Vous êtes sur la page 1sur 5

This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2408053, IEEE Transactions on Power Electronics

IEEE TRANSACTIONS ON POWER ELECTRONICS

Proportional-Resonant Current Controllers


Design based on Desired Transient Performance
A. Kuperman, Senior Member, IEEE

 (and vice-versa) frame transformation in addition to associated


Abstract—Stationary reference frame proportional- need for a phase-locked loop [6]. A viable alternative to
resonant controllers have recently became a viable synchronous frame PI controllers are stationary frame
alternative to rotation reference frame proportional- proportional-resonant (PR) compensators [7] – [11], capable
integrative compensators in AC applications because of of achieving zero steady-state error at AC frequency [12] –
their ability of achieving zero steady-state error without [14]. In fact, utilization of PR controllers in AC systems is a
the need for computational-intensive reference frame straightforward outcome of internal model principle [15] –
transformations. While extensive effort have been put into [17], stating that in case to achieve zero steady-state error in a
performance comparison of the two control schemes, stable feedback system, the loop gain must include model
design of proportional-resonant controllers according to which can generate the required reference and disturbance
desired closed-loop time-domain transient performance signals, calling for integrative controllers in DC systems and
was barely investigated. This letter proposes a method for resonant controllers in AC systems [18] - [21]. Nevertheless,
deriving proportional-resonant controller structure and derivation of PR converters is mostly performed either from
coefficients according to desired transient behavior of AC equivalent synchronous frame PI controllers [6], [10], [22] or
signal amplitude, applied to typical power converter using conventional Bode diagram tools. The latter is relatively
current loop. The method is based on the fact that if AC simple since the resonant term of conventional PR controller
signal envelope is perceived as DC signal, its transient has little or no contribution outside the resonant frequency [9],
behavior may be easily shaped utilizing well-known [10], leaving the proportional gain to shape the frequency
approaches employed in DC systems loop shaping while response. However, while PI controller design according to
keeping zero phase tracking error at all times. On the desired time-domain specifications in DC systems is popular
other hand, while desired transient performance is easily and linked to frequency domain figures of merit [23], the only
achieved, the relation between crossover frequency and attempt of similar PR controller design was recently made in
transient processes time constant is not as straightforward [24], where a methodology to assess and optimize the transient
as in DC systems. The validity of presented theoretical response of digital proportional-resonant current controllers
analysis is evaluated by simulations. was proposed. Moreover, while in DC systems transient
performance is tightly related to loop gain behavior in
Index Terms—Proportional-resonant controller, current proximity of crossover frequency, the same was not proven for
control, transient performance PR-controller based AC systems. Another issue worth
emphasizing is the fact that active and reactive power
I. INTRODUCTION transferred by AC current depend on its amplitude and phase,
ECENT advances in distributed power generation have consequently their transient response is of specific importance.

R emphasized the control issues related to utility interfacing


[1]. It was shown that grid-side current controllers play a
This letter makes an attempt to design a PR current controller
based on desired time-domain tracking performance.
crucial role in coping power quality issues and faults ride- Compared to conventional PR controller, resulting structure
through [2]. Consequently, thorough investigation of different contains additional resonant term. Moreover, it is shown that
current control methods was carried out [3], [4]. In order to unlike in DC systems, resulting crossover frequency is
achieve zero steady-state current tracking error, proportional- nonlinearly linked to desired time-domain specifications. The
integral (PI) control in synchronous frame is the validity of the presented theoretical analysis is confirmed by
conventionally utilized technique [3], [5]. However, it is not simulation results.
well-suited for single phase systems and requires significant
computational effort to perform stationary-to-synchronous II. GENERALIZED CURRENT LOOP DESCRIPTION
Typical large signal averaged equivalent circuit of a
Copyright (c) 2015 IEEE. Personal use of this material is permitted. generalized power converter current loop is shown in Fig. 1,
However, permission to use this material for any other purposes must be where vC denotes controlled converter output voltage, vL
obtained from the IEEE by sending a request to pubs-permissions@ieee.org. symbolizes uncontrolled voltage (e.g. grid voltage or back
Alon Kuperman is with Hybrid Energy Sources R&D Laboratory, the electromotive force) and (L, R) are separating impedance
Dept. of Electrical Engineering and Electronics, Ariel University, Ariel
40700, Israel (email: alonku@ariel.ac.il). inductance and resistance, respectively [25].

0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2408053, IEEE Transactions on Power Electronics

IEEE TRANSACTIONS ON POWER ELECTRONICS

i.e. zero steady state error and transient time constant of 𝜔𝐶−1
are requested. Transforming (9) into Laplace domain results in
1 1  A C
I (s)  A      I * ( s )  TDC ( s ) (10)
 s s  C  s s  C
with
C ( s)  Ls  R 
1
I ( s) C
TDC ( s)   . (11)
I ( s) s  C 1  C ( s)  Ls  R 1
Fig. 1. Averaged equivalent circuit of a generalized power converter current *
loop.
Desired loop gain is then
The controlled converter output voltage is given by
C
C ( s )  Ls  R  
1
vC (t )  d (t )  vDC (t ), (1) , (12)
s
where vDC(t) is equivalent DC link voltage of the converter satisfied by the following PI controller,
C  Ls  R 
and d(t) is current controller output, satisfying
0  d 1 KI
(2) CDC ( s)   K P , DC  (13)
in case of DC-DC converter and s s
1  d  1 (3) with
in case of DC-AC or AC-DC converter (i.e. the power flow of KP, DC  LC , KI  RC . (14)
the circuit in Fig.1 may be bidirectional). Current loop
dynamics is then governed by B. AC system
di (t ) Next, consider an AC system with the following reference
L  d (t )vDC (t )  Ri (t )  vL (t ) (4)
dt current,
and is nonlinear in general since vDC(t) is a regulated quantity i* (t )  A sin 0t  u(t ) (15)
as well. Control closed-loop block diagram of the current loop or
is shown in Fig. 2. The error between the reference current i*
A0
and actual current i is processed by the compensator C(s). I * (s)  (16)
Here, the uncontrolled voltage and the equivalent DC-linked s  02
2

voltages are assumed measurable and used as feedforward in Laplace domain. Utilizing (13) would lead to
signals to the controller in order to decouple the system
A0      s  C 1 
dynamics from external disturbances [26], [27], i.e. the control I ( s)   C  A 20 C 2  2  .
signal is given by s  0 s  C
2 2
C  0  s  0 s  C 
2

1
d (t )   vN (t )  vL (t )  , (5)
Transforming (17) into time domain results in
(17)
vDC (t )
where vN(t) is the compensator output signal. The closed-loop C   1 0  C t

i (t )  A  sin   t  tg    e  u (t ),
C2  02   C 
system dynamics is then given by 0 0

di (t ) 
L  Ri (t )  vN (t ). (6) (18)
dt i.e. steady-state error is expected both in amplitude and phase
responses. Both can be minimized by increasing ωC,
necessitating high control bandwidth which is not always
available. Alternatively, it is proposed to define the desired
time-domain behavior by
 
i(t )  A 1  eC t sin 0t  u(t ), (19)
Fig. 2. Closed-loop system block diagram. requesting zero steady state amplitude error with transient
time constant of 𝜔𝐶−1 and zero instantaneous steady state phase
III. COMPENSATOR DESIGN
error. Transforming (19) into Laplace domain results in
A. DC system
A0 A0
Assume a DC system with reference current given by I ( s)   
i (t )  A  u (t )
*
(7) s  0 ( s  C )2  02
2 2
(20)
or A0 2C s  C2
A  2   I * ( s)  TAC ( s )
I * (s)  (8) s  0 ( s  C )  0
2 2 2

s with
in Laplace domain. Desired time-domain output current
behavior is defined by
 
i(t )  A 1  eC t u(t ), (9)

0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2408053, IEEE Transactions on Power Electronics

IEEE TRANSACTIONS ON POWER ELECTRONICS

C ( s)  Ls  R 
1
I ( s) 2C s  C2
TAC ( s)   . 40
I * ( s) 1  C ( s)  Ls  R 1 ( s  C ) 2  02 20
(21)

i* [A]
0
It is interesting to note that
1
TDC ( s  j0 )  TDC ( s  j0 )  .
-20
TAC ( s )  (22) DC AC
2 -40
0 0.1 0.2 0.3 0.4 0.5 0.6
Desired loop gain is then obtained as
2C s  C2
C ( s )  Ls  R  
1 40
, (23)
s 2  02 20
satisfied by the following PR controller,
 Ls  R   2C s  C2 

i [A]
0

C AC ( s )   -20
s 2  02 (24) AC DC
-40
K s K 0 0.1 0.2 0.3 0.4 0.5 0.6
 K P , AC  2 R1 2  2 R 2 2 time [s]
s  0 s  0 Fig. 3. Simulation results. Top: reference signals, Bottom: output signals for
with ωC = 10π [rad/s].
K P , AC  2 LC , K R1  LC2  2 RC ,
(25) In order to demonstrate the tracking quality, Fig. 4 presents
K R 2  RC2  2 LC02 . AC system simulation results. It may be concluded that phase
Note that unlike PR controllers proposed in the literature by tracking error remains zero at all times while output signal
far, the numerator of (24) includes additional term KR2, amplitude converges to reference signal amplitude at desired
allowing to obtain the desired transient behavior. In addition, rate.
unlike stated in [9], [27], it can be easily shown that typical
30
PR controller utilizing coefficients of PI controller (14),
20
2K I s 2 R s
C ( s )  K P , DC   LC  2 C 2
current [A]

(26) 10
s  0
2 2
s  0 0

would yield different transient performance (still achieving -10


zero steady-state error though) since the resulting loop gain -20
i
*
i
will be plant dependent in contrary to (21). -30
0 0.1 0.2 0.3 0.4 0.5 0.6
To conclude, the controller (24) was designed to satisfy
the three following goals:
1. Zero steady-state amplitude error. 20
2. Zero instantaneous phase error. 10
error [A]

3. Desired transient amplitude behavior.


0
Therefore, proposed PR controller parameters are chosen to
-10
satisfy goal 3 while its structure assures that goals 1 and 2 are
achieved. -20
0 0.1 0.2 0.3 0.4 0.5 0.6
time [s]
IV. VERIFICATION
Fig. 4. Simulation results for ωC = 10π [rad/s]. Top: reference and output
In order to demonstrate the performance of the proposed signals, Bottom: tracking error.
controller, two following PSIM simulations (L = 1mH, R =
0.1 Ω, vDC = 400 V, vL(t) = 230√2∙sin(ω0t - π/4) V, i*(t) = V. REMARK
16√2∙(u(t) – 0.5∙u(t-0.2) + 0.5∙u(t-0.4))∙cos(ω0t)) A were
It should be emphasized that the loop gain crossover
carried out with ωC = 10π [rad/s]:
frequency ωC,AC of AC system utilizing the proposed
- DC system with ω0 = 0 and compensator (13);
controller (23) differs from the loop gain crossover frequency
- Three-level grid-connected inverter (AC system) with ω0
ωC,DC = ωC of DC system utilizing PI controller (13) as
= 100π and compensator (24) (switching frequency 20
follows. According to (22), the following holds at the
kHz, sampling frequency 20 kHz, 12-bit A/D, C-code
crossover frequency ω = ωC,AC,
based controller).
Fig. 3 presents simulation results of both systems. It may be j 2C , DCC , AC  C2 , DC
 1. (26)
concluded that time-domain response of DC system and the C2 , AC  02
envelope of AC system are identical as required.
Solving, there is

0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2408053, IEEE Transactions on Power Electronics

IEEE TRANSACTIONS ON POWER ELECTRONICS

2 2
VI. CONCLUSION
C , AC     0  A method for deriving proportional-resonant controller
  0   2  4    5 (27)
C , DC 
 C , DC
structure and coefficients according to desired transient
 C , DC   behavior of AC signal amplitude, applied to typical power
for ωC,DC > ω0 which is a realistic assumption when high converter current loop was presented. Since AC signal
performance is required. Equation (27) is shown graphically in envelope was treated as DC signal, its transient behavior was
Fig. 5, indicating that the ratio of AC system crossover shaped utilizing well-known approach employed in DC
frequency and DC system crossover frequency varies from systems loop shaping. Moreover, zero phase tracking error
2.45 for ωC,DC = ω0 down to 2.05 for ωC,DC >> ω0. was assured at all times. While desired transient performance
was perfectly achieved, the crossover frequency and transient
2.45
processes time constant were shown to be nonlinearly related,
2.4
which is the main drawback of the proposed method.
Nevertheless, for practical cases the resulting bandwidth was
2.35 shown increase two-fold compared to equivalent DC system
bandwidth.
2.3
C,AC/C,DC

REFERENCES
2.25
[1] F. Blaabjerg, Z. Chen and S. Kjaer, “Power electronics as efficient
interface in dispersed power generation systems,” IEEE Trans. Power
2.2
Electron., vol. 19, no. 5, pp. 1184 – 1194, Sep. 2004.
[2] A. Timbus, M. Liserre, R. Teodorescu, P. Rodriguez and F. Blaabjerg,
2.15 “Evaluation of current controllers for distributed power generation
systems,” IEEE Trans. Power Electron., vol. 24, no. 3, pp. 654 – 664,
2.1 Mar. 2009.
[3] M. Kazmierkowski and L. Malesani, “Current control techniques for
2.05 three-phase voltage-source PWM converters: A survey,” IEEE Trans.
0 0.2 0.4 0.6 0.8 1 Ind. Electron., vol. 45, no. 5, pp. 691 – 703, Oct. 1998.
 / [4] F. Blaabjerg, R. Teodorescu, M. Liserre and A. Timbus, “Overview of
0 C,DC
Fig. 5. Relation between crossover frequencies of AC and DC systems control and grid synchronization for distributed power generation
systems,” IEEE Trans. Ind. Electron., vol. 53, no. 5, pp. 1398 – 1409,
Oct. 2006.
The finding is verified by Fig. 6, where Bode diagrams of (12) [5] X. Bao, F. Zhuo, Y. Tian and P. Tan, “Simplified feedback linearization
and (22) are plotted for ω0 =100π rad/s and different values of control of three phase photovoltaic inverter with an LCL filter,” IEEE
ωC. Consequently, AC system crossover frequency is not Trans. Power Electron., vol. 28, no. 6, pp. 2739 – 2752, Jun. 2013.
[6] J. Hwang, P. Lehn and M. Winkelnkemper, “A generalized class of
linked directly to transient dynamics of output signal stationary frame current controllers for grid-connected AC-DC
envelope, contrary to DC system crossover frequency. converters,” IEEE Trans. Power Del., vol. 25, no. 4, pp. 2742 – 2751,
Nevertheless, for practical cases with ωC,DC >> ω0 it may be Oct. 2010.
assumed that ωC,AC ≈ 2ωC,DC, indicating that two-fold control [7] Y. Sato, T. Ishizuka, K. Nezu and T. Kataoka, “A new control strategy
for voltage-type PWM rectifiers to realize zero steady-state control error
bandwidth increase would be required. to input current,” IEEE Trans. Ind. Appl., vol. 34, no. 3, pp. 480 – 486,
May 1998.
[8] D. Zmood, D. Holmes and G. Bode, “Frequency domain analysis of
100 three-phase linear current regulators,” IEEE Trans. Ind. Appl., vol. 37,
AC,  = 100
C no. 2, pp. 601 – 610, Mar. 2001.
DC,  = 100 [9] D. Zmood and G. Holmes, “Stationary frame current regulation of PWM
80 C
AC,  = 500
C
inverters with zero steady-state error,” IEEE Trans. Power Electron.,
DC, C = 5100
vol. 18, no. 3, pp. 814 – 822, May 2003.
60
[10] R. Teodorescu, F. Blaabjerg, M. Liserre and P. Loh, “Proportional-
AC,  = 2000
Magnitude (dB)

C resonant controllers and filters for grid connected voltage-source


40 DC,  = 2000
C
converters,” IEE Proc. Electr. Power Appl., vol. 153, no. 5, pp. 750 –
762, Sep. 2006.
[11] D. Holmes, T. Lipo, B. McGrath and W. Kong, “Optimized design of
20
stationary frame three phase AC current regulators,” IEEE Trans. Power
Electron., vol. 24, no. 22, pp. 2417 – 2426, Nov. 2009.
0 [12] L. Chiang, M. Newman, D. Zmood and D. Holmes, “A comparative
analysis of multiloop voltage regulation strategies for single and three-
-20
phase UPS systems,” IEEE Trans. Power Electron., vol. 18, no. 5, pp.
1176 – 1185, Sep. 2005.
[13] Z. Wei, D. Lu and V. Agelidis, “Current control of grid-connected boost
-40
2 3 4
inverter with zero steady-state error,” IEEE Trans. Power Electron., vol.
10 10 10 26, no. 10, pp. 2825 – 2834, Oct. 2011.
Frequency (rad/s) [14] D. Dong, T. Thacker, R. Burgos W. Fei and D. Boroyevich, “On zero
Fig. 6. Loop gain bode plots of AC and DC systems for ωC = 100π and steady-state error voltage control of single-phase PWM inverters with
different values of ωC. different load types,” IEEE Trans. Power Electron., vol. 26, no. 11, pp.
3285 – 3297, Nov. 2011.
[15] B. Francis and W. Wonham, “The internal model principle for linear
multivariable regulators,” Appl. Math. Optim., vol. 2, no. 2, pp. 170 –
194, 1975.

0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2408053, IEEE Transactions on Power Electronics

IEEE TRANSACTIONS ON POWER ELECTRONICS


[16] B. Francis and W. Wonham, “The internal model principle of control [23] S. Hamamci and N. Tan, “Design of PI controllers for achieving time
theory,” Automatica, vol. 12, pp. 457 – 465, 1976. and frequency domain specifications simultaneously,” ISA Trans., vol.
[17] C. Garcia and M. Morari, “Internal model control: A unifying review 45, no. 4, pp. 529 – 543, 2006.
and some new results,” Ind. Eng. Chem. Process Des. Dev., vol. 21, pp. [24] A. Vidal, F. D. Freijedo, A. G. Yepes, P. Fernandez-Comesana, J.
308 – 323, 1982. Malvar, O. Lopez and J. Doval-Gandoy, “Assessment and optimization
[18] L. Harnefors and H.-P. Nee, “Model-based current control of AC of the transient response of proportional-resonant current controllers for
machines using the internal model control method,” IEEE Trans. Ind. distributed power generation systems,” IEEE Trans. Ind. Electron., vol.
Appl., vol. 34, no. 1, pp. 133 – 141, Jan. 1998. 60, no. 4, pp. 1367 – 1383, Apr. 2013.
[19] S. Fukuda and T. Yoda, “A novel current-tracking method for active [25] M. Liserre, R. Teodorescu and F. Blaajberg, “Stability of photovoltaic
filters based on a sinusoidal internal model,” IEEE Trans. Ind. Appl., and wind turbine grid-connected inverters for a large set of grid
vol. 37, no. 3, pp. 888 – 895, May 2001. impedance values,” IEEE Trans. Power Electron., vol. 21, no. 1, pp. 263
[20] S. Fukuda and R. Imamura, “Application of a sinusoidal internal model – 272, Jan. 2006.
to current control of three-phase utility-interface converters,” IEEE [26] A. Yazdani and R. Iravani, Voltage-Sourced Converters in Power
Trans. Ind. Electron., vol. 52, no. 2, pp. 420 – 426, Apr. 2005. Systems: Modeling, Control, and Applications. Hoboken, NJ: Wiley,
[21] M. Yazdanian and A. Mehrizi-Sani, “Internal model-based current 2009, ch. 7.
control of the RL filter-based voltage-sourced converter,” ,” IEEE [27] M. Castilla, J. Miret, A. Camacho, J. Matas and L. Garcia de Vicuna,
Trans. Energy Convers., vol. 29, no. 4, pp. 873 – 881, Dec. 2014. “Reduction of current harmonic distortion in three-phase grid-connected
[22] C. Zou, B. Liu, S. Duan and R. Li, “Stationary frame equivalent model photovoltaic inverters via resonant current control,” IEEE Trans. Ind.
of proportional-integral controller in dq synchronous frame,” IEEE Electron., vol. 60, no. 4, pp. 1464 – 1472, Apr. 2013.
Trans. Power Electron., vol. 29, no. 9, pp. 4461 – 4465, Sep. 2014.

0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

Vous aimerez peut-être aussi