DSP Implementation of an Improved DTC Technique for Induction ...

Abstract
DSP Implementation of an Improved DTC Technique for
Induction Motor Drives
H.I. Okumus & D. Holliday
University of Bristol
Department of Electrical & Electronic Engineering
University Walk, Queens Building
BS8 1TR Bristol, UK
H.I.Okumus@bristol.ac.uk Fax: + 44 (0) 117 954 5206
D.Holliday@bristol.ac.uk Fax: + 44 (0) 117 954 5206
A simple stator resistance voltage-drop compensation method that eliminates the stator resistance voltage-drop at
low speed in a Direct Torque Controlled induction motor drive is proposed. The resulting “stator voltage drop
feedback” is used to correct the stator flux estimate resulting in a reduced low speed flux ripple. The performance
of the technique is analysed by simulation and validated by experiments. A low-cost, commercial TMS320C31
DSP board is used for the implementation of the control algorithm.
Introduction
Technological improvements in power semiconductor and microprocessor technology have facilitated the
application of advanced control techniques to ac motor drive systems [1]. There exist two main types of highperformance
torque controlled ac drives: Vector Controlled drives and Direct Torque Controlled (DTC) drives [2].
Although the principles of DTC were introduced more than ten years ago [3,4] it is only recently that industrial
DTC drives have become available [5].
Direct Torque Control (DTC) has received the attention of motor drive designers due to its relatively simple
implementation. The strategy also has the advantage that it does not require a rotor position sensor. Recent trends
and developments are such that speed and position sensorless high dynamic performance drives are attracting
considerable interest due to factors such as increased reliability and cost reduction which can result from the
elimination of the speed transducer [5].
The DTC technique requires a measurement of the machine’s stator resistance to facilitate the estimation of the
stator flux. An accurate estimate of flux linkage is possible at high operating speed [6] but is less accurate at low
speed. This paper describes a simple technique that eliminates the stator resistance voltage-drop thereby
improving the estimated value of flux linkage, particularly at low speed. The performance of a DTC drive
incorporating the “voltage drop feedback” method is analysed by simulation and validated by experiment.
General Description of Direct Torque Control
A block diagram of the DTC scheme is presented in Figure 1. DTC comprises three basic functions: hysteresis
control for torque and flux, an optimal switching vector look-up table and a motor model. The motor model
calculates the actual torque, stator flux and shaft speed based on the measurements of two stator phase currents and
the dc link voltage. Torque and flux references are compared with the actual values and control signals are
produced by using a torque (three-level) and flux (two-level) hysteresis control method.
ω ref
Teref
ψ ref
Speed
Control
+
ω r
-
+
+
-
Te
-
ψ act
Flux
Control
Torque
Control
θψ
dψ
dT
Switching
Table
Flux and
Torque
Estimator
Figure 1 DTC Controlled Voltage Source Inverter Drive
The switching vector look-up table, as shown in Table 1, gives the optimum selection of the switching vectors for
all the possible stator flux-linkage space-vector positions (six positions corresponding to the six sectors). In
1
S
a,
b,
c
Vdc
Voltage
Source
Inverter
IM

switching Table 1, dψ = 1 and dψ = 0 indicates the stator flux needs to be increased and reduced respectively.
Similarly, dT = 1 indicates the motor torque needs to be increased, and dT = 0 and dT = -1 indicate the torque needs
to be reduced slowly and quickly respectively. Speed control is achieved using a PI controller. In a DTC drive, as
shown in Figure 1, the torque reference is also obtained from the PI speed controller.
Table 1 Stator Voltage Switching Table.
Induction Motor Model in the Stationary Reference Frame
The stationary reference frame induction motor model, with stator and rotor fluxes as state variables, is defined by
equation (1) and forms the basis for the simulation results in this paper.
d
dt
⎡
⎢
⎣
s
r
⎡ Rs
⎢ −
⎤
⎢
s
⎥ =
⎦ ⎢ Rr
M
⎢
⎣ s
Lr
Rs
M ⎤
⎥
s
Lr
⎥
⎡
⎛ R ⎢
r ⎞⎥
j ⎣
⎜ r −
⎟
⎟⎥
⎝
r ⎠⎦
s
r
⎤ ⎡1⎤
⎥ + ⎢ ⎥V
⎦ ⎣0⎦
where s and r are stator and rotor flux space vectors, s V is stator voltage space vector, s R and R r are the
stator and rotor resistances, L s and L r are stator and rotor self-inductances, M is the mutual inductance,
2
= 1 - M Ls
Lr
is the leakage coefficient and r
calculated from equations (2) and (3) respectively.
i
i
ds
qs
2
s
is rotor angular speed. Stator dq current components are
ds − md
= (2)
L − M
s
qs − mq
= (3)
L − M
s
X aq X aq
X aq X aq
where, md = ds +
dr , mq =
qs +
qr
L − M L − M
L − M L − M
s
and [ ] 1 −
X = ( 1/M + 1/ ( L − M ) + 1/ ( L − M ) .
aq
dψ dT N1 N2 N3 N4 N5 N6
1 V2(110) V3(010) V4(011) V5(001) V6(101) V1(100)
1 0 V7(111) V0(000) V7(111) V0(000) V7(111) V0(000)
-1 V6(101) V1(100) V2(110) V3(010) V4(011) V5(001)
1 V3(010) V4(011) V5(001) V6(101) V1(100) V2(110)
0 0 V0(000) V7(111) V0(000) V7(111) V0(000) V7(111)
-1 V5(001) V6(101) V1(100) V2(110) V3(010) V4(011)
s
r
r
s
The electromagnetic torque can be expressed in terms of stator flux and current as shown in equation (4)
( i i )
⎛ 3 ⎞⎛
P ⎞
Te = ⎜ ⎟⎜
⎟ ds qs −
⎝ 2 ⎠⎝
2 ⎠
qs ds
where P is the number of poles in the induction motor.
Compensation of the Effect of the Stator Voltage-Drop
A major problem in the application of DTC at low motor speed is the effect of the IsRs voltage drop. In order to
improve flux control at low speed, optimum switching vector selection and a technique that eliminates the effects
of the IsRs voltage drop in the machine stator is therefore considered.
Figure 2 shows the switching voltage space vectors and stator resistance voltage drop. It can be seen that at high
speed, the amplitude of the stator voltage vector sh V is large and the voltage drop across R s is relatively small and
can be neglected, as shown in Figure 2(a). However, it can be seen from Figure 2(b) that, at low speed, the voltage
drop across s R may be the major portion of measured terminal voltage V sl . Moreover, as the temperature of the
r
(1)
(4)

stator windings increases, the value of resistance Rs will change significantly resulting in improper flux estimation
and deterioration in controller performance.
sQ
(a) High Speed (b) Low Speed
Figure 2 Effect of the Voltage Drop on the Stator Voltage Space Vectors.
Flux Estimator with Voltage-Drop Feedback
In classical DTC methods the stator flux is calculated by means of the stator voltage vector and current [7] as
shown in equation (5)
= V R I dt
s
∫ ( − )
(5)
s s s
This gives an accurate flux linkage estimate at high speed. However, as described in the previous section, errors in
the flux estimate increase as motor speed reduces.
To compensate the RsIs voltage drop at low speed, a measurement of the stator winding resistance voltage drop is
incorporated into the flux estimator. It can be seen from Figure 2 that the stator voltage drop vector is the function
of the stator current. The voltage drop vector can therefore be calculated from a present stator current sample and
used as feedback to compensate the present voltage drop at the next sample instant. The improved flux estimator
is described by equation (6).
Simulation Results
vqsh
I
α
v
dsh
V − R I
s
Vsh
sh
α
s s
RsI
s
sD
= V k R I k I k 1 dt
s
∫[
( ) − ( ( ) − ( − ))]
(6)
s s s s
Simulation studies were carried out using the MATLAB software package. Figures 3, 4 and 5 show a comparison
between the proposed compensation strategy and the classical DTC scheme. Torque and speed responses of both
schemes are shown in Figure 3. It can be observed that there is no significant difference between both schemes.
Figure 4(a) shows that the compensation technique results in a sinusoidal stator flux waveform whilst Figure 4(b)
shows that the flux waveform obtained from the conventional DTC implementation is clearly not sinusoidal. Also,
it may be seen that, with the modified DTC strategy, the actual stator flux magnitude is equal to the reference value
of 0.8Wb whilst with the classical method there is a significant error between the two flux values. Figure 5(a)
shows a that the stator flux vector locus is circular compared to the conventional, hexagonal flux locus of Figure
5(b).
speed (rad/s) and torque (Nm)
torque (Nm)
2
1
0
20
10
0
-10
0.05 0.1 0.15
time (sec.)
0.2 0.25
speed
torque
0.05 0.1 0.15 0.2
time (sec.)
3
speed (rad/s), torque (Nm)
torque (Nm)
vqsl
2
1
sQ
α
vdsl
Vsl
V − R I
sl
α
Is
s
RsI
s
s
sD
0
0 0.05 0.1 0.15 0.2 0.25
time (sec.)
20
10
0
-10
speed
torque
0.05 0.1 0.15 0.2 0.25
time (sec.)
(a) Proposed Method (b) Conventional Method
Figure 3 Torque and Speed Responses of Induction Motor (ωr =20 rad/s and TL=1.2 Nm).

stator d-axis flux (Wb)
stator d-axis current (A)
(a) Proposed Method (b) Conventional Method
Figure 4 Phase Current and d-axis Stator Flux (ωr =20 rad/s and TL=1.2 Nm).
0.5
-0.5
(a) Proposed Method (b) Conventional Method
Figure 5 Stator Flux Trajectories of Induction Motor (ωr =20 rad/s and TL=1.2 Nm).
Experimental Results
1
0
-1
0.5
0
-0.5
0.1 0.2 0.3
time (sec.)
0.4
0.1 0.2 0.3
time (sec.)
0.4 0.5
q-axis stator flux (Wb)
1
0
-1
-1 -0.5 0 0.5 1
d-axis stator flux (Wb)
Experimental tests were carried out on a 0.37kW squirrel-cage induction motor using a TMS320C31 floating point
Digital Signal Processor (DSP) as the controller. The basic TMS320C31 DSP has been modified by the addition
of 64k-words of memory, an 8-channel Analogue to Digital Converter (ADC) and 4-channel Digital to Analogue
Converter (DAC). The communication between the external board and the DSP board is performed by “built-in”
control signals and external user ports.
The modified DTC method is compared with conventional DTC under the same operating conditions. The torque
and speed responses of both schemes are shown in Figure 6. It can be observed that there is no significant
difference between the responses under each of the schemes. Figure 8 illustrates the stator current waveforms
obtained using both the proposed DTC method and conventional DTC. Figure 7(a) shows that the current
waveform is more sinusoidal under the proposed DTC scheme. The experimental results for stator flux locus are
shown in Figure 8. It can be seen from Figure 8(a) that the stator flux trajectory is more circular than that under
the conventional DTC scheme.
1V=2.35rad/s
1V=0.25Nm
q-axis stator flux (Wb)
4
stator d-axis flux (Wb)
1
0.5
0
-0.5
stator d-axis current (A)
2
1
0
-1
-2
0.05 0.1 0.15 0.2 0.25 0.3 0.35
time (sec.)
0.5
0
-0.5
0.05 0.1 0.15 0.2 0.25 0.3 0.35
time (sec.)
-1
-1 -0.5 0 0.5 1
d-axis stator flux (Wb)
1V=2.35rad/s
1V=0.25Nm
(a) Proposed Method (b) Conventional Method
Figure 6 Torque and Speed Responses of Induction Motor (ωref =20 rad/s and TL=1.2 Nm).

(a) Proposed Method (b) Conventional Method
Figure 7 Phase Current of Induction Motor (ωref =20 rad/s and TL=1.2 Nm).
(a) Proposed Method (b) Conventional Method
Figure 8 Stator Flux Trajectories of Induction Motor (ψref = 0.8Wb)
Conclusion
In this paper, a new compensation strategy for Direct Torque Control of induction motor has been presented. The
results have been compared with those obtained using conventional DTC under the same operating conditions.
Simulation and experimental results show that the new strategy can be applied to DTC to improve the flux control
at low speeds.
Acknowledgements
The authors would like to acknowledge the University of Bristol for support and the use of facilities. Thanks are
also due to Dr. Naim Dahnoun for his useful discussion regarding DSP.
References
1V=0.4A
[1] Ludtke, M.G. Jayne, “A New Direct Torque Control Strategy”, IEE, Savoy Place, London 1995, pp. 5/1-5/4.
[2] P. Vas, “Vector Control of ac Machines”, Oxford University Press, 1990.
[3] Manfred Depenbrock, “Direct Self Control (DSC) of Inverter-Fed Induction Machine IEEE Trans. on Power
Electronics, Vol. 3, No. 4, October 1992, pp. 420-429.
[4] I. Takahashi, T. Noguchi, “A New Quick-Response and High-Efficiency Control Strategy of an Induction
Motor”, IEEE Trans. on Industry Applications, Vol.IA-22, No. 5, September/October 1986, pp. 820-827.
[5] P. Vas, “Sensorless vector and direct torque control”, Oxford University Press, 1998.
[6] A. Monti, F. Pironi, F. Sartogo, P. Vas, “A New State Observer For Sensorless DTC Control”, Power
Electronics and Variable Speed Drives, 21-23 September 1998, No. 456, pp. 311-17.
[7] J.R.G. Schofield, “Variable speed drives using induction motors and direct torque control”, ABB Industrial
Systems Ltd., IEE, Savoy Place, London 1998, pp. 5/1-5/7.
5
1V=0.4A
1V=0.2Wb 1V=0.2Wb