Proposed Voltage Vector to Optimize Efficiency of Direct Torque Control

International Journal of Power Electronics and Drive System (IJPEDS)
Vol. 4, No. 4, December 2014, pp. 578~586
ISSN: 2088-8694  578
Journal homepage: http://iaesjournal.com/online/index.php/IJPEDS
Proposed Voltage Vector to Optimize Efficiency of Direct
Torque Control
Goh Wee Yen*, Ali Monadi*, Nik Rumzi Nik Idris*, Auzani Jidin**, Tole Sutikno**
*Department of Electrical Power Engineering, Universiti Teknologi Malaysia (UTM), Johor Bahru, Malaysia
** Department of Power Electronics and Drives, Universiti Teknikal Malaysia Melaka (UTeM), Malacca, Malaysia
*** Department of Electrical Engineering, Universitas Ahmad Dahlan (UAD), Yogyakarta, Indonesia
Article Info ABSTRACT
Article history:
Received Aug 22, 2014
Revised Oct 26, 2014
Accepted Nov 6, 2014
Compared to field-oriented control (FOC) system, direct torque control
(DTC) system has gained attractiveness in control drive system because of its
simpler control structure and faster dynamic control. However, supplying the
drive system with rated flux at light load will decrease the power factor and
efficiency of the system. Thus, an optimal flux has been applied during
steady-state in order to maximize the efficiency of drive system. But when a
torque is suddenly needed, for example during acceleration, the dynamic of
the torque response would be degraded and it is not suitable for electric
vehicle (EV) applications. Therefore, a modification to the voltage vector as
well as look-up table has been proposed in order to improve the performance
of torque response.
Keyword:
Direct torque control
Induction machine
Proposed voltage vector
Search controller
Torque response Copyright © 2014 Institute of Advanced Engineering and Science.
All rights reserved.
Corresponding Author:
Nik Rumzi Nik Idris,
Department of Electrical Power Engineering,
Universiti Teknologi Malaysia (UTM),
81310 UTM Skudai, Johor, Malaysia.
Email: nikrumzi@fke.utm.my
1. INTRODUCTION
A simple control structure of DTC that has been proposed by Takahashi [1] has gained popularity in
industrial motor drive applications. Due to its simpler control structure and faster dynamic control compared
to FOC system, the popularity of DTC system is increased rapidly in the past decades [1-3]. In FOC, the
torque and flux are controlled based on stator current components whereas in DTC, the torque and flux are
controlled directly and independently via an optimized selection of voltage vectors using look-up table.
As illustrated in Figure 1, the simple control structure of DTC consists of three-phase voltage source
inverter (VSI), hysteresis comparators, stator flux and torque estimators, as well as look-up table. By using
two-level and three-level hysteresis comparators, the stator flux and electromagnetic torque can be controlled
independently. Based on look-up table, an appropriate voltage vector is selected to satisfy its flux and torque
requirement. The selected voltage vector is then applied to activate the VSI in which it will in turn operate
the induction machine.
A fast instantaneous control of torque and flux occurs because of de-coupled control of torque and
flux, in which it leads to the faster dynamic control of DTC system compared to the FOC system. To achieve
more accurate flux estimation, the current model is applied during low speed operations whereas the voltage
model is employed in high speed operations. Only stator resistance and terminal quantities such as stator
voltages and currents are required in the estimation of voltage model.

IJPEDS ISSN: 2088-8694 
Proposed Voltage Vector to Optimize Efficiency of Direct Torque Control (Goh Wee Yen)
579
Figure 1. Simple Control Structure of DTC
In order to fully utilized the power and lengthen the life span of induction motor, an optimal
efficiency of the drive system is an important factor to be implemented in EV applications. Usually,
induction motors are operated at light load and thus, supplying the motor at its rated flux will decrease the
power factor and efficiency of the drive [4]. Therefore, researchers have been working on the efficiency
optimization of drive system in recent years but there is still no suitable method to achieve the fast
instantaneous torque response of DTC drive.
Two main methods have been proposed to maximize efficiency in DTC drive system. These
methods are known as flux search controller (SC) [4-12] and loss model controller (LMC) [13-16]. The
former method measures input power or stator current of the system while decreasing the flux value in a
consecutive step. When the input power or stator current is at its minimized value, the optimal flux is
obtained. Meanwhile, by applying loss model equations, the optimal flux of latter method is determined.
When copper losses are approximately equal to iron losses, the optimize efficiency of drive system is
achieved.
Instead of just concentrated on searching for the optimal flux, improving the dynamic torque is also
an essential factor to be considered in order to optimize the efficiency of DTC system. This is because
supplying the drive system at its optimum flux will cause the torque response to be degraded when a rated
torque is suddenly needed. Therefore, a modification to look-up table as well as DTC algorithm has been
done so that the dynamic torque is achieved during transient state.
2. EFFECTS OF VOLTAGE VECTOR
The effect of voltage vector on torque response has been studied in order to improve the
performance of torque response, as shown in Figure 2. Based on Figure 2 (b), in sector 4, the voltage vectors,
vs,5 and vs,6, are applied to increase and decrease the stator flux, respectively. In Figure 2 (a), vs,5 is activated
to increase the stator flux and at the same time, it is capable to increase the output torque dynamically. But
activating vs,6 to decrease the flux causes the output torque to increase slightly, and thus, it degraded the
torque performance. This case is worsening when the flux is set to its optimized value for efficiency
purposes.
In sector 4, vs,5 is considered as the most optimized voltage vector compared to vs,6 because it has
larger tangential to the stator flux and consequently, it can produce dynamically torque. Note that, at the very
beginning of sector 4, the response of torque is more dynamic when vs,5 is activated for a longer time because
this voltage vector is tangential to stator flux. Therefore, the voltage vector that is applied to decrease the
stator flux has to be modified so that the proposed voltage vector can produce larger tangential to stator flux
in order to improve torque performance.
Flux and Torque Estimators
(i.e. only using the voltage model)
Switching
Table
VSI
Sector
Te
Te, ref
IM
Sa, Sb, Sc
Ψs, ref
Ψs
ia, ib, ic
va, vb, vc

 ISSN: 2088-8694
IJPEDS Vol. 4, No. 4, December 2014 : 578 – 586
580
(a) (b)
Figure 2. Voltage Vector (a) Effects on Torque Response, and (b) Controlling Stator Flux
3. RESEARCH METHOD
A proposed voltage vector is produced between two conventional voltage vectors by applying
vector’s parallelogram law. Compared to the conventional voltage vector, the proposed voltage vector has
longer amplitude and an angle of 30° adjacent to the conventional voltage vector. For instance, addition of
vs,5 to vs,6 will obtain vs,5-6, as shown in Figure 3, and their respective equations are calculated in (1).
, , , , , , 1

Figure 3. Vector’s Parallelogram Law
In Figure 4, the red line indicates the proposed voltage vector whereas the black line represents the
conventional voltage vector. In proposed method, vs,5-6 is activated instead of vs,6 when the flux is required to
be reduced. The proposed voltage vector has a larger tangential to stator flux in which it is believed to
improve the torque performance when decreasing the flux. In conventional DTC system, the switching of
voltage vector is more regular in the middle of a sector compared to the beginning and end of a sector.
Consequently, activating the proposed voltage vector also reduces the switching of voltage vector when it is
in the middle of a sector and thus, it increases the torque dynamically.
30°
vs,5
vs,6
vs,5-6
Sector
1
Sector
2
Sector
3
Sector
4
Sector
5
Sector
6
Ψs
δsr
Ψrvs,6
vs,5

581

Figure 4 Proposed Voltage Vector
By applying the proposed method, the amplitude of proposed voltage vector is almost twice of the
amplitude of conventional method. The increasing amplitude of voltage vector will cause the value of stator
flux and output torque to be increased as well. Therefore, the amplitude of proposed voltage vector can be
reduced by estimating a ratio between proposed and conventional voltage vector, as follows:
,
,
                                                                                2
where Vs,k-k is the amplitude of proposed voltage vector and Vs,k is the amplitude of conventional voltage
vector.
After introducing the estimated ratio, it can be seen that the d-q axis of proposed voltage vector is
exchanged with the d-q axis of conventional voltage vector. Therefore, the d-q axis of proposed voltage
vector is given in (3) and (4):
,
2
3
0.866 0.866                                                             3
,
2
3
0.5 0.5                                                             4
The switching state of VSI can be implemented in the look-up table with modified DTC algorithm
since d-q axis of proposed voltage vector is exchanged with d-q axis of conventional voltage vector. The
proposed look-up table with modified DTC algorithm is shown in Table 1. In order to improve the dynamic
of output torque, the modified look-up table with DTC algorithm is implemented only during transient state.
Meanwhile, the conventional look-up table is applied during steady-state.
Table 1 Modified Look-Up Table
Stator Flux
Error Status,
Ψs
+
Torque
Error
Status,
Tstat
Sector
1
Sector
2
Sector
3
Sector
4
Sector
5
Sector
6
1
1 vs,2 vs,3 vs,4 vs,5 vs,6 vs,1
0 vs,0 vs,7 vs,0 vs,7 vs,0 vs,7
-1 vs,5-6 (vs,6) vs,1-6 (vs,5) vs,1-2 (vs,4) vs,2-3 (vs,3) vs,3-4 (vs,2) vs,4-5 (vs,1)
0
1 vs,2-3 (vs,3) vs,3-4 (vs,2) vs,4-5 (vs,1) vs,5-6 (vs,6) vs,1-6 (vs,5) vs,1-2 (vs,4)
0 vs,7 vs,0 vs,7 vs,0 vs,7 vs,0
-1 vs,5 vs,6 vs,1 vs,2 vs,3 vs,4
Sector IV
vs,5
vs,6
vs,5-6
Hysteresis Band

 ISSN: 2088-8694
582
As shown in Table 1, the respective proposed voltage vector can be obtained when the voltage
vector in bracket is activated. In other words, the voltage vector in bracket indicates the switching state of the
respective proposed voltage vector. For example, in sector 1, vs,2 is applied to increase the flux and vs,2-3 is
activated to decrease the flux. Both of these voltage vectors are capable to increase the output torque. But in
order to activate vs,2-3, the switching state of vs,3 has to be implemented.
4. RESULTS AND ANALYSIS
The simulation of DTC drive system has been constructed using MATLAB’s SIMULINK blocks, as
shown in Figure 5. The specifications and parameters of induction machine used in the simulation are given
in Table 2. The modified look-up table with DTC algorithm has been attached in parallel to the conventional
look-up table with DTC algorithm. The modified look-up table with DTC algorithm has been activated for
5ms only during transient state whereas during steady-state, the conventional look-up table with DTC
algorithm has been implemented.
Figure 5. Simulation of DTC Drive System
Table 2 Specifications and Parameters of Induction Motor
Parameters Values
DC voltage 340V
Stator resistance, Rs 0.25Ω
Rotor resistance, Rr 0.2Ω
Stator inductance, Ls 0.0971H
Rotor inductance, Lr 0.0971H
Mutual inductance, Lm 0.0955H
Frequency,f 50Hz
Inertia motor, J 0.046kgm2
Pole pairs, p 2
Sampling time 50µs
Rated flux 1.04Wb
Rated torque 150Nm
In order to optimize the efficiency of drive system, the SC method has been implemented in DTC
drive system. The SC is activated at t=1s with step flux of 0.043Wb, sample time of 0.1s and at its rated
speed. Basically, the rated flux is applied to the system and after the system has reached its steady-state, the
corresponding current value is measured. Then, the flux is decreased with step flux and the corresponding
current value is measured again. When the new stator current (Is, k) is smaller than the previous stator current
(Is, k-1), the flux value is decreased with step flux, and vice-versa. The SC method is continuing until it

583
reaches its optimum flux value. From Figure 6, the optimal flux is obtained after 0.3s with flux value of
0.89Wb.
Figure 6. Flux Search Controller
3.1. Voltage Vector
As discussed earlier, the conventional and proposed voltage vector has been proven in Figure 7. The
blue line indicates the conventional voltage vector whereas the red line denotes the proposed voltage vector.
Based on this figure, the proposed voltage vector has same amplitude as the conventional voltage vector and
it is 30° adjacent to the conventional voltage vector.

Figure 7. Voltage Vector
3.2. Torque Performance
In order to examine the effectiveness of proposed voltage vector towards the dynamic torque at light
load, a step flux and torque is applied from 0.89Wb to 1.04Wb and from 0Nm to 150Nm respectively at
beginning, middle and end of a sector, as shown in Figure 8. In Figure 8 (a) and (b), the stator flux of
proposed method is increased beyond rated flux for a short duration because d-axis stator flux is slightly
decreased and q-axis stator flux is slightly increased compared to conventional method. However, the output
torque is not affected by the slightly increased of stator flux. Meanwhile, in Figure 8 (a), the output torque is
decreased to 120Nm at t=0.505s because the flux is decreased to its rated value.
vs,5-6
vs,1-6
vs,6 vs,1
vs,2-3
vs,1-2
vs,2
vs,3
vs,3-4
vs,4
vs,5
vs,4-5

 ISSN: 2088-8694
584
At beginning of a sector, the conventional and proposed method achieved the rated torque in 2ms.
But when a step torque is applied at the middle of a sector, the conventional method requires 2.8ms to
achieve its rated torque whereas the proposed method needs 2.4ms to attain its rated torque. The proposed
method has improved the performance of torque response by 0.4ms compared to conventional method.
Meanwhile, at the end of a sector, the conventional and proposed method requires 3.5ms and 3.0ms to reach
their rated torque, respectively. By implementing the proposed method, the torque response can be improved
by 0.5ms.
(a) (b)
(c)
Figure 8 Torque Performance on (a) Beginning, (b) Middle, and (c) End, of a Sector

585
From Figure 8 (a), the applied voltage vector is activated for a longer period since the voltage vector
is tangential with respect to stator flux. As mentioned earlier, only the voltage vector that is used to decrease
the stator flux is modified; hence, the conventional and proposed method applied the same voltage vector to
increase flux in which it caused the conventional and proposed method to reach rated torque at the same time.
Based on Figure 8 (b) and (c), it can be seen that the torque performance is improved when the voltage vector
is activated for a longer period, in which it can be controlled by generating a larger tangential to the stator
flux. Besides that, voltage vector that has larger tangential respective to stator flux is able to reduce the
switching of voltage vector. Consequently, it is necessary to modify the angle of voltage vector so that a
dynamic torque can be produced during transient state.
5. CONCLUSION
During steady-state, the flux has to be set to an optimum value in order to optimize the efficiency of
DTC drive system. However, the dynamic of output torque would be degraded when a torque is suddenly
needed and it is not suitable to be implemented in EV applications. Therefore, an adjustment to the look-up
table as well as DTC algorithm has been constructed by modifying the angle of voltage vector so that a larger
tangential with respect to the flux is yielded. Based on the results, it can be concluded that the proposed
voltage vector improves the performance of torque response during transient state. Therefore, this method is
believed to optimize the efficiency of DTC drive system and at the same time, the dynamic of torque
response is improved.
ACKNOWLEDGEMENTS
The authors would like to express their appreciation to Universiti Teknologi Malaysia (UTM) for
providing Zamalah’s Scholarship and Ministry of Education for fund research grant (R.J130000.7823.4F380)
in this research.
REFERENCES
[1] I. Takahashi and T. Noguchi, "A New Quick-Response and High-Efficiency Control Strategy of an Induction
Motor," Industry Applications, IEEE Transactions on, vol. IA-22, pp. 820-827, 1986.
[2] T. Abe, T. G. Habetler, F. Profumo, and G. Griva, "Evaluation of a high performance induction motor drive using
direct torque control," in Power Conversion Conference, 1993. Yokohama 1993., Conference Record of the, 1993,
pp. 444-449.
[3] I. Takahashi and T. Noguchi, "Take a look back upon the past decade of direct torque control [of induction
motors]," in Industrial Electronics, Control and Instrumentation, 1997. IECON 97. 23rd International Conference
on, 1997, pp. 546-551 vol.2.
[4] S. Kaboli, E. Vahdati-Khajeh, M. R. Zolghadri, and A. Homaifar, "A fast optimal flux search controller with
improved steady state behavior for DTC based induction motor drives," in Electric Machines and Drives, 2005
IEEE International Conference on, 2005, pp. 1732-1736.
[5] S. Kaboli, M. R. Zlghadri, and A. Emadi, "A Fast Flux Search Controller for DTC Based Induction Motor Drives,"
in Power Electronics Specialists Conference, 2005. PESC '05. IEEE 36th, 2005, pp. 739-744.
[6] S. Kaboli, M. R. Zolghadri, and E. Vahdati-Khajeh, "A Fast Flux Search Controller for DTC-Based Induction
Motor Drives," Industrial Electronics, IEEE Transactions on, vol. 54, pp. 2407-2416, 2007.
[7] S. Kaboli, E. Vahdati-Khajeh, M. R. Zolghadri, and A. Homaifar, "on the Performance of Optimal Flux Search
Controller for DTC Based Induction Motor Drives," in Electric Machines and Drives, 2005 IEEE International
Conference on, 2005, pp. 1752-1756.
[8] S. Kaboli, M. R. Zolghadri, D. Roye, and A. Emadi, "Online optimal flux controller for DTC based induction
motor drives," in Industrial Electronics Society, 2004. IECON 2004. 30th Annual Conference of IEEE, 2004, pp.
1391-1395 Vol. 2.
[9] N. Sadati, S. Kaboli, H. Adeli, E. Hajipour, and M. Ferdowsi, "Online Optimal Neuro-Fuzzy Flux Controller for
DTC Based Induction Motor Drives," in Applied Power Electronics Conference and Exposition, 2009. APEC 2009.
Twenty-Fourth Annual IEEE, 2009, pp. 210-215.
[10] S. Vamsidhar and B. G. Fernandas, "Design and development of energy efficient sensorless direct torque controlled
induction motor drive based on real time simulation," in Industrial Electronics Society, 2004. IECON 2004. 30th
Annual Conference of IEEE, 2004, pp. 1349-1354 Vol. 2.
[11] G. Calzada-Lara, F. Pazos-Flores, and R. Alvarez-Salas, "A new Direct Torque Control for a better efficiency of
the induction motor," in Power Electronics Congress (CIEP), 2010 12th International, 2010, pp. 78-83.
[12] I. Kioskeridis and N. Margaris, "Loss minimization in scalar-controlled induction motor drives with search
controllers," Power Electronics, IEEE Transactions on, vol. 11, pp. 213-220, 1996.
[13] D. Gan and O. Ojo, "Efficiency Optimizing Control of Induction Motor Using Natural Variables," Industrial
Electronics, IEEE Transactions on, vol. 53, pp. 1791-1798, 2006.

 ISSN: 2088-8694
586
[14] X. Zhang, H. Zuo, and Z. Sun, "Efficiency optimization of direct torque controlled induction motor drives for
electric vehicles," in Electrical Machines and Systems (ICEMS), 2011 International Conference on, 2011, pp. 1-5.
[15] G. Bhuvaneswari and A. P. Satapathy, "ANN based optimal flux determination for efficiency improvement in
Direct Torque controlled induction motor drives," in Power and Energy Society General Meeting, 2010 IEEE,
2010, pp. 1-6.
[16] I. Kioskeridis and N. Margaris, "Loss minimization in induction motor adjustable-speed drives," Industrial
Electronics, IEEE Transactions on, vol. 43, pp. 226-231, 1996.

Proposed Voltage Vector to Optimize Efficiency of Direct Torque Control

More Related Content

What's hot (18)

Similar to Proposed Voltage Vector to Optimize Efficiency of Direct Torque Control (20)

More from IAES-IJPEDS (20)

Recently uploaded (20)

Proposed Voltage Vector to Optimize Efficiency of Direct Torque Control