 40 60 80
3
Vartiation of the moment of inertia (%)

2.5
Figure 7. The rise time.
2
Torque (Nm)

1.5
30
1
25
0.5
Fuzzy vector control
20
Speed (rad/s)

Vector control
3.2 3.4 3.6 3.8 4
Time (s) 15

10

0.35 5
Rotor flux (wb)

0
0.3 0 20 40 60 80
Variation of Rr (%)

0.25 25

20
3.2 3.4 3.6 3.8 4
Speed (rad/s)

Time (s)
15 Fuzzy vector control
a) Vector control Vector control
10
3

2.5 5

2
Torque (Nm)

0
0 20 40 60 80
1.5
Variation of the moment of inertia (%)
1
Figure 8. The Overshoot of the speed.
0.5

0
3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9
Time (s)
IV. IMPLEMENTATION OF THE FUZZY VECTOR CONTROL
The experimental system consists of two essentials parts:
(1) the power driver, (2) the control system. Figure. 8 shows
a synoptic scheme of the experimental platform.

978-1-4799-0275-0/13/$31.00 ©2013 IEEE

Proceedings of the 3rd International Conference on WeAB.1

various operating conditions to highlight the performance of

the proposed algorithm. Some selected experimental results
are presented here.
In test 1, we aimed to analyze the performance of an
induction motor with constant speed (150 rad/sec) of the new
fuzzy vector control at no load torque. It can be seen from
the Figure. 10 that the actual speed converges to the
reference value without any overshoot and undershoot and
with small steady-state error.
200

Figure 9. Synoptic scheme of the experimental platform. 150

Speed (s)
100

The control system is based on the DS1104 Controller 50

Board by dSPACE plugged into a computer. Its development 0

0 2 4 6 8 10 12 14 16 18
Time (s)
software operates under Matlab/Simulink environment and is
divided into two main components: Real Time Interface 2

(RTI) which is the implementation software and ControlDesk 1

Phase current (A)

which is the experimentation software. RTI is a Simulink 0

toolbox which provides blocks to configure models. These -1

blocks allow the users to access the dSPACE hardware. -2

-3
ControlDesk allows, as for it, the users to control and 4.18 4.2 4.22 4.24
Time (s)
4.26 4.28 4.3

monitor the real-time operation by using a lot of virtual 1

instruments and building a control window. 0.5

The power driver is composed with an induction machine,

Torque (Nm)

0
equipped with an incremental encoder position sensor
delivering 1024 pulses per revolution mounted on the rotor -0.5

shaft of the IM, powder brake coupled to the IM via torque -1

0 2 4 6 8 10 12 14 16 18

transducer and power circuit for the drive consist a Semikron Time (s)

IGBT based voltage source inverter with opto-isolation and Figure 11. The performance of the fuzzy vector control.
gate driver circuit SKHI22A. The DC voltage for the VSI is
In test 2, the variation of rotor speed is presented. It can be
achieved through a three-phase diode bridge rectifier
noted how the measured speed tracks accurately the
module. A capacitive filter is used at the DC link of this
reference speed. These results are presented in Figure.11. It
module to reduce the voltage ripples. Figure. 9 displays a
can be seen also the better performance when the changing
photo of the experimental system.
of the rotational direction of the induction motor within 0.5
seconds.
200

180

160

140
Speed (rad/s)

120

100

80

60

40

20

0
0 2 4 6 8 10 12 14 16 18
Time (s)

200

150

100
150
50
Speed (rad/s)

100

Figure 10. Photo of the experimental system. 0

Speed (rad/s)

50

-50 0

-50
-100

In this experiment the induction motor parameters are -150

-100
-150

determined using the no-load test and blocked rotor test. -200
0 2 4 6 8
Time (s)
10 12 14
12.9
16
13 13.1 13.2 13.3
Time (s)
18
13.4

These parameters are presented in APPENDIX. It is Figure 12. The experimentation variation of rotor speed.
important to notify that all experimental results are obtained
after running the motor for at least 15 min. This period will
warm up the resistances; which gives an uncertainty in the
parameters of this motor. Experiments were carried out on

978-1-4799-0275-0/13/$31.00 ©2013 IEEE

Proceedings of the 3rd International Conference on WeAB.1

V. CONCLUSION [5] Kuo-Kai Shyu, Faa-Jeng Lin, Hsin-Jeng Lin, Bor-Sen Juang, Robust
variable structure speed control for induction motor drive, IEEE
The intelligent vector control of induction motor has been Transactions on Aerospace and Electronic Systems, Vol. 35, pp 215 –
successfully implemented in real-time. The validity of the 224, 1999.
proposed control technique has been established both in [6] L.Barazane, C.Larbes, R.Ouiguini, Y. Sellami, Commande hybride,
linéarisante et à structure variables d’un moteur asynchrone,
simulation and experiment at different operating conditions. ICEE’2000, Algeria 2000.
This control technique has revealed very interesting [7] P.A. Egiguren, Robust adaptive Variable Structure Control of
characteristics. The novelty in this control, compared to induction motor drives, IEEE International Symposium on Industrial
previous works in the literature, is to replace the FOC block Electronics, pp. 582 – 587, 2008.
by the fuzzy vector control block to control the induction [8] Jinpeng Yu, Bing Chen, Haisheng Yu, Position tracking control of
induction motors via adaptive fuzzy backstepping, Energy
motor. This control maintains the decoupling and provides a Conversion and Management 51,pp 2345–2352, 2010.
high dynamic and statistic performance. Furthermore, [9] C.C. Gastaldini, R.P. Vieira, R.Z. A olin, .A. r ndling, An
compared with to the conventional vector control, the fuzzy adaptive feedback linearization control for induction motor, 2010 XIX
vector control is simple to implement as it uses a reduced International Conference on Electrical Machines (ICEM), 2010.
model, reduced time of simulation and its insensitivity to [10] M.W.Dunnigan, S. Wade, B.W. Williams, X. Yu, Position control of
a vector controlled induction machine using slotine’s sliding mode
parametric variation. Another advantage of this control is the
control approach, IEE Proceedings on Electrical Power Applications
absence of d-q current controllers. Therefore, leading the 145 (3), pp 231-238, 1998.
induction motor with fuzzy vector control presents a low cost [11] Tian-Jun Fu, Wen-Fang XieA, novel sliding-mode control of
system for many applications. induction motor using space vector modulation technique, ISA
Transactions 44, pp 481–490, 2005.
APPENDIX [12] B. K. Bose, Sliding mode control of induction motor, in Proc. IEEE-
IAS Annu. Meeting, pp.1985-1998, 2007.
Nomenclature: [13] Y.H.Hwang, K.K.Park, H.W.Yang, Robust adaptive backstepping
Rs, Rr: stator and rotor resistances, respectively, control for efficiency optimization of induction motors with
Ls, Lr: stator and rotor inductances, respectively, uncertainties, IEEE International Symposium on Industrial
Lm: mutual inductance, Electronics, pp.878-883, 2008.
J: moment of inertia, [14] N. Ezziani, N. Essounbouli, A. Hamzaoui, A backstepping fuzzy
P: the number of pole pairs, adaptive controller of induction motor, 16 Mediterranean conference
Ω: rotor speed, of control and automation, pp 1622-1627, 2008.
Φdr, Φqr: rotor flux d-q components, [15] Alma Y. Alanis, Real-time Discrete Backstepping Neural Control for
Ids, Iqs: stator current components, Induction Motors, IEEE Transactions on control systems technology,
Vol. 19, pp 359 – 366, 2011.
Vds, Vqs: stator voltage components (control signals),
Φr: reference rotor flux, [16] L. Barazane, C. Larbes and M.S. Boucherit.: Robust hybrid control
based on feedback linearization and field-oriented controls of
Ωref : reference rotor speed,
induction motor, Archives of Electrical Engineering. Vol. 3, No 2, pp.
Cr: the load torque, 115-135, 2003.
Ce : the electromagnetic torque,
[17] J. Bocker, S. Mathapati, State of The Art of Induction motor Control,
ωs: stator pulsation.
IEEE Trans on Industrial Electronics, vol. 5, pp. 1459-1464, 2007.
The motor data are as follows:
[18] S. Peresada, A. Tilli, and A. Tonielli, Theoretical and Experimental
Electric power: P =1.5KW,
Comparison of Indirect Field-Oriented Controllers for Induction
Stator voltage: V = 380/ 220V, Motors, IEEE Transaction on Power Electronics, vol. 18, No. 1, pp.
Number of Poles: p = 2, 151-163, January 2003.
Rated speed: Ω =150 rad /sec, [19] A.K. Rathore, Improved Performance of Fuzzy Logic Based Direct
Frequency: f = 50Hz, Field Oriented Controlled Induction Motor, 9th IEEE International
Stator resistance: Rs = 4.70Ω, Congress Power Electronics, pp. 152-157, 2008.
Rotor resistance: Rr = 4.32Ω, [20] S. Rafa and all, Fuzzy vector control of induction motor, the 2013
Stator inductance: Ls = 0.276H, IEEE International Conference on Networking Sensing and Control
Rotor inductance: Lr = 0.276H, (ICNSC 13), IEEE, Paris, France, avril 2013.
Mutual inductance: Lm = 0.262H,
Moment of inertia: J = 0.0023kg.m2,
Viscous friction coefficient f=0.00088 Nms.

REFERENCES
[1] F. Blaschke, The principle of field orientation applied to the new
transvector closed-loop control system for rotating field machines,
Siemens-Rev, Vol. 39, pp. 217-220, 1972.
[2] W. Leonhard, Control of Electric Drives. Berlin, Germany; Springer-
Verlag, 1995.
[3] D.W. Novotny, T. A. Lipo, Vector Control and Dynamics of AC
Drives. Oxford, U.K.: Clarendon, 1996.
[4] Ramón Blasco Giménez, High performance sensorless vector control
of induction motor drives, Thesis of the University of Nottingham,
December 1995.

978-1-4799-0275-0/13/$31.00 ©2013 IEEE