Documente Academic
Documente Profesional
Documente Cultură
4, JULY/AUGUST 2010
1509
I. I NTRODUCTION
1510
dq
d i dq
1
u dq
u
i dq
C =
Cf C j C
dt
Lf c
d u dq
1
i dq
i dq
u dq
Cf =
L C j Cf .
dt
Cf
(1)
1
1
UCf (s)
=
2
2
UC (s)
Lc Cf s + Res
(3)
and taking into account ICf (s) = sCf UCf (s), this gives
GIcf
LCL (s) =
1
s
ICf (s) UCf (s)
=
2 .
UCf (s) UC (s)
Lc s2 + Res
(4)
Lf g + Lf c
3 Tc
TI = 9 Tc .
(5)
DANNEHL et al.: INVESTIGATION OF ACTIVE DAMPING APPROACHES FOR PI-BASED CURRENT CONTROL
1511
TABLE I
C LOSED -L OOP T RANSFER F UNCTIONS W ITH UCf AND ICf F EEDBACK FOR D IFFERENT F EEDBACK T YPES
GLCL (s)
K(s) + GLCL (s)
(6)
whereas for UCf feedback (3) is used and (4) for ICf feedback.
In this paper, feedback on the converter voltage references
is considered. Note that feedback on the current references,
instead of voltage references, is also possible. This can be
beneficial with respect to current overshoot during transients
[6]. In [7], [14], this is used as part of the concept of a so-called
virtual resistor. In [6], it is used as part of deadbeat state space
control. Similar analyses including a PI controller in the loop
give the required types of feedback. As this is beyond the scope
of this paper, results are not shown here.
A. Filter Capacitor Voltage Feedback
Evaluations of (3) and (6) for various types of feedback are
shown in Table I. For proportional feedback, the transfer function of a resonator is obtained, whereas the natural resonance
frequency Res is shifted in proportion with K. As can be seen
1512
Fig. 4. Impact of different continuous feedback characteristics on the system pole-zero locations in the continuous s-plane (PWM neglected): (top) Filter
capacitor voltage feedback and (bottom) filter capacitor current feedback.
DANNEHL et al.: INVESTIGATION OF ACTIVE DAMPING APPROACHES FOR PI-BASED CURRENT CONTROL
1513
TABLE II
S YSTEM PARAMETERS
TABLE III
LCL F ILTER S ETTINGS AND C ORRESPONDING
R ESONANCE F REQUENCIES
1514
Fig. 7. ICf feedback: Impact of feedback gain KiCf on the pole zero map for LCL filters with different resonance frequencies: (left) fRes = 1.69 kHz;
(middle) fRes = 1.19 kHz; (right) fRes = 0.69 kHz.
Fig. 8. ICf feedback: Impact of feedback gain KiCf on the step response for
LCL filters with different resonance frequencies: (top) fRes = 1.69 kHz and
(bottom) fRes = 1.19 kHz.
DANNEHL et al.: INVESTIGATION OF ACTIVE DAMPING APPROACHES FOR PI-BASED CURRENT CONTROL
1515
Fig. 9. UCf feedback: Impact of feedback gain KuCf on the pole zero map for LCL filters with different resonance frequencies: (left) fRes = 1.69 kHz;
(middle) fRes = 1.19 kHz; (right) fRes = 0.69 kHz.
Fig. 12. ICf feedback: measured step response of (left) dq converter current
(right) and line current at medium resonance frequency.
Fig. 10. UCf feedback: Impact of feedback gain KuCf on the step response
for LCL filters with different resonance frequencies: (top) fRes = 1.19 kHz
and (bottom) fRes = 0.69 kHz.
Fig. 13. UCf feedback: measured step response of (left) dq converter current
and (right) line current at medium resonance frequency.
1516
Fig. 14. Measured steady state spectra of (top) phase line currents and (bottom) converter currents for medium resonance frequency (left) without any additional
feedback, (middle) ICf feedback, and (right) UCf feedback (sampling frequency: 100 kHz).
TABLE IV
C OMPARISON OF ICf AND UCf F EEDBACK
DANNEHL et al.: INVESTIGATION OF ACTIVE DAMPING APPROACHES FOR PI-BASED CURRENT CONTROL
1517