Middle-East Journal of Scientific Research 20 (5): 586-597, 2014

ISSN 1990-9233
© IDOSI Publications, 2014
DOI: 10.5829/idosi.mejsr.2014.20.05.11341

Design and Comparison of Vector and

Direct Torque Control of 3-Phase Induction Motor Drive

B. Vamsi Krishna

Department of Electrical and Electronics Engineering,

Bharath University, Chennai, India

Abstract: The aim of this paper is to design and compare the performance of vector and direct torque control
of 3-phase Induction motor drive using Matlab/Simulink tool and In conventional vector and direct torque
control (DTC) has, over the years, become of the most popular methods of control for induction motors (IM)
drive system. In DTC a single stator voltage vector of the inverter standard topology is selected every control
sampling period and it is maintained for the whole period. This makes the torque and stator flux reference values
to be reached quickly, with a small number of inverter switching, resulting in a very good dynamic response.
However, by this switching technique, based on hysteresis, large and small torque and flux errors are not
distinguished, which cause an extra torque ripple in motor steady state operation According to the simulation,
it can be verified that this Method not only Decrease the oscillation of torque and flux, but can simplify the
design of control system as well. The analysis has been based both on an analytical and experimental approach.
The results have shown that the vector control method, compared to the direct torque control of three phase
induction motor drive. The direct comparison of the performances of the motor control technique has been
made both for constant load and constant dissipated power conditions.

Key words: Induction Motor Vector control Direct torque control Torque Ripple Hysteresis controller
Switching Frequency

INTRODUCTION Ideally, a vector control of an induction motor drive

operates like a separately exited dc motor drive. In a dc
The induction motor drives with squirrel cage–type machine, neglecting the armature reaction and field
have been the workhorses in industry for variable speed saturation, the developed torque is given by:
applications in a wide power range that covers from
fractional horse power to the multi mega-watts because of Te = Kt Ia If (1)
their ruggedness and low cost [1]. The control and
estimation of ac drives in general more complex than those where Ia = armature current, If = field current. The
of dc drives and this complexity increase substantially if construction of a dc machine is such that the field flux f
high performance is needed. The main reason for this produced by the current If is perpendicular to the armature
complexity in the control of induction motor are the need flux a produced by the armature current Ia. Theses space
of variable frequency, harmonically optimum converter vectors, which are stationary in space, are orthogonal or
power supplies and the difficulties of processing feedback decoupled in nature [4].This means that when torque is
signals in the presence of harmonics [2]. The induction controlled by controlling the current Ia, the flux f is not
motor (IM) thanks to its well known advantages of simple affected and we get fast transient response and high
construction, reliability, ruggedness and low cost has torque/ampere ratio with rated f. When the field current
found very wide industrial applications. These If is controlled, it affects the field flux only, but not
advantages are superseded by control problems when armature flux. Because of inherent coupling problem an
using an IM in industrial drives with high Performance induction motor cannot generally give such a fast
demands [3]. response [5].

Corresponding Author: B. Vamsi Krishna, Department of Electrical and Electronics Engineering,

Bharath University, Chennai, India.
586
 Middle-East J. Sci. Res., 20 (5): 586-597, 2014

Dc machine like performance can also be extended to switching frequency are removed. The principle of
an induction motor if the machine controls is considered operation of vector control and DTC control is detail in
in a synchronously rotating reference frame (de-qe), where the next Topics.
the sinusoidal variable appear as dc quantities in a steady
state. The control current inputs, ids* and iqs*. These are Vector Control: The vector control based on current
the direct axis component and quadrature axis component, controlled PWM inverter which is very popular method
respectively, in asynchronously rotating reference frame. used in industry. The fundamentals of vector control
with vector control, ids is analogous to the field current implementation can be explained with the help of Figure 1,
and iqs is analogous to armature current ia of a dc machine. where the machine model is represented in a
Therefore the torque can be expressed as: synchronously rotating reference frame. Assume that the
inverter has unity current gain, that is, it generates
Te = Kt ids iqs (2) currents ia, ib, ic as dictated by the corresponding command
current ia*, ib*, ic* from the controller. A machine model
This means that when iqs* controlled, it affects the with internal conversions is shown on the right. The
actual iqs current only, but does not affect the flux r. machine terminal phase currents ia, ib, ic are converted to
Similarly when ids* is controlled it controls the flux only i dss, iqss components by 3 /2 transformation. These are
and does not affect the iqs component of current [6]. then converted to synchronously rotating frame by the
Based on the method of generation of the unit vector for unit vector components cos e and sin e before applying
finding the speed of rotation of the synchronous frame, them to the de–qe model as shown. The transformation
the vector control is classified as direct or indirect equations will be given later in this chapter. The controller
method. The speed of rotation is directly found from the makes of inverse transformation as shown so, that the
machine model or the estimator in the case of direct vector control currents ids * and iqs * correspond to the machine
control, whereas in the case of indirect vector control it is currents ids and iqs respectively. In addition the unit vector
found by adding the rotor speed with help of slip speed. assures correct alignment of i ds, current with the flux
The reduced version of the Vector control is Direct vector r. There are essentially two general methods of
Torque Control [7]. vector control. One is direct control and other is indirect
Direct Torque Control (DTC) is emerging technique control depends on the how unit vector is generated for
for variable speed control of inverter driven induction the control [9].
motor. This method allows the direct control of stator flux The key component of the Field oriented control
and instantaneous torque through simple algorithm. (FOC) strategy is the Clarke and Park transform blocks.
Basically, it uses torque and stator flux control loops, These map the three phase stator currents onto a direct
where the feedback signals are estimated from the and quadrature rotating reference frame that is aligned
machine terminal voltages and currents. The torque with the rotor flux. This decouples the torque and flux
command can be generated by the speed loop. The loop producing components of the stator currents allowing the
errors are processed through hysteresis bands and fed to induction motor to be controlled in much the same was as
a voltage vector look-up table. The flux Control loop has a separately excited DC machine. The d-axis component of
outputs +1 and -1, whereas the torque Control loop has the stator current is related to the rotor flux magnitude via
three outputs, +1, 0 and -1. The inverter voltage vector Equation below.
table also gets the information about the location of the
stator flux vector. From the three inputs, the voltage dr = Lm * ids (3)
vector table selects an appropriate voltage vector to
control the PWM inverter switches. This is analogous to where, Lm is mutual inductance of the motor.
the hysteresis based current control in the case of vector In the indirect vector control the unit vectors are in
control. Due to the absence of several controllers and the feed forward manner. The fundamental principle of
transformations in DTC, the delay in processing signal is indirect vector control is explained with the help of
drastically reduced [8]. phasor diagram as shown in Figure 2. With the help of
Because of the simplicity and superior performance of phasor diagram. The d e-q e axes are fixed on the stationary
DTC make it more popular in industrial applications, if but the d r-qr axes, which are fixed on the rotor are moving
some of the disadvantages like high ripple, variable at speed r as shown. Synchronously rotating axes de- qe

587
 Middle-East J. Sci. Res., 20 (5): 586-597, 2014

Fig. 1: Vector control implementation principle with machine d e – qe model

Fig. 2: Basic Three Phase Voltage Source Inverter

are rotating ahead of the dr –qr axes by the positive slip sl = ((Lm* Rr)/(Lr* r))*iqs (7)
angle sl corresponding to slip frequency sl. Since the
rotor pole is directed on the de axis and = r + sl, we can where, r = dr has been substituted.
write If rotor flux r = constant, which is usually the case
then from equation (6).
e= e dt = ( r + sl ) dt = r + sl
(4)
r = Lm * ids (8)
The rotor pole position is not absolute, but is slipping
with respect to the rotor at frequency sl. The phasor In other words the rotor flux is directly proportional to
diagram suggests that for decoupling control, the stator current i ds.
flux component current ids should be aligned on the de axis
and the torque component of current iqs should be on the Overview of Space Vectors: Representing three phase
qe axis. quantities as Space Vectors is particularly useful for
For decoupling control it is desirable that qr =0, power electronic applications. Essentially this
That is, methods defines a three phase system with a single unity
vector:
d dr /dt = 0 (5)
us = (1 + a + a2) (9)
So that the total rotor fluxes r is directed on the d
e

axis. where a = e j2 /3
Using this unity vector, a space vector representation
Then, of phase voltages va, vb, vc is therefore:

(Lr/ r )* (d r/ dt) + r = Lm * ids (6) V = (2/3) (va + a vb + a2vc) (10)

588
 Middle-East J. Sci. Res., 20 (5): 586-597, 2014

Fig. 3: Switching Voltage Space Vectors

Table 2.1: Space vector switching pattern over a time period Tpwm to be equal to the average voltage
a b c Va Vb Vc Vab Vbc Vca of the reference space vector in that period. This is done
0 0 0 0 0 0 0 0 0 by time modulating the two adjacent switching vectors
1 0 0 2/3 -1/3 -1/3 1 0 -1
that set the bounds for the sector the reference vector is
1 1 0 1/3 1/3 -2/3 0 1 -1
0 1 0 -1/3 2/3 -1/3 -1 1 0
currently in. The binary representation for two adjacent
0 1 1 -2/3 1/3 1/3 -1 0 1 switching vectors differs by one bit so only one of the
0 0 1 -1/3 -1/3 2/3 0 -1 1 upper transistors needs to change. This improves the
1 0 1 1/3 -2/3 1/3 1 -1 0 inverter’s performance and significantly simplifies the
1 1 1 0 0 0 0 0 0 digital implementation of the algorithm.
If we assume that the switching frequency is high
The 2/3 scaling factor is necessary to ensure that the and that the change in Vref over this period is small then
system remains power invariant. the modulation scheme can be represented using the
following equation:
The Space Vector Switching Pattern: There are eight
possible combinations of switching patterns for the three Tpwm Vref = 2/3(T1Vswitch x +T2 V switch x+1 + T3 VNull ) (11)
upper MOSFETs of the inverter switching legs. The phase
and line to line voltages generated by each of these where Vswitch x and Vswitch x + 1 represent the adjacent
combinations can be calculated from equation and are switching vectors for sector x. VNull is the null switching
expressed as a fraction of Vdc. The results of this are vector and T1, T2 and T3 are the durations for each
presented in Table 2.1. switching vector. The sum of T1 and T2 is less than
For each switching combination a voltage space or e qual to Vref so the null vector VNull is activated for
vector can be constructed using the phase voltages and time T3 to make up the rest of the switching period such
equation (10) When these space vectors are plotted on a that:
set of real and imaginary axes the switching diagram in
Figure (3) is obtained. The switching space vectors divide Tpwm = T1 + T2 + T3 (12)
the axes into 6 equally sized sectors. The two null vectors
000 and 111 are located at the origin. The length of the reference vector determines the
magnitude of the output voltage, while the speed with
Generating a Reference Voltage: The objective of which the vector rotates around the circle determines the
SVPWM is to approximate a reference space vector Vref, frequency of the three phase system. In a motor
somewhere within the transcribed circle of Figure (4) application the direction of the rotor depends on
using a combination of the eight switching vectors. whether the reference vector is rotating clockwise or
One method is to set the average voltage of the inverter anti-clockwise.

589
 Middle-East J. Sci. Res., 20 (5): 586-597, 2014

machine as shown in Figure (6). Where ds-qs corresponds

to stator direct and quadrature axes and dr-qr corresponds
to the rotor direct and quadrature axes.
In this two-phase model the time varying inductances
can be eliminated by referring the stator and rotor
variables to a common reference which may rotate at any
speed.
Synchronously rotating de-qe axes rotates at
synchronous speed e with respect to the ds-ds axes and
the angle e = et. The two phase ds-qs windings are
transformed into the hypothetical windings mounted on
the de-qe axes. The voltage on the ds-qs axes can be
converted into the de-qe qxes as follows:

Vqse = vqss cos e- vdss sin e (13)

Fig. 4: Vref Mapped onto the Switching Pattern
U2 Vdse = vqss sin e+ vdss cos e (14)

Again resolving the synchronously rotating frame

(2/3).T1. SECTOR parameters into a stationary frame the relation are:
Vdc 111 1VIref
Vqss = vqse cos vdse sin (15)
(2/3).T2. U1 e+ e

Vdc Vdss = -vqse sin vdse cos (16)

e+ e

Fig. 5: Generating Vref in Sector 1

Direct Torque Control: Direct Torque Control (DTC) is an
entirely different approach to induction motor control that
was developed to overcome Vector Control’s relatively
poor transient response and reliance on induction motor
parameters. The standard closed loop Vector Control
structure consists of

A number of coordinate transforms to decouple flux

and torque
A current model to estimate the rotor flux angle
Two PI controllers for the direct and quadrature
currents
An additional PI controller to regulate the speed.

In DTC the first three components are removed and

Fig. 6: Equivalent two-phase machine for 3-phase machine replaced by two hysteresis comparators and a look up
table. Another benefit is that it can calculate torque
Dynamic D-Q Model of Induction Machine: Vector control without the need for complex observer algorithms or
is based on the dynamic d-q model of the machine. The mechanical speed sensors.
dynamic performance of an ac machine is somewhat
complex because the three-phase rotor windings move Principle of Operation of Direct Torque Control: In the
continuously with respect to three-phase stator windings. Direct Torque Control method instead of using coordinate
The machine model can be described by differential transforms and PI controllers to determine the output
equations with time varying mutual inductance, but such reference voltages for the inverter, it uses a look up table.
a model tends to be very complex. The three phase Flux and torque are used as feedback signals for the
machine can be represented by an equivalent two-phase controller. The errors in torque, stator flux magnitude and

590
 Middle-East J. Sci. Res., 20 (5): 586-597, 2014

Fig. 7: Block diagram of Direct torque & Flux control

method

angle are used as the inputs for this table. Each of these
variables is discretised into a specific number of levels Fig. 8: Trajectory of stator flux vector in DTC control
using a hysteresis comparator. The error in magnitude of
the stator flux is 1 if it is low and 0 otherwise. Torque error H = 1 for E > + HB (21)
is represented using three levels with -1 being too high,
0 being acceptable and 1 being too low. Finally the stator H = -1 for E < - HB (22)
flux angle is discretised into six, 60 degree sectors
corresponding to the regions bounded by the PWM where,
space vectors. The black diagram of the Direct Torque 2HB = total hysteresis band width of the flux controller.
control Induction Motor drive is shown in Figure (7).
The circular trajectory of the command flux vector
(17) s* with hysteresis band rotates in anti-clock wise
=s ∫ (v s − Rs i s )dt direction, as shown in Figure (8).
The d and q components of s are given by: The actual stator flux s within the hysteresis band
and it tracks the command is constrained flux in zigzag
(18) path. The torque control loop has three levels of digital
=ds ∫ (vds − Rsids )dt output as follows:

(19) HTe = 1 for ET > +HBT (23)

=qs ∫ (vqs − Rsiqs )dt
The electromagnetic torque is an important output HTe = -1 for ET < -HBT (24)
variable that determines such mechanical dynamics of the
machine as the rotor position and speed. The HTe = 0 for -HBT <ET < +HBT (25)
electromagnetic torque of an induction motor in stator
reference frame is given by: The feedback flux and torque are calculated from the
machine terminal voltage and currents.
3p (20)
=Te [iqs ds − ids qs ]
22 Optimal Switching Logic: The sector number S (K)is
We can estimate the electromagnetic torque of also calculated from signal commutation black in which
induction motor: the flux vector s lies. There are six sectors, each 60
degree angle wide, as shown in Figure (8) The voltage
Flux and Torque Hysteresis Controllers: The diagram for vector table black receives the input signals H , HTe and
direct torque and flux control is shown in Figure (6) The S(K) and generates the appropriate voltage vector for the
command stator flux s* and torque te* magnitudes are inverter by a look up table which is shown in Table (2)
compared with the respective estimated values and the Neglecting the stator resistance Rs of the machine, we can
errors are processed through hysteresis band controllers. write:
As shown. The flux loop controller has two levels of
digital output according to the following quations. s = Vs * t (26)

591
 Middle-East J. Sci. Res., 20 (5): 586-597, 2014

Table 2: Switching table of inverter voltage Table 3: Parameters of 3HP 460V, 50Hz 4 Pole Induction Machine
H HTe S(1) S(2) S(3) S(4) S(5) S(6) Parameter Values
1 1 V2 V3 V4 V5 V6 V1 Stator Resistance (Rs) 1.115
1 0 V0 V7 V0 V7 V0 V7 Rotor Resistance (Rr) 1.083
1 -1 V6 V1 V2 V3 V4 V5 Stator Inductance (Ls) 0.005974mH
-1 1 V3 V4 V5 V6 V1 V2 Rotor Inductance (Lr) 0.005974mH
-1 0 V7 V0 V7 V0 V7 V0 Magnetizing Inductance (Lm) 0.0347mH
-1 1 V5 V6 V1 V2 V3 V4 Inertia (J) 0.02Kg/m2
Rated Current 5.0 Amp
Rated Torque 22.38 Nm
Which means that s can be changed incrementally Rotor speed 1500 rpm
by applying stator voltage vector Vs, for time increment
t. Table 4: Inverter Data
The voltage vector look-up table for DTC control is Electrical Quantity Rating
shown in table for the three inputs H , HTe and S (K). Power 3 KW
Depending upon these three input signals the inverter Voltage 460V -AC
gets respective voltage vector. The zero vector short
Table 5:
circuits the machine terminal and keeps the flux and
Parameter Vector Control DTC
torque essentially unchanged.
Torque Ripple (N-m) -0.2 to 1.08 -1.7 to 1.7
Flux ripple (Wb) 0.19 to 0.4 0.97 to 1.2
RESULTS AND DISCUSSIONS

The ripples in the electromagnetic torque at the

The Complete model of Vector and Direct Torque
steady state i.e. rotor speed settles to its reference speed
Control of induction motor drive is developed by
were shown in Figure (11). In the steady state
Matlab/Simulink tool box. The major problems associated
electromagnetic torque should be the one. Figure (12)
with Vector Control drive is feedback signal estimation is
shows the stator flux ripples and Figure (13) shows the
complex and it requires a number of coordinate
stator q-axes and d-axes fluxes which can be controlled
transformations. The DTC induction motor drive provides
independently.
superior performance and here is no need of coordinate
The simulated waveform of stator flux in d-q plane is
transformation. The main drawback of DTC drive is the
shown in Figure 15. The stator d-axes flux was measured
flux and torque ripples. This chapter briefly discusses the on the x-axes and stator q-axes flux was measured on
simulation results of the Vector and Direct Torque Control y-axes and the corresponding graph was shown in
induction motor drive. Figure 15. The circular trajectory of flux vector rotates in
anti-clock wise direction with the hysteresis band.
Motor Parameters: The rating and parameters of the Simulated waveforms of stator current,
induction machine and inverter data used for simulation electromagnetic torque and rotor speed shown in
are given in Table (3) and (4). Figure (16) The reference speed was 100 rad/sec. The
The Matlab/Simulink implementation of conventional actual speed of the rotor settled to its reference speed in
Vector Control induction motor drive is shown in 0.26 seconds. After reaching the rotor speed to its
Figure 9. The three phase quantities are converted into reference speed the electromagnetic torque becomes zero.
two phase quantities and vice versa by using Clark and In the steady state the ripples in the electromagnetic
Park transformations. The simulated waveforms of stator torque and stator flux was shown in Figure (17) and (18)
current, electromagnetic torque and rotor speed (rad/sec) respectively. The torque ripples is between -1.7 to 1.7 and
was shown in Figure 10. the stator flux ripples is between 0.97 to 1.2. And these
The reference speed given is 100 rad/sec. After values are tabulated in Table (5) and compared with
0.75 seconds the step given to the reference speed is vector control induction motor drive [10-13].
150 rad/sec. The corresponding changes in rotor speed,
stator current and electromagnetic torque was shown in Comparison of Results: The Matlab/Simulink simulated
Figure 10. After reaching rotor speed to its reference results of the Torque ripples and stator flux ripples in
speed the electromagnetic torque settles to the rated steady state for the Vector and Direct Torque control of
value. induction motor drive is tabulated in Table (5).

592
 Middle-East J. Sci. Res., 20 (5): 586-597, 2014

Fig. 9: Matlab/Simulink implementation of Vector Control Induction Motor Drive

Fig. 10: Stator current, rotor speed and electromagnetic torque waveforms of Vector Control induction motor drive

593
 Middle-East J. Sci. Res., 20 (5): 586-597, 2014

Fig. 11: Ripples in electromagnetic torque

Fig. 12: Ripples in stator flux

Fig. 13: Stator q-axes and d-axes fluxes which can be controlled independently

594
 Middle-East J. Sci. Res., 20 (5): 586-597, 2014

Fig. 14: Matlab/Simulink implementation of Direct Torque Control Induction Motor Drive

Fig. 15: Stator d-q axes flux in x-y plane

Fig. 16: Simulated waveforms of stator current, electromagnetic torque and rotor speed for DTC induction motor drive

595
 Middle-East J. Sci. Res., 20 (5): 586-597, 2014

Fig. 4.9: Ripples in electromagnetic torque

Fig. 18: Ripples in stator flux

CONCLUSION REFERENCES

Induction motor with squirrel cage type is best suited 1. Bose, Bimal K., Modern Power Electrics and AC
for industrial applications because of their low cast, Drives, Prentice hall of India.
ruggedness. The advent of the Power Electronics and 2. Vas, P., 1998. Sensorless Vector and Direct torque
semiconductor devices has contributed to rapid control, Oxford University Press.
developments in the field of motor control. The two best 3. Takahashi and T. Noguchi, 1986. A new quick
methods of the controlling of the induction motor drive response and high efficiency control strategy of
are Vector and Direct Torque Controls. This project report an induction motor, IEEE Trans. Ind. Appl.,
briefly explains the Principle of operation of Vector and 22(5): 820-827.
Direct Torque Controls, their salient feature, advantages 4. Marino, P., M. D’Incecco and N. Visciano, 2001. A
and disadvantages. The Vector and Direct Torque Control comparison of direct torque control methodologies
induction motor drive is implemented using for induction motor, Procc. IEEE Porto. Power Tech
Matlab/Simulink tool box. The simulation results are Conference, Portugal.
obtained, analysed and reported. From the simulated 5. Takahashi, I. and Y. Ohmori, 1989.
results it is observed that the torque ripples in the DTC High-performance direct torque control of
induction motor drive is high compared to Vector control induction motor, IEEE Trans. Ind. Appl.,
drive. The transient response is good. with DTC drive. 25(2): 257-264.

596
 Middle-East J. Sci. Res., 20 (5): 586-597, 2014

6. Habetler, T.G., F. Profumo, M. Pastorelli and 11. Kishwar Sultana, Najm ul Hassan Khan and Khadija
L.M. Tolbert, 1992. Direct torque control of induction Shahid, 2013. Efficient Solvent Free Synthesis and X
machines using space vector modulation, IEEE Ray Crystal Structure of Some Cyclic Moieties
Trans. Ind. Appl., 28(5): 1045-1053. Containing N-Aryl Imide and Amide, Middle-East
7. Casadei, D., G. Sera and A. Tani, 0000. Stator flux Journal of Scientific Research, 18(4): 438-443.
vector control for high performance induction motor 12. Pattanayak, Monalisa. and P.L. Nayak, 2013. Green
drives using space vector modulation, Electro Synthesis of Gold Nanoparticles Using Elettaria
motion, 2(2): 79-86. cardamomum (ELAICHI) Aqueous Extract World
8. Lascu, C., I. Boldea and F. Blaabjerg, 2000. A Journal of Nano Science & Technology, 2(1): 01-05.
modified direct torque control for induction motor 13. Chahataray, Rajashree and P.L. Nayak, 2013.
sensorless drive, IEEE Trans. Industry Applications, Synthesis and Characterization of Conducting
36(1): 122-130. Polymers Multi Walled Carbon Nanotube-Chitosan
9. Casadei, D., G. Sera and A. Tani, 2000. Composites Coupled with Poly (P-Aminophenol)
Implementation of a direct torque Control algorithm World Journal of Nano Science and Technology,
for induction motors based on discrete space 2(1): 18-25.
vector Modulation, IEEE Trans. Power Electronics,
15(4): 769-777.
10. Shafaq Sherazi and Habib Ahmad, 2014. Volatility of
Stock Market and Capital Flow Middle-East Journal
of Scientific Research, 19(5): 688-692.

597