Design and implementation of a Direct Torque Controller ... · and new type of AC machine drive control technology developed after vector control technology. ... dimension leads to

International Journal of Applied Engineering Research ISSN 0973-4562 Volume 14, Number 22 (2019) pp. 4181-4187

© Research India Publications. http://www.ripublication.com

4181

Design and implementation of a Direct Torque Controller for a Three Phase

Induction Motor based on DSP

Tesfaye Meberate Anteneh 1, Ayodeji Olalekan Salau 2, Takele Ferede Agajie 3, Engidaw Abel Hailu 4

1, 3, 4 Department of Electrical and Computer Engineering, Debre Markos University, Debre Markos, Ethiopia

2 Department of Electrical and Computer Engineering, Afe Babalola University, Ado-Ekiti, Nigeria

[email protected] 1, [email protected] 2, [email protected] 3, [email protected] 4

Corresponding Author: Tesfaye Meberate Anteneh, Email: [email protected]

NB: Authors declare no funding and conflict of interest with regards to this manuscript.

Abstract

Direct torque control (DTC) technology is a superior, modern,

and new type of AC machine drive control technology

developed after vector control technology. In this paper, a DTC

scheme for Induction Motor Drive using Space Vector

Modulation (SVM) technique is implemented using Texas

instrument TMDSHVMTRPFCKIT. This scheme is based on

directly controlling the instantaneous value of the

electromagnetic torque and stator flux-linkage through the

control of stator flux linkage and electromagnetic torque in a

stationary coordinate system. This has the advantage of rapid

torque response, simple control structure, and easy

digitalization. The software inside the code composer studio is

composed of a two-level hysteresis comparator for stator flux

control, a three-level hysteresis comparator for torque control,

a switching table for voltage vector selection, stator flux

position identifier, a three-phase voltage source inverter (VSI)

for generation of pulse width modulated (PWM) signals, flux

and torque estimators, and proportional-integral (PI) controller

for speed estimator. The hardware circuit is composed of a

three-phase induction machine, TMDSHVMTRPFCKIT with

other necessary equipment. The switching table is employed for

selecting the optimum inverter output voltage vectors so as to

attain a fast torque response, low inverter switching frequency,

and low harmonic losses in the code. The

TMDSHVMTRPFCKIT with TMS320F28035 control card is

programmed using code composer studio for the hardware

implementation of the DTC scheme. Steady state estimated

stator flux module was observed to fluctuate from the reference

with an error of 2.8% maximum ripple.

Key words: Direct torque control, code composer studio, flux

and torque estimator, speed estimator, voltage source inverter,

induction motor.

INTRODUCTION

In this age of rapid technological advancement, induction

motors (IMs) are a widely used technology in most electronic

applications. The control methods for IMs can be divided into

two parts: vector control and scalar control strategies. Many

applications require a variable motor speed with constant

torque. The speed of an induction motor (IM) can be changed

in different ways which include: Changing the number of pole

pairs, varying the magnitude of the supplied voltage or varying

the frequency of the supplied voltage [1, 2]. Changing the

number of pole pairs causes huge curve shifting, while varying

the magnitude of the supplied voltage gives a solution only in

a small range. The most promising possibility to change the

speed of the motor is varying the motors frequency. However,

this is not the best solution as the motors impedance increases

and the currents decrease with a variation in frequency. This

offers a possibility to control the motor through changes in both

the voltage and the frequency. This idea is the basis of the

control method (Volts/Hertz constant method) which is capable

of controlling both speed and torque of an IM [3]. Scalar

controlled drives (SCD) give a somewhat inferior performance

but are easy to implement. The importance of SCD has

diminished in recent times because of the superior performance

of vector-controlled drives which are in high demand in many

applications. Field-oriented control (FOC) is one of the vector

control methods in which the stator currents of a three-phase

AC electric motor is identified as two orthogonal components.

One component defines the magnetic flux of the motor, while

the other defines the torque. FOC is classified as Direct Field

Oriented Control (DFOC) and Indirect Field Oriented Control

(IFOC). The FOC method has been shown to be an effective

method but it suffers from some drawbacks such as the

requirement of coordinate transformations, current controllers,

sensitivity to parameter variations, switching frequency, rotor

position measurement, and control tuning loops [4]. Similarly,

the use of speed encoder is associated with some drawbacks

such as the requirement of shaft extension, reduction of

mechanical robustness of the motor drive which is costlier,

reduces the drives reliability, and is not suitable for hostile

environments [5]. The speed control system of the drive

calculates the corresponding current and torque references from

the flux. Proportional-integral (PI) controllers are used to keep

the measured current components at their reference values. The

high-performance frequency controlled PWM inverter fed IM

drive should be characterized by quick flux and torque

mailto:[email protected]

mailto:[email protected]



4182

response, constant switching frequency, unipolar voltage

PWM, low flux ripple and low torque ripple and robustness for

parameter variation [6]. The key to high-performance AC

motor drive is the instantaneous torque control. These features

depend on the applied control strategy. The main goal of the

chosen a control method is to provide simplicity (simple

algorithm, simple tuning, and operation with small controller

dimension leads to the low price of the final product). Direct

torque control (DTC) is another torque control method

following vector control. Vector control is to control the torque

by controlling the decoupled current. DTC directly focuses on

the control of the torque, and this “direct control” concept is not

only used for torque control, but also for the flux chain control

[7-9].

DIRECT TORQUE CONTROL IMPLEMENTATION

Currently, DTC is implemented by one of the most well-known

drive manufacturing companies [10]. The important elements

of the control structure are the power supply circuit, a three-

phase voltage source inverter, three phase IM, the speed

controller to generate the torque command and the DTC

controller. The DTC controller again consists of torque and flux

estimation block, two hysteresis controllers and sector selection

block; the output of the DTC controller is the gating pulses for

the inverter which can generating the PWM signals. A typical

DTC system control principle is shown in Figure 1.

During DTC implementation, the feature considered are the

motor stator flux linkage and torque are analyzed in stationary

reference frame fixed to stator based on the concept of space

vector. It eliminates complicated transformations and

calculations such as vector rotation transformations. Therefore,

the signal processing required by it is particularly simple and

the control signals used to make it possible for the observer to

make a direct and explicit determination of the physical process

of the AC motor. In addition to above the stator flux linkage is

used for the magnetic field orientation. As long as the stator

resistance is known, it can be observed and the control structure

is simple and easy to implement.

Figure 1. Diagram of the implemented DTC scheme for IM drives.

The implemented DTC in Figure 1 consist of a rectifier, voltage

source inverter, driver protection circuit, operational amplifiers

used as a current sensor, voltage divider used as a dc bus

voltage sensor. Some of these hardware components are found

in the TMDSHVMTRPFCKIT. The other part comprise of the

software design in which the DTC algorithm is implemented.

This function is achieved by the TMS320F28035 Control Card.

Furthermore, ADC and DAC conversions are achieved using C

programming language with code composer studio. Also, we

ran the TMS320f28035 controller in Real-time to get a better

results without using floating variables. The variables used in

this implementation are called Integer – Quotient (IQ)

variables. These variables are efficient for motor control

applications. All control algorithms in this kit work based on

the IQ math variables and in per unit system (PU). Stator flux

estimators are used to control the speed of IMs. This estimation

is based on two approaches, namely: the current model, and the

voltage model of IMs.

i. Stator Flux Estimators

The stator winding of an IM is governed by Equations (1) to

(4). Electromagnetic quantities in the equations are represented



4183

with complex space-vector notation, namely: currents (is),

voltages (vs) and fluxes ( ). The voltage model equation is

represented by:

( )( ) ( )s

s s s

d tv t R i t

dt

(1)

( ) [ ( ) ( )]d s ss

d s dt v t R i t dt (2)

( ) [ ( ) ( )]s s sq q s qt v t R i t dt (3)

where Equations (2) and (3) represent the stator and rotor

fluxes, respectively.

In a digital implementation, backward approximation is one of

the methods to change continuous signals to discrete. Replace 1

𝑠 term (the integral term) to discrete using the relation

1

𝑠 =

𝑇𝑠𝑍

𝑍−1.

This gives Equation (4).

1

1

( )

( )

s s s s

s s s s

s

q s q s q q

s

d s d s d d

T v R i

T v R i

(4)

After sampling was performed, we use a lowpass filter with

cutoff frequency 3𝑟𝑎𝑑

𝑠𝑒𝑐 to filter out noise.

Figure 2. Low pass filter for stator flux estimation.

The transfer function of the lowpass filter used is 1

1+𝑠𝜏𝑐 , where

𝜏𝑐 =1

2𝜋𝑓𝑐.

ᴪ𝑑�̂� =1

1+𝑠𝜏𝑐ᴪ𝑑𝑠 (5)

ᴪ𝑞𝑠̂ =1

1+𝑠𝜏𝑐ᴪ𝑞𝑠 (6)

Also, using backward approximation technique to replace 1

𝑠 =

𝑇𝑠𝑍

𝑍−1 in the stator flux estimation gives Equation (7).

2 2( ) ( ( )) ( ( ))d ss

s qt t t (7)

( )tan 2

( )

s

d s

q

s

ta

t

(8)

ii. Torque Estimation

Torque estimation can be obtained using Equation (9).

3

( )2 2 s s s se d q q d

pT i i (9)

iii. Speed Estimation

Equations (10) and (11) are used for the estimation of speed in

the IM.

( )( tan 2 )

( )

s

d s

qssyn

td da

dt dt t

(10)

2

( ) ( ) ( ) ( )d s ss s

q q d

syn

s

d dt t t t

dt dt

(11)

Equation 11 does not give an exact estimate, rather it gives a

disturbed higher frequency noise signal due to modeling errors.

During digital implementation, this can be removed by using a

first order low pass filter [6]. In this stage, a low pass filter with

cutoff frequency 5 rad/sec was used as shown in Figure 3.

Figure 3. Low pass filter for synchronous speed estimation.

The equations of operation of the low pass filter are given by

Equations (12) to (16).

𝜔𝑠𝑦�̂� = 𝜔𝑠𝑦𝑛1

1+𝑠𝜏𝑐 (12)

c synssyn syn

(13)

1

s c

syn

s c s c

Tsynsyn T T

(14)

2

2

3

r estimated

slip

r

R T

polepair

(15)

r syn slip (16)



4184

EXPERIMENTAL SETUP, RESULTS AND

DISCUSSIONS

In this section, we present the experimental results obtain after

the experiment was performed. The experimental setup is

shown in Figure 4.

Figure 4. Hardware setup of DTC based induction motor drive.

The experimental results were obtained after the dc bus voltage

was set to 120V with a reference speed of the step input given

as 1080 rpm rotor speed (0.3pu) and 0.1𝑣𝑜𝑙𝑡−𝑠𝑒𝑐

𝑟𝑎𝑑 stator flux

reference. All the results were captured with code composer

studio when the motor was running at the reference operating

points. In Figures 5 and 6, the values on the horizontal axis

represent sampling frequency, while the vertical axis represent

the stator current.

Figure 5. Direct axis stator current.

Figure 5 shows the direct axis stator current of the IM flows at

a steady rate which doesn’t exceed 0.002pu. This shows the IM

is operating at its normal condition. Similarly, this is seen with

the Quadrature axis stator current as shown in Figure 6.

Figure 6. Quadrature axis stator current.

The results of Figures 5 and 6 were obtained after the present

state of stator winding phase A current and stator winding

phase B current using OPA2350 and using ADC register were

measured. These results are converted to two phase orthogonal

stationary stator winding current (d-q) through a transformation

in the code. The d-q currents are used for estimation purpose.

The d-q voltage in Figures 7 and 8 were obtained from the

reconstructed stator phase voltage by using transformation. The

reconstructed stator phase voltage can also be obtained from

sensed dc bus voltage and the corresponding phase switching

function (1 or 0) which comes from the space vector PWM.

Figure 7. Direct axis stator phase voltage.

Figure 8. Quadrature axis stator phase voltage.

Figures 9 and 10 were obtained based on simulations performed

using Equations 5 and 6. From Figures 5-10, the real-time

graphs of direct axis and quadrature axis variables (current,

voltage, and flux) waveforms are sinusoidal with a varying

amplitude. This proves the reference frame is a stationary

reference frame. In addition, the angle between the direct axis



4185

and quadrature axis is 900 phase shift. Furthermore, it was

observed that waveforms of direct axis stator flux and

quadrature axis stator flux has a 900 shift with a sinusoidal type

waveform with an amplitude of the reference flux value (the

same magnitude). From the real time code composer studio

graph shown in Figures 5-10, we obtain each variable values of

the motor by multiplying the base values of each variables with

their corresponding per unit values from the real time graph as

shown in Figures 5-10.

Figure 9. Direct axis stator flux.

Figure 10. Quadrature axis stator flux.

The estimation of the stator flux angle is seen to vary from 0

rad (00) to 2𝜋 𝑟𝑎𝑑 (3600). Since our analysis is in per unit,

Atan2pu function returns the value 0 to 1 by considering 2𝜋 rad

as a base. But when per unit analysis is not considered, this

function returns the value from - 𝜋 to 𝜋. This means it covers

all four quadrants of the angle. From Figures 11 and 12, the

real-time graph indicates the value of the estimated resultant

stator flux (the square root of the estimated direct and

quadrature axis stator flux). This value is around the reference

resultant stator flux value 0.1Wb, but not exactly 0.1Wb as

shown in Figure 13 because it has some fluctuation from the

reference value with a maximum ripple of 2.8%. I.e. maximum

experimental result -reference value is divided by the reference

value.

Figure 11. Stator flux angle.

Figure 12. Estimated stator flux modulus.

Figure 13. Estimated stator flux in zoom view.

In Figure 14, the graph of the direct axis rotor flux in the

stationary reference frame is presented. The rotor flux is seen

to fluctuate in a sinusoidal manner. The rotor speed is seen to

also change drastically as shown in Figure 15. This eventually

causes the IM torques speed to also increase accordingly as

shown in Figure 16. The specifications of the squirrel cage IM

used are presented in Table 1 and the Base values for the IPM

voltage source inverter and induction motor are presented in

Table 2.

Figure 14. Direct axis rotor flux in the stationary reference

frame.

Figure 15. Estimated rotor speed.



4186

Figure 16. Developed electromagnetic torque.

Table 1. System specification of squirrel cage induction

motor.

Parameters Values

Rated power (P) 370W

Stator resistance (Rs) 11.05Ω

Rotor resistance (Rr) 6.11Ω

Stator linkage inductance (Ls) 0.316423H

Rotor linkage inductance (Lr) 0.316423H

Mutual inductance (Lm) 0.293939H

Moment of inertia (J) 0.009 kg.m2

Number of poles (p) 4

Stator rated current (is) 1.1A

Rated frequency (f) 50Hz

Rated electrical (mechanical)

rotor speed

2640 (1320) rpm

Rated stator phase rms voltage 220V

Table 2. Base values for the IPM voltage source inverter and

induction motor.

Base parameters Values

Base phase voltage 220V

Base phase current 10A

Base flux 2.93939 volt-sec/rad

Base frequency 50Hz

Base torque 22.604485 N.M

Switching frequency of IPM 10kHz

Base Dc Bus Voltage 300V

CONCLUSIONS

In this paper, we have presented the implementation of direct

torque-controlled IM drive without using a speed sensor (using

speed estimation techniques). The stationary frame of reference

is used for our analysis using code composer studio c

programming direct torque-controlled drive algorithm.

Experimental results show that the DTC system based on the

voltage model of flux linkage achieves the rotation speed of the

stator flux linkage by applying the voltage vector while

maintaining the constant value of the stator flux linkage.

Although, DTC technology is a new type of modern AC motor

control technology with numerous advantages, there are still

many problems to be solved in the future either in theory or

practice.

REFERENCES

[1] Datta, R. and Ranganathan, V. T., 2001, “A simple

position-sensorless algorithm for rotor-side field-

oriented control of wound-rotor induction

machine,” IEEE Transactions on Industrial

Electronics, 48(4), pp. 786-793. DOI:

10.1109/41.937411

[2] Toh, C. L., Idris, N. R. N., Yatim, A. H. M. and Patkar,

F., 2005, “Design and implementation of a direct

torque control of induction machine utilizing a digital

signal processor and the field programmable gate

arrays,” International Conference on Power

Electronics and Drives Systems, Kuala Lumpur, pp.

72-77. DOI: 10.1109/PEDS.2005.1619663

[3] Levi, E., Aware, M. V., Pandit, J. K. and Nemade, R.,

2017, “Direct torque control scheme for a six-phase

induction motor with reduced torque ripple,” IEEE

Transactions on Power Electronics, 32 (9), pp. 7118-

7129.

[4] Muchande, S., Kadam, A., Unni, K. and Thale, S.,

2013, “Design and implementation of a direct torque

control space vector modulated three phase induction

motor drive,” In: Unnikrishnan S., Surve S., Bhoir D.

(eds) Advances in Computing, Communication, and

Control. Communications in Computer and

Information Science, 361. Springer, Berlin,

Heidelberg.

[5] BabaLawan, M., Samaila, Y. A., Tijjani, I., Mustapha,

M. and Sarki, M. A., 2019, “Review of the rotor

position and speed estimation method of induction

motor drives,” ASPL International Journal of

Information and Technology, 8(1), pp. 74 – 91.

[6] Kumar, D. and Thakur, I., 2014, “A review on direct

torque control for induction motor,” International

Journal of Science and Research (IJSR), 3(7), pp. 694-

698.

[7] Ozkop, E. and Okumus, H. I., 2008, “Direct torque

control of induction motor using space vector

modulation (SVM-DTC),” 12th International Middle-



4187

East Power System Conference, Aswan, pp. 368-372.

DOI: 10.1109/MEPCON.2008.4562350

[8] Toh, C. L., Idris, N. R. N. and Yatim, A. H. M., 2009,

“Design and implementation of TMS320C31 DSP and

FPGA for conventional direct torque control (dtc) of

induction machines,” International Scholarly and

Scientific Research and Innovation, 3(2), pp. 225- 230.

[9] Abbou, A., Sayouti, Y., Mahmoudi, H. and Akherraz,

M., 2010, “dSPACE direct torque control

implementation for induction motor,” 18th

Mediterranean Conference on Control and

Automation, MED'10, Marrakech, pp. 1121-1126.

DOI: 10.1109/MED.2010.5547641

[10] Stojic, D. M., 2012, “An algorithm for induction

motor stator flux estimation,” Advances in Electrical

and Computer Engineering, 12(3), pp. 47-52. DOI:

10.4316/AECE.2012.03007

Documents

Design and implementation of a Direct Torque Controller ... · and new type of AC machine drive control technology developed after vector control technology. ... dimension leads to