New Direct Torque Control Scheme of Induction Motor for

Electric Vehicles
M.Vasudevan* and Dr.R.Arumugam✝

*Department of Electrical and Electronics Engineering, Anna University, Chennai-25,

e-mail: vasumame@yahoo.com
✝
Department of Electrical and Electronics Engineering, Anna University, Chennai-25,
e-mail: arumugam@annauniv.edu

Abstract

In this paper, new scheme of direct torque
control of induction motor for electric vehicles is
proposed and the results of an investigation into
suitable torque control schemes are also presented.
The electric vehicle drive consists of rewound
induction motors and a three-level IGBT inverter.
The schemes investigated are Field Oriented Control,
Direct Torque Control (DTC), and DTC using Space
Vector Modulation. The results of Matlab-Simulink
simulations and a comparison between the control
schemes are presented. It is found that the DTC using
Space Vector Modulation scheme is best for this
application.
Keywords: Induction Motor, Field Oriented Control,
Direct Torque Control, Space Vector Modulation
1 Introduction
Induction motors are today the most widely used ac
machines due to the advantageous mix of cost,
reliability, and performances. Torque control in
induction motors can be achieved according to
different techniques ranging from the inexpensive
constant voltage/frequency ratio strategy to the
sophisticated sensorless control schemes. Now-a-
days, such efficient schemes have been implemented
in electric vehicles. The University of Canterbury has
developed an electric vehicle over many years as the
test bed for a variety of projects. It is based on a 1962
Austin Farina body with 20 lead acid batteries
providing a nominal 240-volt DC supply. Two 2.2kW
50Hz induction motors that have been rewound to
operate with an applied frequency of up to 150Hz
drive it. An SCR inverter originally drove the motors,
but recently new three-level IGBT inverters have
been developed for each motor. This has the advantage
of being able to produce three different levels of output
voltage compared to the standard two levels. With a
greater number of available levels for the output voltage
the desired sinusoidal voltage can be achieved more
accurately without increasing the switching frequency.
This results in less current harmonics and therefore more
efficient utilization of the available energy. The three-
level IGBT inverter that was developed was only tested
under open loop conditions, as no suitable controller
was available.
A suitable closed loop torque controller is required
to enable this new three-level inverter to be used in the
electric vehicle. Torque control is preferred for electric
vehicle applications instead of precise closed loop speed
control because it mimics the operation of an internal
combustion engine. It is important to make an electric
vehicle drive like a standard vehicle. An ideal electric
vehicle motor drive system would have high efficiency
and produce low torque ripple. By reducing the current
distortion, losses in the motor are reduced. Increasing
the inverter switching frequency could reduce current
distortion, but this would result in increased switching
losses in the inverter. Therefore it is important for the
control scheme to produce minimal current distortion for
a given switching frequency. In this paper, four different
control schemes are investigated to determine their
suitability for this application. A brief overview of the
operation of each scheme is presented followed by
results from Matlab-Simulink simulations. A
comparison of the advantages and disadvantages of
these three schemes is included.
2 Control methods
A number of different control schemes for accurate
torque control of an induction motor for this electric
 Figure 1: Field Oriented Control Block Diagram
vehicle application, have been investigated. Field
Oriented Control (FOC) and Direct Torque Control
(DTC) were chosen for simulation, as they are
standard induction motor control techniques. An
improvement to DTC is DTC using Space Vector
Modulation, a new torque control scheme proposed is
also simulated. These are briefly described in the
following sections.
2.1 Field oriented control
Field Oriented Control refers to induction motor
operation in a synchronously rotating d-q reference
frame that is aligned with one of the motor fluxes,
typically the rotor flux [2]. In this mode of operation,
control of the torque and flux is decoupled such that
the d-axis component of the stator current controls
the rotor flux magnitude and the q-axis component
controls the output torque.
The required d-axis component of the stator
current, ids, to achieve a given rotor flux magnitude
demand, Ȝ*dr can be determined from Equation (1).
λdr* = Lmids (1)
where, Lm = magnetizing inductance
The required q-axis component of the stator current,
iqs, for a given torque demand (T*em ),can be
determined from Equation (2).
3 P Lm
T * em = λ * dr i qs (2)
2 2 Lr
For the d-axis of the synchronously rotating reference
frame to be aligned with the rotor flux the ‘slip
relation’, Equation (3), must be maintained.
r r i qs
ϖ e − ϖ r = (3)
L r i ds
A diagram of the field-oriented controller is shown in
Figure 1. A proportional-integral controller regulates the
stator voltage to achieve the calculated stator
current. The required voltage is then synthesized by the
inverter using space vector modulation (SVM).During
motor operation the actual rotor resistance and
inductance can vary, for example with temperature. The
resulting errors between the values used and the actual
parameters cause an incomplete decoupling between
torque and flux.
2.2 Direct Torque Control
In principle, the DTC method selects one of the
inverter’s six voltage vectors and two zero vectors in
order to keep the stator flux and torque within a
hysteresis band around the demand flux and torque
magnitudes. The torque produced by the induction
motor can be expressed as Equation (4),
3 P Lm
Tem = λ r λ s . sin α (4)
2 Ls ' Lr
which shows the torque produced is dependent on the
stator flux magnitude, rotor flux magnitude, and the
phase angle between the stator and rotor flux vectors.
The induction motor stator equation, (5),
d λs
Vs = − Is.rs (5)
dt
can be approximated as Equation (6) over a short time
period if the stator resistance is ignored.
∆ λs = Vs.∆t (6)
This means that the applied voltage vector as shown TABLE 1: OPTIMAL VECTOR SELECTION TABLE
in Figure 2(b) determines the change in the stator flux (2-LEVEL) SECTOR
vector as shown in figure 2(a). If a voltage vector is Sector
applied that changes the stator flux to increase the ǻȜ ǻTem 1 2 3 4 5 6
phase angle between the stator flux and rotor flux Ĺ V2 V3 V4 V5 V6 V1
vectors, then the torque produced will increase. Ĺ 0 V0 V0 V0 V0 V0 V0
Ļ V6 V1 V2 V3 V4 V5
Ĺ V3 V4 V5 V6 V1 V2
It Ļ 0 V0 V0 V0 V0 V0 V0
Ļ V5 V6 V1 V2 V3 V4
It
3 It 2
3 The stator flux vector can be estimated using the stator
4 1 voltage equation, (5), as given in Equation (7).

)
4 It
5 6 λ s = ∫ ( Vs − rs I s dt (7)
As shown in Figure 3, the estimated stator flux magnitude
5 6 and torque output is compared to the demand values. A
voltage vector is then selected that will drive the torque and
λ flux towards the demanded values.

Figure 2(a): Torque change corresponding to the

position of flux

Figure 3: Direct Torque Control Block Diagram

2.3 Direct torque control using space vector

modulation

Figure 2(b): Direct Torque Control Operation
(2-level)
Since a two-level inverter is only capable of
producing six non-zero voltage vectors and two zero
vectors, it is possible to create a table that determines
the voltage vector to apply based on the position of
the stator flux and the required changes in stator flux
magnitude and torque. This is called the optimal
vector selection table and is shown as Table 1 for the
case of a two level inverter. It is possible to expand
the optimal vector selection table to include the larger
number of voltage vectors produced by a three-level
inverter.
Based on instantaneous current and voltage
measurements it is possible to calculate the voltage
required to drive the current output torque and stator
flux to the demanded values within a fixed time period.
This calculated voltage is then synthesized using Space
Vector Modulation.
The applied voltage and motor current can be used
to estimate the instantaneous stator flux and output
torque. From these the required change in output torque
and stator flux in order to reach the demanded values by
the end of the current switching period can be
calculated.
Equation (8) shows that a change in torque
corresponds to a change in stator current. The voltage,
 V, required to achieve that change in stator current
can be calculated knowing the back EMF of the
motor. The back EMF can be calculated from the
stator flux and current.
∆ T em =
3 P
2 2
(
λs X∆ Is ) (8)
As shown by Equation (6), the change in the stator
flux vector is only dependent on V. To solve for the
unknown voltage vector requires two equations, one
for the required change in torque, (9), and another for
the change in stator flux, (10).
∆Tem =
3 P ∆t
2 2 Ls '
(
λ s X (V − E ) ) (9)
Figure 4: Circuit Diagram of Space Vector Modulation
2.4 Simulation and results
All three of the control schemes have been simulated
using Matlab-Simulink. All simulations were
performed using a model of the motors that are used in
the electric vehicle. For the control schemes that
operate at a fixed switching frequency, an inverter
switching frequency of 10kHz was used. For the
standard DTC simulations, torque and flux hysteresis
bands of 1Nm and 0.0016Wb respectively were used to
give a switching frequency close to 10kHz at the
∆λ s = V .∆t (10)
These two equations can be solved to find the
smallest voltage required to drive both the torque and
flux to the demand values. The required stator voltage
can be calculated by adding on the voltage drop across
the stator resistance calculated using the current
measured from the last cycle. The calculated voltage is
then synthesized directly using space vector modulation.
If the inverter is not capable of generating the required
voltage then the voltage vector that will drive the torque
and flux towards the demand value is chosen and held
for the complete cycle. The complete circuit diagram of
space vector pulse width modulation is shown in
figure4.
chosen motor speed and load. A minimum vector hold
time of 25ms was chosen to simulate the time required to
sample currents and voltages, and calculate the new
vector. In sensored indirect field oriented control, the
complete system is simulated and explored the
performance of speed control. The simulation results are
presented based on MATLAB programming. The indirect
field oriented control speed closed-loop system is using
the estimated speed as the speed feedback. All physical
variables such as voltages, currents, flux linkages, flux
angle, speed and torque are normalized in per-unit. The
reference speed is set at 900 r.p.m. or 0.5 p.u. with
1800 r.p.m. base. The rotor angular velocity, estimated
rotor flux and torque response are shown in figures 6, 8
and 9 respectively.

In Sensorless direct field oriented control

scheme, rotor flux, its angle and rotor speed have been
estimated. The DFOC speed closed-loop system is
using the estimated speed as the speed feedback. The
stator currents and DC-bus voltage are measured and
all physical variables such as voltages, currents, flux
linkages, flux angle, speed, and torque are normalized
in per-unit. Torque demand response of sensorless
field oriented control is shown in figure8. Motor per
unit stator current is also shown in figure 10. Figure 7: Direct Field oriented Control torque
demand response

Figure 5: DTC using Space vector modulation

response to step change in torque demand
Figure 8:Estimated Rotor flux in Indirect Field
Oriented Control

Figure 6: Plot of rotor angular velocity in indirect Figure 9: Response to step change in torque demand
field oriented control in Indirect Field oriented Control

Figure 10: Motor per unit stator current
3. Comparison
Direct Torque Control was developed as an alternative
to FOC that had been in use for a number of years.
DTC has the advantage of not requiring speed or
position encoders and uses voltage and current
measurements only. It also has a faster dynamic
response due to the absence of the PI current regulator.
A disadvantage of DTC is its comparatively high
current distortion and torque ripple. DTC using SVM
is a different approach that involves a direct predictive
calculation of the voltage required to be applied to the
induction motor to change the stator flux and torque
magnitudes to the demanded values. It combines the
best features of the direct torque control schemes such
as fast torque response, no fixed switching frequency.
DTC using SVM was chosen as the torque control
scheme in this electric vehicle application. From the
above discussions, advantages and disadvantages of all
the control schemes are summarized as follows:
3.1 Field-oriented control
Advantages
i) Relatively simple high performance
scheme.
ii) A proven technique that has been
used for some time.
Disadvantages:
i) Relatively low dynamic performance
due to the PI current regulator
ii) Parameter detuning causes high
torque and flux magnitude errors.
3.2. Direct Torque control
Advantages:
i) Fast torque response
ii) Relatively simple
iii) No speed or position encoder is
required
Disadvantages;
i) High current distortion
ii) High torque ripple
iii) Switching frequency changes with
motor speed.
3.3 DTC using Space Vector Modulation
Advantages:
i) Fast torque response that is equivalent
to standard direct torque control
ii) Low steady state torque ripple and
current distortion that is equal to FOC
iii) Constant switching frequency
iv) Not as parameter sensitive as FOC
v) No speed or position encoder is
required
Disadvantages:
i) The control algorithm is very complex
compared to other control schemes.
ii) It requires a relatively fast processor to
implement at the desired switching
frequency.
4. Conclusions
Three different control schemes have been evaluated for
induction motor torque control in an electric vehicle
application. All other direct torque control techniques
exhibited improved transient torque response compared to
standard rotor flux FOC. The disadvantage of these
schemes is increased current distortion. Since high
efficiency is one of the prime requirements for an electric
vehicle drive and high current distortion results in
increase motor losses, this is unacceptable. DTC using
space vector modulation is an exception that exhibited
fast torque response and low current distortion and torque
ripple.
References
[1] X. Li, R. Duke and S. Round. Development of a three-
phase three-level inverter for an electric vehicle.
Australasian Universities PowerEngineering Conf.,
Darwin, Australia, 1999, pp 247-251.
[2] D. W. Novotny and T. A. Lipo. Vector Control and
Dynamics of AC Drives, Oxford University Press,1997.
[3] J. Holtz. Pulse width Modulation for Electronic Power
Conversion. Proc. of the IEEE, Vol. 82, No.8, pp. 1194 –
1213, Aug 1994.
[4] I. Takahashi and T. Noguchi. A new quick response
and high efficiency control strategy for an induction
motor. IEEE Trans. IndustryApplications. Vol. IA-22,
No. 5, pp. 820-827,Sept/Oct 1986.
[5] T.G. Habetler, F. Profumo, M. Pastorelli, and program,” IEEE transactions. Energy conversion, vol,
L.M.Tolbert. Direct Torque Control of Induction EC-2, pp. 662-668, Dec. 1987.
Machines using Space Vector Modulation. IEEE
[13] A.Keyhani and H.Tsai, “IGSPICE simulation of
Trans. Industry Applications. Vol. 28, No. 5,
induction machines with saturable inductances,” IEEE
Sept/Oct 1992
transactions. Energy conversion, vol,4, pp. 118-125, Mar.
[6] J-K. Kang and S-K. Sul. New Direct Torque
1989.
Control of Induction Motor for Minimum Torque
Ripple and Constant Switching Frequency. IEEE [14] Pierre Biden, Thierry Lebey, Member, IEEE, Gerard
Trans. Industry Applications, Vol. 35, No. 5, Montrery, Clandice Nealsu and Jacques saint- Michel
Sept/Oct 1999. “Transient voltage distribution in inverter fed motor
[7] T.A. Lipo and A. Consoli, “Modeling and windings: Experimental study and Modeling.” IEEE
simulation of induction motors with saturable leakage transactions on Power Electronics, vol.16, No.1., Jan.
reactances,” IEEE Trans. Ind. Applications, vol. IA-20, 2000.
pp. 180-198,jan./Feb.1984.
[15] Robert T. Novotnak, Member, IEEE, John Chiasson,
[8] S.I.Moon and A.Keyhani, “Estimation of induction Member, IEEE, and Marc Bodson, Senior Member,
machine parameters from standstill domain data,” IEEE,”High – Performance Motion control of an
IEEE Trans.Ind. Applications, vol.30 pp. 1609-1615. Induction motor with Magnetic saturation. IEEE
Nov/Dec. 1994. transactions on Control systems Technology, vol.7,
No.3., May. 1999.
[9] Seung-Iii Moon,Ali Keyhani,Srinivas Pillutla,
“Nonlinear Neural Network Modeling of an Induction [16] B.J.A. Krose and P.Patrick van der smagt, An
Machine”, IEEE Trans. On Control systems Introduction to Neural Networks, The University of
technology, vol.7,no2, pp. 203-211. March1999. Amsterdam, The Netherlands, Sept. 1991.
[10] “The MATLAB compilers user’s guide” in Math
works hand book Math works 1994.
[11] Pawel z. Grabowski, Marian.P. Kazmierkowski,
B.K.Bose and Frede Blaabjerg’” A simple Direct
Torque Neuro-Fuzzy control of PWM inverter Fed
Induction Motor Drive.
[12] G.J.Rogers and D.Shirmohammdi, “Induction
machine modeling for electromagnetic transient