Beruflich Dokumente
Kultur Dokumente
Summary This paper presents a comparison between two control strategies for Permanent Magnet Synchronous Motors (PMSM):
Field Oriented Control (FOC) and Direct Torque Control (DTC). These two strategies can be considered among the family of Vector
Control (VC) methods and provide a solution for high-performance drives. This paper presents the implementation of both strategies
in PMSM drives. Advantages and disadvantages of both methods are discussed and different simulation tests are performed to
illustrate the features of both methods. The criteria followed to establish a fair comparison between both control methods is also
presented.
PM, both inductances have similar values and for control scheme of the FOC strategy is shown in Fig. 2:
simplification they can be considered equal. The model L sq i s q
is completed with the mechanical equation, which is
r
i sd* -
+- PI + d,q
defined as: PMSM
*
r
+ PI + PI + a,b,c
- - +
d m r r L sd i s d s
J = e l Bm (6) i sd
d,q
i sa
dt i sq i sb
r = Pm (7) a,b,c
d r
dt
s
that d and q axis are not completely independent and
d
sd
q
stator flux position by means of the stator voltage their respective hysteresis bands.
applied to the motor. The desired decoupled control of
*
the stator flux modulus and torque is achieved by acting s
+-
s = 2 + L2sq isq
2
(13)
Torque [Nm]
10 Torque reference from Speed PI
5
In order to carry out a comparative analysis of FOC Torque load
Speed [rad/s]
Regarding transients the main characteristic to be 0
-30
analysed is the time response to a torque step. This test
-60
can be performed at different speeds. For the steady- 0 0.05 0.1 0.15 0.2 0.25
state performance, two different characteristics can be 5000
frequency [Hz]
analysed regarding flux and torque responses: the
Switching
2500
average error as defined in (15) and the oscillation or
ripple in the torque and stator flux that can be calculated 0
0 0.05 0.1 0.15 0.2 0.25
by means of the standard deviation as defined in (16) Time [s]
(being n the number of samples): Fig. 6. Variation of switching frequency during a transient in
DTC
n
1
e= ei (15) In this case the values employed for the hysteresis
n i =1 bands are 2.5% for flux and 7.5% for torque.
1/ 2 In [8] it is also said that the sampling time of the
n
1
( ei e )
2
= (16) control loop cannot be equal in both systems. First of all
n 1 i =1 equal sampling times would imply different switching
frequencies. Secondly the main advantages of DTC are
Another interesting feature in steady-state operation the simplicity and lower calculation requirements when
is the distortion of the stator phase currents. This can be compared to FOC, and therefore these advantages must
evaluated by means of the current spectrum and the be exploited to have a fair comparison.
Total Harmonic Distortion (THD). The presence of
harmonics in the audible noise band can be noticed from 6. SIMULATION RESULTS
this analysis.
In [8], where both control methods are compared for Some simulation tests have been designed and
the IM, it is discussed how FOC and DTC can be fairly executed to obtain comparative results of both systems.
compared. The authors claim that a fair comparison can The parameters shown in Table 2 corresponding to a
only be made if the average switching frequency of the real motor have been used in the simulation model. The
inverter is approximately the same. In FOC the sampling time used in the control loop is 25 s for DTC
switching frequency is adjusted by the PWM period. and 100 s for FOC. In the FOC system the current PI
DTC, however, has variable frequency due to the regulators have been tuned using the Absolute Value
hysteresis blocks, which depends on the operating point. Optimum (AVO) criterion (Kp =8.86, Ki=778.6), and the
The variation of the switching frequency depending on Symmetric Optimum (SO) criterion has been used for
the speed and load torque is illustrated in Fig. 5. the speed PI (Kp =0.0934, Ki=3.18) [10].
Tab. 2. PMSM characteristics (Siemens 1KF7)
Nominal Output Power (Pn) 2135W
4000
Nominal Speed ( n) 3000rpm
Switch in g freq uen cy [Hz]
3800
Nominal Torque (Mn) 6.8Nm
3600 Nominal Current (In) 4.4A
3400 Number of pole pairs (P) 4
3200
Stator resistance (Rs) 1.09Ohm
Stator inductance (Lsd and Lsq) 0.0124H
4.15e-4 Kgm2
3000
0 Inertia (J)
2 300
4 200
250 Permanent Magnet Flux ( ) 0.1821Wb
150
Load Torque [Nm] 6 100 Speed [rad/s]
50
The first test performed is shown in Fig. 7. It
Fig. 5. Variation of switching frequency in DTC of PMSM for consists on a torque step change from 0 to nominal
steady-state depending on the operating point torque at 3 different speeds. It can be seen how the
response time is considerably smaller for DTC when
Fig. 6 presents the variation of the switching compared to FOC. Table 3 contains the settling time of
frequency during a transient, when a step change of the both systems. It can also be observed a higher torque
load torque at 0 rad/s speed occurs. During this transient ripple for DTC.
torque error is bigger and active VSI vectors are applied Tab. 3. Torque settling time at different speeds
for longer time, reducing the switching frequency. Electrical speed DTC FOC
In order to have similar switching frequency in the 0 rad/s 0.22ms 6ms
DTC and FOC systems the hysteresis bands of the DTC 300 rad/s 0.32ms 5ms
scheme must be adjusted. 1200 rad/s 1ms 15ms
Advances in Electrical and Electronic Engineering 80
10 10
5 5
0.2 0.2
0 0 0.1 0.1
0 0
0 0.05 0.1 0.15 0.2 0 0.05 0.1 0.15 0.2
0.08 0.085 0.09 0.095 0.08 0.085 0.09 0.095
20 20
isq [A]
10 10
0 0
5 5 -20 -20
0 0.05 0.1 0.15 0.2 0 0.05 0.1 0.15 0.2
5 0.5
0 0
isd [A]
0 0
0.08 0.085 0.09 0.095 0.08 0.085 0.09 0.095
-5 -0.5
Time [s] Time [s] 0 0.05 0.1 0.15 0.2 0 0.05 0.1 0.15 0.2
Time[s] Time[s]
Fig. 7. Torque transients for DTC (left) and FOC (right) at
0 rad/s (up), 300 rad/s (middle) and 1200 rad/s (down) Fig. 9. Torque and stator flux behaviour during a speed step
change for DTC (left) and FOC (right)
The second test performed for transient conditions
consists on the response to a speed step change. Fig. 8 It can be seen in Fig. 9 how the higher torque ripple
shows the stator flux path in the - plane during the is not affecting the speed response due to the inertia of
test. Fig. 9 shows the speed, torque, stator flux modulus, the motor.
isd and isq responses. It can also be seen how the torque Some additional tests have been carried out to assess
response for DTC performs a better tracking of its the behaviour of both systems in steady-state conditions.
reference and as a result the rise time of the speed The torque and stator flux average errors and standard
response is slightly smaller. It can also be noticed the deviation have been calculated for both systems at 9
higher ripple of the DTC system regarding the flux and different operation points obtained combining three
torque responses. different speeds (0, 150 and 300 rad/s) with three
different levels of load torque (0, 3.4 and 6.8 Nm). The
0.3 0.3 resulting values have been averaged and are presented
0.2 0.2 in Table 4. They are presented as a percentage of the
0.1 0.1 nominal values of the stator flux modulus and torque.
Flux Beta
Flux Beta
0 0
Tab. 4. Steady-state performance indexes
-0.1 -0.1
DTC FOC
-0.2 -0.2 Stator Flux
0.3% 0.1%
-0.3 -0.3 Average Error
-0.2 0 0.2 -0.2 0 0.2 Stator Flux
Flux Alpha Flux Alpha 4.85% 0.15%
Standard Deviation
Fig. 8. Stator flux circular path in - coordinates during the Torque
1.8% 0.08%
speed step response for DTC (left) and FOC (right) Average Error
Torque
12.8% 2.81%
Standard Deviation
FFT window: 2 of 9.512 cycles of selected signal FFT window: 2 of 9.54 8 cycles of selecte d signal
5 5
Current [A]
0 0
-5 -5
0.28 0.29 0.3 0.31 0.32 0.33 0.34 0.35 0.28 0.29 0.3 0.31 0.32 0 .33 0.34 0.35
Time (s) Time (s)
10 10
Mag (% of Fundamental)
8 Fundamental (23.78Hz) = 3.177 , THD= 22.93% 8 Fundamental (23.87Hz) = 3.146 , THD= 2.84%
6 6
4 4
2 2
0 0
0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5 000
Frequency (Hz) Frequency (Hz)
Fig. 10. Steady-state stator phase currents and harmonic spectrum at 150 rad/s and 3.4 Nm for DTC (left) and FOC (right)
81 Comparison between FOC and DTCstrategies