16

Current Proportional-Resonant
Control

The proportional-resonant (PR) controller is one of the most popular controllers used for grid-
connected inverters to regulate the current injected into the grid. In this chapter, the PR current
controller is designed and implemented for three-phase inverters, in the stationary reference
frame and in the natural reference frame.

16.1 Proportional-resonant Controller

For inverters, the controller deals with sinusoidal signals, which makes it difficult to design
the controller with the correct gain that is able to regulate the performance at the fundamental
frequency and also to reject harmonic disturbances. PI controllers, having a pole (with an
infinite gain) at the zero frequency, are not able to eliminate the steady-state error at the
fundamental frequency (Blaabjerg et al. 2006) unless it is adopted in the dq frame, as done in
Chapter 15. Alternatively, PR controllers can be used.
A PR controller is the combination of a proportional term and a resonant term given by

s
CPR (s) = K p + K i (16.1)
s 2 + ω2

where ω is the resonant frequency. Such a controller has a high gain around the resonant
frequency and, thus, is capable of eliminating the steady-state error when tracking or rejecting
a sinusoidal signal (Blaabjerg et al. 2006; Sera et al. 2005; Timbus et al. 2006b), according
to the internal model principle (Francis and Wonham 1975). As a result, PR controllers are
widely used in inverter control. In order to improve the performance of handling harmonics,
a harmonic compensator given by
 s
CHC (s) = K ih (16.2)
h=3,5,7,..
s 2 + (ωh)2

Control of Power Inverters in Renewable Energy and Smart Grid Integration, First Edition.
Qing-Chang Zhong and Tomas Hornik.
© 2013 John Wiley & Sons, Ltd. Published 2013 by John Wiley & Sons, Ltd.
260 Control of Power Inverters in Renewable Energy and Smart Grid Integration

where h is the harmonic order, can be easily added to a PR controller (Timbus et al. 2006b).
It is worth noting that, in order to maintain good performance of the controller, the resonant
frequency should be maintained close to the system frequency. If the system frequency varies
significantly, adaptive mechanisms, e.g. the one reported in (Timbus et al. 2006a), can be
adopted to adjust the resonant frequency according to the system frequency.

16.2 Control Structure

16.2.1 In the Stationary Reference (αβ) Frame
A PR controller is usually adopted in the stationary reference (αβ) frame for inverter control
(Sera et al. 2005; Teodorescu et al. 2006; Timbus et al. 2006a; Zmood et al. 1999, 2001), but
it can easily be implemented in the abc frame as well.
The block diagram of a current-controlled VSI in the stationary reference frame with the
PR controller is shown in Figure 16.1. The three-phase currents are transformed into α-,
β-components i α and i β using the abc → αβ transformation. The α-component is the same
as the Phase a component and the β-component is the combination of Phase b and Phase c
components (when the currents are balanced). Hence, only two current channels i α and i β
need to be controlled, with two separate PR controllers. A PLL is adopted to generate the
phase information of the grid voltage, which is needed by the dq → αβ transformation to
convert the d-, q-current references Id∗ and Iq∗ into the reference currents i α∗ and i β∗ . The PR
controllers, often equipped with the harmonic compensator in (16.2), forces i α and i β to track
the reference currents i α∗ and i β∗ . The output of the PR controllers is then converted back to
the abc frame before being converted into PWM signals to drive the switches. The real power
and reactive power exchanged with the grid are directly determined by Id∗ and Iq∗ because the
voltage is determined by the grid.
The grid voltages are feed-forwarded via a phase-lead low-pass filter to improve the per-
formance, as shown in Chapter 3, etc. The filter, chosen as (3.1), is introduced to compensate

Figure 16.1 Block diagram of a current-controlled VSI in the stationary reference frame (αβ)
Current Proportional-Resonant Control 261

the phase shift and gain attenuation caused by the computational delay, PWM modulation, the
inverter bridge and the LC filter. It also attenuates the harmonics in the feed-forwarded grid
voltages and improves the dynamics during grid voltage fluctuations (Timbus et al. 2009).

16.2.2 Equivalent Controller in the abc Frame

As mentioned before, the PR controllers can easily be implemented in the abc frame. Actually,
it is straightforward to do so. Instead of using the abc → αβ transformation, each phase can
be equipped with a PR controller, with the current reference generated from Id∗ and Iq∗ with
the dq → abc transformation. According to (Timbus et al. 2009), the controller matrix in the
abc frame is given as

⎡ ⎤
K p + K i s 2 +ω
s
2 0 0
CPRabc (s) = ⎣ 0 K p + K i s 2 +ω
s
2 0 ⎦. (16.3)
0 0 K p + K i s 2 +ω2
s

Because of this, the PR controller can be designed for each phase and then applied to the
stationary reference frame.

16.3 Controller Design

16.3.1 Model of the Plant
The model of the control plant can be derived from the single-phase diagram of the inverter
shown in Figure 16.2, which mainly consists of the LCL filter of the inverter. The inverter is
operated so that the average of u f over a switching period is the same as u and, hence, the
effect of the PWM conversion and the switching is negligible.
According to Kirchhoff’s current law (KCL), there is

ic = i1 − i2 . (16.4)

+ VDC -
Sc
PWM Inverter uf uo
bridge ug
u Rf Lf i1 ic Rg Lg i2
filter inductor Cf grid interface inductor
uc grid
Rd
neutral

Figure 16.2 Single-phase representation of an inverter

262 Control of Power Inverters in Renewable Energy and Smart Grid Integration

According to Kirchhoff’s voltage law (KVL), there are

1
u f = i 1 (Rf + s L f ) + i c + Rd , (16.5)
sCf

and

1
u g = −i 2 (Rg + s L g ) + i c + Rd , (16.6)
sCf

where s is the Laplace operator. Re-writing (16.4)–(16.6) in terms of impedances, then

    
uf Z 11 −Z 12 i1
= (16.7)
ug Z 21 −Z 22 i2

with

Z 11 = Rf + s L f + Rd + 1
sCf
, Z 12 = Rd + 1
sCf
,

Z 21 = Rd + 1
sCf
, Z 22 = Rg + s L g + Rd + 1
sCf
.

The transfer function from the inverter voltage u f to the grid output current i 2 is

HP (s) = i2
uf
= Z 21
Z 11 Z 22 −Z 12 Z 21

sCf Rd + 1
= ( ) ( )
s 3 Cf L f L g + s 2 Cf L f Rg + Rf L g + L g Rd + L f Rd + sCf Rf Rg + Rd Rg + Rf Rd + L f + L g + Rf + Rg

(16.8)

16.3.2 Design Example

For a PR controller, there are two parameters K p and K i , in addition to the known system
frequency ω. The parameter K i is the gain for the resonant term so it can be chosen to be large.
What is to be determined is the gain K p . Here, this is done with the root-locus approach.
The closed-loop system transfer function in the discrete-time domain is given by

CPR (z)HP (z)HD (z)

G CL (z) = , (16.9)
1 + CPR (z)HP (z)HD (z)

where CPR (z) and HP (z) are the discretised version of the controller (16.1) and the plant
transfer function (16.8), respectively. The processing delay of the PWM inverter is represented
by HD (z) = z −1 (Sera et al. 2005).
For the parameters given in Table 3.1 and the sampling frequency of 5 kHz, the root locus of
the open-loop transfer function CPR (z)HP (z)HD (z) is shown in Figure 16.3 for K i /K p = 200.
When the proportional gain K p is chosen as K p = 1.12, the damping ratio of the system is
Current Proportional-Resonant Control 263

1.5
Gain: 1.12
Pole: 0.504 + 0.341i
1 Damping: 0.641
0.5π/T
0.6π/T
0.4π/T 0.1 Overshoot (%): 7.23
0.7π/T 0.3π/T
0.2
0.3
0.5 0.8π/T 0.4
0.5 0.2π/T
Imaginary Axis

0.6
0.7
0.9π/T 0.8 0.1π/T
0.9
π/T
0
π/T
0.9π/T 0.1π/T

−0.5 0.8π/T 0.2π/T

0.7π/T 0.3π/T
0.6π/T
0.4π/T
0.5π/T
−1

−1.5
−2 −1.5 −1 −0.5 0 0.5 1 1.5 2
Real Axis

Figure 16.3 Root locus for the controller design

ξ = 0.641. Figure 16.4 shows the Bode plots of the open-loop system for different integral
gain K i with K p = 1.12, from which the integral gain is chosen as K i = 200. Since the
controller has a very high gain at the resonant frequency, it has a very good ability to reduce
the steady-state error. The resulting controller in the discrete-time domain is
1.12z 2 − 2.196z + 1.08
CPR (z) = . (16.10)
z 2 − 1.996z + 1

16.4 Experimental Results

The PR controllers implemented in both the αβ frame and the abc frame were evaluated with
the test rig described in Chapter 15 in the grid-connected mode under five different scenarios
to test the steady-state responses without a local load and with a resistive, non-linear and
unbalanced resistive local load connected to the system, and the transient response without a
local load.

16.4.1 Steady-state Performance

16.4.1.1 Without a Local Load
In the steady state, the grid current reference Id∗ was set at 3 A. The reactive power was set at 0
Var (Iq∗ = 0). This corresponds to the unity power factor. All the active power was transferred
to the grid via the step-up transformer. The grid current output i a and its spectra, the current
reference i ref and the tracking error ei are shown in the left column of Figure 16.5 for the
264 Control of Power Inverters in Renewable Energy and Smart Grid Integration

200
Ki=25
Magnitude (dB) Ki=50
100 Ki=200
Ki=500

−100
180
Phase (deg)

−180

−360
1 2 3 4
10 10 10 10
Frequency (rad/sec)

Figure 16.4 Bode plots of the open-loop system for different K i with K p = 1.12

5 5
i i i i
ref a ref a
2.5 2.5
Current [A]

Current [A]

0 0

−2.5 −2.5

−5 −5

1.5 1.5
1 ei 1 ei
Error [A]

Error [A]

0.5 0.5
0 0
−0.5 −0.5
−1 −1
−1.5 −1.5
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time [s] Time [s]

(a) Grid output current , its reference and current tracking error
5 5
THD of ia=4.14% THD of ia=3.84%
Magnitude [%]

Magnitude [%]

4 4
3 3
2 2
1 1
0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Harmonics order Harmonics order

(b) Spectra of the grid output current

Figure 16.5 Experimental results for a PR controller without a local load implemented in the αβ frame
(left column) and in the abc frame (right column)
Current Proportional-Resonant Control 265

5 5
i i i i
ref a ref a
2.5 2.5
Current [A]

Current [A]
0 0

−2.5 −2.5

−5 −5

1.5 1.5
1 ei 1 ei
Error [A]

Error [A]
0.5 0.5
0 0
−0.5 −0.5
−1 −1
−1.5 −1.5
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time [s] Time [s]

(a) Grid output current , its reference and current tracking error
5 5
THD of ia=5.46% THD of ia=5.79%
Magnitude [%]

Magnitude [%]
4 4
3 3
2 2
1 1
0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Harmonics order Harmonics order

(b) Spectra of the grid output current

Figure 16.6 Experimental results for a PR controller with a resistive local load implemented in the αβ
frame (left column) and in the abc frame (right column)

implementation in the αβ frame and in the right column of Figure 16.5 for the implementation
in the abc frame. The recorded current THD, shown in Figure 16.5(b), was 3.84% (right) in
the abc frame compared to 4.14% in the αβ frame (left). Both the tracking performance and
the recorded current THD in the abc frame were better than that in the αβ frame.

16.4.1.2 With a Resistive Load

In this experiment a balanced resistive local load with RA = RB = RC = 12  was connected
to the system. The grid current reference Id∗ was set at 2 A, after connecting the inverter to
the grid. The reactive power was set at 0 Var (Iq∗ = 0), which corresponds to the unity power
factor. The grid current output i a and its spectra, the reference current i ref and the current
tracking error ei are shown in the left column of Figure 16.6 for the implementation in the αβ
frame and in the right column of Figure 16.6 for the implementation in the abc frame. The
PR controller implemented in the abc frame again demonstrated better tracking performance,
although with a higher THD due to the increased 3rd harmonics.

16.4.1.3 With an Unbalanced Load

In this experiment, an unbalanced resistive local load with RA = 12 , RB = ∞ and RC =
12  was connected to the system. The grid output current reference Id∗ was set at 2 A (after
connecting the inverter to the grid) and the reactive power was set at 0 Var (Iq∗ = 0). The output
266 Control of Power Inverters in Renewable Energy and Smart Grid Integration

5 5
i i i i
ref a ref a
2.5 2.5
Current [A]

Current [A]
0 0

−2.5 −2.5

−5 −5

1.5 1.5
1 ei 1 ei
Error [A]

Error [A]
0.5 0.5
0 0
−0.5 −0.5
−1 −1
−1.5 −1.5
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time [s] Time [s]

(a) Grid output current , its reference and current tracking error
5 5
THD of ia=5.22% THD of ia=5.39%
Magnitude [%]

Magnitude [%]
4 4
3 3
2 2
1 1
0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Harmonics order Harmonics order

(b) Grid output current harmonic spectra

Figure 16.7 Experimental results for a PR controller with an unbalanced resistive local load imple-
mented in the αβ frame (left column) and in the abc frame (right column)

current i a and its spectra, the reference current i ref and the corresponding current tracking error
ei in the steady state are shown in the left column of Figure 16.7 for the implementation in
the αβ frame and in the right column of Figure 16.7 for the implementation in the abc frame.
The tracking performance of the PR controller implemented in the αβ frame was remarkably
worse than the one implemented in the abc frame, although the recorded THD in the αβ frame
was slightly better than that in the abc frame.

16.4.1.4 With a Non-linear Load

In this experiment, a three-phase uncontrolled rectifier loaded with an LC filter L = 150 μH,
C = 1000 μF and a resistor R = 20  was connected to the system. The grid current output
i a and its spectra, the reference current i ref and the current tracking error ei are shown in the
left column of Figure 16.8 for the implementation in the αβ frame and in the right column of
Figure 16.8 for the implementation in the abc frame. While the tracking performance in both
frames were quite similar, the current THD in the abc frame was slightly better than that in
the αβ frame.

16.4.2 Transient Performance

In this experiment, a step change in the current Id∗ reference from 2 A to 3 A was applied
(while keeping Iq∗ = 0). The grid current output i a , the current reference i ref and the tracking
Current Proportional-Resonant Control 267

5 5
iref ia iref ia
2.5 2.5
Current [A]

Current [A]
0 0

−2.5 −2.5

−5 −5

1.5 1.5
1 ei 1 ei
Error [A]

Error [A]
0.5 0.5
0 0
−0.5 −0.5
−1 −1
−1.5 −1.5
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time [s] Time [s]

(a) Grid output current , its reference and current tracking error
15 15
THD of i =17.64% THD of i =16.71%
Magnitude [%]

Magnitude [%]
a a
10 10

5 5

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Harmonics order Harmonics order

(b) Grid output current harmonic spectra

Figure 16.8 Experimental results for a PR controller with a non-linear local load implemented in the
αβ frame (left column) and in the abc frame (right column)

error ei are shown in the left column of Figure 16.9 for the implementation in the αβ frame and
in the right column of Figure 16.9 for the implementation in the abc frame. Both controllers
demonstrated very fast dynamics, following the reference signal i ref closely. The implemen-
tation in the abc frame demonstrated better performance than the implementation in the
αβ frame.

5 5
i i i i
ref a ref a
2.5 2.5
Current [A]

Current [A]

0 0

−2.5 −2.5

−5 −5

1.5 1.5
1 ei 1 ei
Error [A]

Error [A]

0.5 0.5
0 0
−0.5 −0.5
−1 −1
−1.5 −1.5
−0.04 −0.03 −0.02 −0.01 0 0.01 0.02 0.03 0.04 −0.04 −0.03 −0.02 −0.01 0 0.01 0.02 0.03 0.04
Time [s] Time [s]

Figure 16.9 Transient response of the PR controller in the grid-connected mode without a local load
implemented in the αβ frame (left column) and in the abc frame (right column)
268 Control of Power Inverters in Renewable Energy and Smart Grid Integration

16.5 Summary
A PR controller is able to eliminate the steady-state error when tracking a sinusoidal signal.
It can be implemented in the stationary reference (αβ) frame and the natural (abc) frame. It
is also possible to add a harmonic compensator to improve the power quality. Experimental
results have shown that the tracking performance in the abc frame is generally better than that
in the αβ frame because the controllers for the three phases in the abc frame are independent
from each other.