energies

Article
A Proportional Resonant Controller Scheme for
Suppressing Resonance in Grid tied
Multilevel Inverter
Hemakesavulu Oruganti ID
, Subranshu Sekhar Dash, Chellammal Nallaperumal ID
and
Sridhar Ramasamy *
Department of Electrical and Electronics Engineering, SRM Institute of Science & Technology, Chennai 603203,
India; hkesavulu6@gmail.com (H.O.); sekhardash.s@ktr.srmuniv.ac.in (S.S.D.);
chellammal.n@ktr.srmuniv.ac.in (C.N.)
* Correspondence: sridharmanly@gmail.com; Tel.: +91-9176652907




Received: 28 February 2018; Accepted: 13 April 2018; Published: 24 April 2018


Abstract: Photovoltaic (PV) resources are connected to power grid through voltage source inverters.
The quality of power output from PV inverter should be in grid compliance of IEEE standard. In this
regard, the deployment of appropriate low pass filters such as inductor (L), capacitor (C) or inductor
capacitor inductor (LCL) is critical as they aid in minimizing the harmonics being injected into
the grid. LCL filters are well entrenched but they bring in stability issue due to resonance and
therefore a damping controller with suitable control logic is needed. In this work, to suppress
resonance, a Proportional Resonant-Derivative (PR-D) controller has been designed, proposed, and
compared with existing counterparts, i.e., two-degree of freedom controller (2DOF) and feedback
current controller. The results exhibits that PR-D controller admits meliorate resonance damping and
constancy when compared with the two other schemes. The whole system has been simulated in
MATLAB/Simulink environment and a prototype has also been made to ensure the performance.

Keywords: photovoltaic; two degrees of freedom; proportional resonant-derivative; feedback current

controller; maximum power point tracking; multilevel inverter

1. Introduction
Photovoltaic (PV) power generation has become the most promising way of power generation
among the renewable sources due its inherent advantages. The major hurdle that PV system faces,
similar to other renewable sources, is the power intermittency with respect to atmospheric conditions
such as irradiation and temperature. With the advent of advanced controllers and efficient dc-dc
converters, the intermittency issues are meritoriously addressed [1]. In a contemporary note, many
novel dc-dc converters have been introduced in research arena, but the appeal remains the same in
case of the three well entrenched converters, i.e. buck, boost, and buck boost converters. The choice of
dc-dc converters depends on the voltage levels of source and load. The role of the dc-dc converters
in PV system is very critical, as it only regulates the voltage but also makes most of the PV power
available through an intelligent maximum power tracking (MPPT) algorithm [2]. MPPT facilitates the
yield of maximum available power for the given atmospheric conditions. Numerous MPPT algorithms
have been archived in research forum [3]. The modified and new MPPT algorithms claimed to be
novel either in alleviating the complexity involved in execution of the algorithm in real time or in
making the algorithm not prone to oscillations and inaccuracies [4]. Nevertheless, two well established
MPPT algorithms, perturb and observe (P&O) and incremental conductance (INC), remain the most
prevalent choices in the PV industry [5]. P&O is a simple and cost effective MPPT controller but the

Energies 2018, 11, 1024; doi:10.3390/en11051024 www.mdpi.com/journal/energies

Energies 2018, 11, 1024 2 of 15

oscillation in the power output due to the inherent comparative search mechanism is indispensable.
INC, on the other hand, is the most promising algorithm, as it is efficient and not oscillatory like its
counterpart [6], and remains the most prominent and highly recommended MPPT algorithm.
PV with MPPT aided dc-dc converters need to be interfaced with inverters when it serves AC
loads or grid. The output voltage profile, the switching losses/stress, stability issues and its total
harmonic distortion (THD) are the direct indicators for the performance of the inverter. The THD of the
inverter is considerably reduced when the choice of the inverter topology and the switching techniques
are done prudently [7]. There are several switching techniques proposed in research platform and
these techniques possess their own merits and demerits. The sinusoidal pulse width modulation
(SPWM) technique still lingers as the simplest and most effective one [8]. Pertaining to the topological
modifications, multilevel inverter (MLI) since its inception has been the most effective power converter
in terms of handling high voltage and high power [9]. The research articles archived pertaining to
several topologies of MLI is huge and among them reduced cascaded MLI is most attracting one [10].
Recently, there has been a considerable increase in the importance of multilevel inverters, since they
do not rely on just two levels of voltage to create an AC signal. Alternatively, different voltage levels
are combined to each other resulting in a smoother stepped waveform with less dv/dt and lower
harmonic distortions. There are several types of topologies of a multilevel inverter to produce a
stepped waveform, including neutral clamped diode MLI (NPCMLI), flying capacitor multilevel
inverter and so on. The inherent quality power output of MLI has attracted researchers to explore new
topologies in MLI [11–13]. Among all MLI topologies, the reverse voltage topology MLI (RVT MLI) is
famous in research arena for its reduced switch count. This particular topology is very compatible with
distributed sources such as PV, fuel cell etc., as the sources of medium power can be easily cascaded
to constitute a higher capacity inverter. On the other hand, since the topology has a cascaded unit
structure, factors such as voltage stress on the switch, switching loss, and the issues with selection of
the switch are easily sorted [14]. Several articles have been archived in the field of modular multilevel
converters; Mehrasa has done extensive research in analyzing the regulation of the modular inverter
and its stability aspects during varying load parameters [15,16].
PV MLI with grid interface should comply with the grid standard and therefore filters selection
and design is critical [17,18]. The selection of the LCL filter is the best choice since it has advanced
aspects [18] in comparison with other filters. PV as a source injects current into the grid once the
inverter output is synchronized with grid with respect to voltage and frequency, and therefore it
is vital that the profile of the current is sinusoidal to lessen the harmonic impact. The typical PI
controller based structure may be a weak candidate for mitigating the current harmonic issue due to
the bandwidth limitation [19–21]. To augment the bandwidth a high proportional controller is always
preferred but the major hurdle witnessed here is that the system stability may become unwieldy
if the LC tunes resonance for a particular ripple frequency. Therefore, control logic and damping
methods become crucial parameters in grid synchronized system. Although some research addressed
PR controller in PV grid interfaces, less research addressed PV grid interfaces through MLI having
PR controller. This study had that unique feature and, in this work, PR-D controller was considered
by which the damping was done up to−0.16db, thus exhibiting improved operation. The proposed
method was duly compared with its similar counterpart 2DOF active damping. The detailed stability
analysis of the proposed structure clearly envisaged the merits of the proposed system. The paper
is as follows. Section 2 deals with PV modeling with MPPT scheme and the grid interactive MI.
The proposed system is described and it is compared with other prevailing techniques in Section 3.
The concluding remarks are presented in Section 4.

2. System Description
Figure 1 shows system topology of grid interfaced reverse voltage topology MLI with LCL filter.
This topology comprises PV Panel, MPPT aided boost converter and a multilevel inverter unit that
Energies 2018, 11, 1024 3 of 15
Energies 2018, 11, x FOR PEER REVIEW 3 of 14
Energies 2018, 11, x FOR PEER REVIEW 3 of 14
facilitates single-phase
that facilitates seven-level
single-phase stepped
seven-level output
stepped and LCL
output filterfilter
and LCL connecting thethe
connecting inverter output
inverter outputto the
gridthat
and controller
to facilitates
the grid and to regulate
single-phase the inverter.
controller seven-level stepped
to regulate the output and LCL filter connecting the inverter output
inverter.
to the grid and controller to regulate the inverter.

Figure 1. Grid interfaced RVT multi-level inverter fed with LCL filter.
Figure 1. Grid interfaced RVT multi-level inverter fed with LCL filter.
Figure 1. Grid interfaced RVT multi-level inverter fed with LCL filter.
2.1. PV Array Modeling
2.1.2.1.
PVPVArray Modeling
Array Modeling
The equivalent circuit of a PV cell is shown in Figure 2.
TheThe
equivalent
equivalentcircuit
circuitof
of aa PV cellisisshown
PV cell shownininFigure
Figure
2. 2.
Rs I
Rs I +
ID Ish
+
ID Ish
Iph
Rsh V
Iph
Rsh V

-
-

Figure 2. PV cell equivalent circuit.

Figure 2. PV cell equivalent circuit.
Figure 2. PV cell equivalent circuit.
The photovoltaic panel can be modeled mathematically as given in Equations (1)–(4).
Module
The photonpanel
photovoltaic current:
can be modeled mathematically as given in Equations (1)–(4).
TheModule
photovoltaic panel can be modeled mathematically as given in Equations (1)–(4).
photon current:
I = [I + K (T − 298)] * λ / 1000 (1)
Module photon current: ph Scr i
I = [I + K (T − 298)] * λ / 1000 (1)
ph Scr i
Modules reverse saturation I phcurrent:
= [ IScr + Ki ( T − 298)] ∗ λ/1000 (1)
Modules reverse saturation current:
I rs = I Scr /[exp( qV oc / N s KAT ) − 1 ] (2)
Modules reverse saturation = I Scr /[exp( qV oc / N s KAT ) − 1 ]
I rs current: (2)
Module saturation current I0 differs with the cell temperature:
Module saturation current I0 differs with the cell temperature:
Irs = IScr /[exp 3 (qVoc /Ns KAT ) − 1] (2)
T   q * Eg 1 1 
I 0 = I rs  3 expq* Eg 0 { − } (3)
 T T  0 { 1 Tr− 1 }T 
Module saturation current II0 differs
0 = I rs    r
with  the
exp  
cell BK
temperature:  (3)
Tr   BK Tr T  
The current output of PV module:
T 3 q ∗ Eg0 1 1
The current output of PV module:
I0 = Irs exp { − } (3)
Tr  BKq * (V T pvr + I T
pv Rs ) 
I pv = N p * I ph − N p * I0 exp{ q * (V pv + I pv Rs ) } −1 (4)
 N s AKT } − 1 
The current output of I pvPV=N *
module:
p I ph − N p 0
* I exp{  (4)
 N s AKT  
By the advanced formulations, the PV module features are predicted, as shown in Figure 3. To
q ∗ (Vpv + I pv Rs )

By the advanced
meet I pv = Np ∗ I ph
necessary formulations,
specifications of−the
the PV ∗ I0 exp
Npmodule
application, { the are
features of the}as
predicted,
output −shown
PV in Figure
1module 3. Toby
is boosted (4)
meet the of
means necessary specifications of the application, the output
a power converter. Ns AKT of the PV module is boosted by
means of a power converter.
By the advanced formulations, the PV module features are predicted, as shown in Figure 3.
To meet the necessary specifications of the application, the output of the PV module is boosted by
means of a power converter.
Energies 2018, 11, 1024 4 of 15
Energies 2018, 11, x FOR PEER REVIEW 4 of 14

Energies 2018, 11, x FOR PEER REVIEW 4 of 14

Figure3.3.I-V
Figure I-Vcharacteristics
characteristicsand
andP-V
P-Vcharacteristics.
characteristics.

Figure 3. I-V characteristics and P-V characteristics.

2.2. MPPT Controller
2.2. MPPT Controller
2.2.
TheMPPT
MPPT Controller
unit will make sure that maximum power is rendered to the grid for the given
The MPPT unit will make sure that maximum power is rendered to the grid for the given
atmospheric conditions.
The conditions.
MPPT unit willManymakealgorithms have
sure thathave
maximum beenpower
administered so far
is rendered in the
to in
the research
grid forum and
for theforum
given
atmospheric Many algorithms been administered so far the research and
this atmospheric
work adopted incremental
conditions. Many conductance
algorithms (INC)(INC)
have been method of MPPT.
administered farThe INC method is superior
and to itsto
this work adopted incremental conductance method of MPPT.soThe in the
INC research
method isforum
superior
its closework
competitor
adoptedperturb andconductance
observe (P&O) in termsofofMPPT.
reliable,
Theoscillation
INC methodfree power to output.
close this
competitor perturbincremental
and observe (P&O) (INC)
in termsmethod
of reliable, oscillation is superior
free power output. The
The its
flowclose competitor
diagram of INCperturb
MPPT andisobserve
given in(P&O)
Figure in 4.
terms of reliable, oscillation free power output.
flow diagram of INC MPPT is given in Figure 4.
The flow diagram of INC MPPT is given in Figure 4.

Figure 4. INC MPPT.

Figure 4. INC MPPT.

Figure 4. INC MPPT.

 Energies 2018, 11, 1024 5 of 15
Energies 2018, 11, x FOR PEER REVIEW 5 of 14

Theincremental
The incremental conduction
conduction MPPT
MPPTchecks
checksforforthethechange
change are are
slope value
slope of theofpower
value with respect
the power with
to voltage
respect and propagates
to voltage towards
and propagates the peak
towards thepower, i.e., if dP/dV
peak power, is negative,
i.e., if dP/dV then MPPT
is negative, will be
then MPPT on
will
the
be onright side ofside
the right the of
current position,
the current while, while,
position, if it is positive, then thethen
if it is positive, peakthe
will be situated
peak on the left
will be situated onof
the
theleft of thesearch
current current search point.
point.

2.3.Grid
2.3. GridInterfaced
InterfacedRVT
RVTMLI
MLI
Bycomparison
By comparisonofofthe
theabove
abovementioned
mentionedMLI MLItypes,
types,the
thereverse
reversevoltage
voltagemultilevel
multilevelinverter
inverterwas
was
projected during 2008 as a new topology to challenge the existing popular topologies. The
projected during 2008 as a new topology to challenge the existing popular topologies. The circuitcircuit
diagramofofRVTMLI
diagram RVTMLIisisshown
shownininFigure
Figure5.5.

PV
S

Full Bridge

6
2
PV 7 9

8 10 Grid
PV
4

Low Frequency
5 Switches
Polarity
Generation
High Frequency
Switching Level
Generation

Figure 5. Single-phase multi-level inverter RV topology.

Figure 5. Single-phase multi-level inverter RV topology.

Modes of Operation of RVT MLI

Modes of Operation of RVT MLI
Switching sequence in this inverter is very simple. While selecting a particular switching
Switching sequence in this inverter is very simple. While selecting a particular switching pattern,
pattern, make sure that switching transitions are minimum. There are six switching patterns to
make sure that switching transitions are minimum. There are six switching patterns to control the
control the inverter. The sequence of switches (2-3-4), (2-3-5), (2-6-5), and (1-5) are used for levels
inverter. The sequence of switches (2-3-4), (2-3-5), (2-6-5), and (1-5) are used for levels 0–3, respectively.
0–3, respectively. The output voltage will be the sum of voltage sources that are included in the
The output voltage will be the sum of voltage sources that are included in the current path. Current
current path. Current path for each level is shown in Figure 6 according to the above switching
path for each level is shown in Figure 6 according to the above switching pattern.
pattern.
The number of components is determined using the equations given below:
The number of components is determined using the equations given below:
M
Mswitch [(M−−
switch ==[(M 1)1) + 4]
+ 4] (5)(5)

MMcarriers
carriers or Mdc = (M − 1)/2
or Mdc = (M − 1)/2 (6)(6)
where,
where,M
Misisthe
thenumber
numberof
oflevels,
levels, M
Mcarriers is the required number of carriers, and Mdc is the required
carriers is the required number of carriers, and Mdc is the required
number of DC sources.
number of DC sources.
The total number of components (Tc) is given by,
Tc = {[(M − 1) + 4] + Mcarriers+ Mdc} (7)
 Energies 2018, 11, 1024 6 of 15

The total number of components (Tc) is given by,

Energies 2018, 11, x FOR PEER REVIEW 6 of 14
Tc = {[(M − 1) + 4] + Mcarriers + Mdc } (7)
Energies 2018, 11, x FOR PEER REVIEW 6 of 14
PV PV
S S

1 1
PV PV
S S

1 1
6 6
2 2
PV PV
6 6
Level 1
3 Level 0 3
2 2
PV PV
Level 1
PV 3 Level 0 PV 3
4 4

PV PV
4 4
5 5

5 5

PV
PV
S S

1 PV 1
PV
S S

1 1
6
6
2 2
PV PV
6
6 Level 2
3 2 PV
3 2 Level 3
PV
Level 2
3 3 Level 3
PV PV
4 4

PV PV
4 4

5 5

5 5

Switchingsequence
Figure6.6.Switching
Figure sequenceofofRVT
RVTMLI.
MLI.

For
For60-degree
60-degreePWMPWMmethod, Figure 6. Switching
method,modulating
modulating sine sequence
sinewave
wave isisflatoftopped
flat RVT MLI.
toppedfor
foraaperiod
periodofof60
60degrees
degreesinin
each half cycle, as shown in Figure 7.
each half cycle, as shown in Figure 7.
For 60-degree PWM method, modulating sine wave is flat topped for a period of 60 degrees in
each half cycle, as shown in Figure 7.

Figure
Figure7.7.Generated
Generatedreference
referencewaveform
waveformusing
using60-degree
60-degreePWM.
PWM.
Figure 7.seven-level
The simulated single-phase Generated reference waveform
output voltage using
and 60-degree
harmonic PWM. with PWM is
spectrum
The simulated single-phase seven-level output voltage and harmonic spectrum with PWM is
shown in Figure 8a,b.
shown inThe Figure
In Figure 8b,8a,b.
simulated single-phase
the sinusoidal outputseven-level
voltage ofoutput voltage
the reverse and harmonic
voltage topology spectrum
multilevelwith PWM is
inverter
shown in Figure 8a,b.
has less total harmonic distortion and no even or triplen harmonics. LCL filter is used to integrate
RVT MLI In andFigure 8b, the function
grid whose sinusoidalis output voltage
to get rid of theofcurrent
the reverse voltage
harmonics topology
injected multilevel
to the inverter
grid. Thus,
has less total harmonic distortion and no even or triplen harmonics. LCL filter
output of the inverter is connected to the LCL filter, which is explained in flowing sections. is used to integrate
RVT MLI and grid whose function is to get rid of the current harmonics injected to the grid. Thus,
output of the inverter is connected to the LCL filter, which is explained in flowing sections.
 Energies 2018, 11, 1024 7 of 15

In Figure 8b, the sinusoidal output voltage of the reverse voltage topology multilevel inverter has
less total harmonic distortion and no even or triplen harmonics. LCL filter is used to integrate RVT
MLI and grid whose function is to get rid of the current harmonics injected to the grid. Thus, output of
Energies 2018, 11,
the inverter is xconnected
FOR PEER REVIEW
to the LCL filter, which is explained in flowing sections. 7 of 14
Energies 2018, 11, x FOR PEER REVIEW 7 of 14

(a)
(a)

(b)
(b)
Figure
Figure8.8. (a) Output
(a)Output
8.(a) voltage
Outputvoltage waveform
waveformofof
voltagewaveform RVTMLI,
ofRVTMLI, and
RVTMLI,and (b)
and(b) harmonic
(b)harmonic spectrum
harmonicspectrum analysis
analysisofof
spectrumanalysis RVTMLI
ofRVTMLI
RVTMLI
Figure
output voltage.
outputvoltage.
voltage.
output
2.4. Grid Interfaced Inverter with LCL Filter
2.4.Grid
2.4. GridInterfaced
InterfacedInverter
Inverterwith
withLCL
LCLFilter
Filter
The PWM inverter is connected to the grid by LCL filter that acts as low pass filter to filter out
ThePWM
The PWMinverter
inverterisisconnected
connectedtotothe
thegrid
gridby
byLCL
LCLfilter
filterthat
thatacts
actsasaslow
lowpass
passfilter
filtertotofilter
filterout
out
higher order harmonics. However, LCL filter injects two complex conjugate poles to the system and
higherorder
higher orderharmonics.
harmonics.However,
However,LCLLCLfilter
filterinjects
injectstwo
twocomplex
complexconjugate
conjugatepoles
polestotothe
thesystem
systemand and
thus generates resonance in the system. The possible topologies are shown in Figure 9.
thusgenerates
thus generatesresonance
resonanceininthe
thesystem.
system.TheThepossible
possibletopologies
topologiesare
areshown
shownininFigure
Figure9.9.

-
-
-
-

VSI VSI VSI

VSI VSI VSI

(a) (b) (c)

(a) (b) (c)

Figure 9. Basic filter topologies: (a) L filter, (b) LC filter and (c) LCL filter.
Figure 9. Basic filter topologies: (a) L filter, (b) LC filter and (c) LCL filter.
Figure 9. Basic filter topologies: (a) L filter, (b) LC filter and (c) LCL filter.
Modeling of grid interfaced inverter with LCL Filter
Modeling of grid interfaced inverter with LCL Filter
Modeling
BY KVLof grid
and interfaced inverter
KCL equations werewith LCL Filter
applied:
BY KVL and KCL equations were applied:
BY KVL and KCL equations were applied:
di 1 1
di gg = 1 v − 1 v (8)
di = 1v c− 1 v g
dt g L c L g (8)(8)
dtdt =L 2L22 vc −LL222v g
di inv 1 1
didi
inv = 1 1 v inv − 11 v c (9)
dt == L1vvinv
inv −− L vvcc (9)(9)
dtdt L1L1 inv LL111
dv c 1 1
dv c = 1 iinv −1 i g (10)
dt = i
C inv − Ci g (10)
dt C C
To evaluate the system, the state-space model should be transformed into the standard
To evaluate the system, the state-space model should be transformed into the standard
canonical form. The state variables selected are
canonical form. The state variables selected are
Energies 2018, 11, x FOR PEER REVIEW 8 of 14

Energies 2018, 11, 1024 x1 = i g 8 of 15

(11)

•
dvc 1 1
= x1 i=invdi−
x2 = g dti g (12)
(10)
dt C C
• should be transformed into the standard canonical
To evaluate the system, the state-space model
2 2
form. The state variables selected are x 3 = x 2 = d i g dt (13)

By differentiating Equation (8), x1 = i g (11)

• 2 .
d i g x2 1= x11= 1 
di g /dt 1 • (12)
x2 = = i − i − Vg (14)
L2x2 C= d2 i g /dt
C 2  L2
2 . inv g
dt x3 = (13)

By differentiating Equation 3
• d (8),
i g 1  1 1  1  1 1  1 ••
x3 =
3
=  vinv − vc  −  vc − vg  − Vg (15)
. 2 d i 1 1 1 1  2 1  2 1 2 .  2
dt L C 2L L L C L L L
g
x2 = 2 = i − i g − Vg (14)
Therefore, the system is formulated L2 C inv form
dt in canonical C as follows:
L2
d3 i g 1.
   
1 1 1 1 1 1 ..
x3 = 3 =   vinv − vc −
.
vc − v g − Vg (15)
dt L2 C  L1  0 1L1 0  xL2 C
x 1  L2 0 L2 L2
 x  =  0 0 1  x  +  0 
. 1
 . 2 in canonical
Therefore, the system is formulated  form  as2follows:
 
  (16)
2
  0 − ω o   
 .  x 3 0  x3   ε cos
 ωt 
x1 

 .    0 1 0 x1 0
 x2  =  0 0 1   x2  +  0 (16)
   

.
The transfer functionxof LCL filter
3 0 is−shown
ωo2 0 belowx3and the filter ε cosmodel
ωt is shown in Figure 10.
Applying KVL and KCL to LCL filter, the transfer function of LCL filter is given by
The transfer function of LCL filter is shown below and the filter model is shown in Figure 10.
1
H LCL =
( )
Applying KVL and KCL to LCL filter, the transfer
3 function of LCL filter is given by (17)
s CL1 L 2 + s L1 + L 2
1
The transfer function of LCL filter = damping
HLCLwith (17)
s3 CL1 L2 +resistor
s( L1 + L(Rd)
2)
is given as

R d Cs(Rd)
The transfer function of LCL filter with damping resistor + 1 is given as
G LCL (s ) =
3
L L Cs + L1 +
GLCL (s) = 1 2 3
(
RdLCs 2 CR
2
)
+ 1d s + L1 + L2 s ( ) (18)
(18)
L1 L2 Cs + ( L1 + L2 )CRd s2 + ( L1 + L2 )s
The cut-off frequency of the filter can be calculated as
The cut-off frequency of the filter can be calculated as
s Li + L g
Fres = Li + L g (19)
Fres = L L g Cf (19)
Lii L g C f

L1 iinv L2 ig
+ + +
ic

Vin C VC Vg

- - -

Figure
Figure 10.
10. Topology
Topology of
of LCL
LCL filter.
filter.

The parameters of the design filter and parameters required for filter design are summarized in
The parameters of the design filter and parameters required for filter design are summarized in
Table 1.
Table 1.
 Energies 2018, 11, 1024 9 of 15
Energies 2018, 11, x FOR PEER REVIEW 9 of 14
Energies 2018, 11, x FOR PEER REVIEW 9 of 14

Table 1. Parameters for LCL filter design.

Table 1. Parameters for LCL filter design.
Inverter side inductance, L1 17.67 mH
Inverter
Inverter side inductance,
side inductance, L1 L1 17.6717.67
mH mH
Grid
GridGrid
side side inductance,
inductance, L2 L2 5.7 mH5.7 mH
side inductance, L2 5.7 mH
Filter capacitor,
Filter capacitor, C C 10 µF 10 µF
Filter
Damping capacitor,
Resistor, Rd C Rd 10 µF
11.3 ohm
Damping Resistor, 11.3 ohm
Damping Resistor, Rd 11.3 ohm

The Bode
Bodeplot
plot of the inverter with the designed LCL without
filter without damping is shown in
The Bode plot of
of the
theinverter with
inverter thethe
with designed LCLLCL
designed filter damping
filter without is shown
damping is in Figurein11
shown
Figure
and the11 and the resonance
resonance peak is peak is between
observed observed0between
and 123 0 and
db at 123 db
1305 Hz.at 1305 Hz.
Figure 11 and the resonance peak is observed between 0 and 123 db at 1305 Hz.
Bode Diagram
Bode Diagram
Phase (deg) Magnitude (dB)
Phase (deg) Magnitude (dB)

100
100
0
0
-100
-100
-90
-90
-135
-135
-180
-180
-225
-225
-270
-270 3 4 5
10
3 10
4 10
5
10 10(rad/s)
Frequency 10
Frequency (rad/s)
Figure 11. Bode
Bode plot
plot of grid connected LCL filter.
Figure11.
Figure 11.Bode plot ofof grid
grid connectedLCL
connected LCLfilter.
filter.

3. PR-D Prevails over Two Degrees and Current Feedback Controller

3.3.PR-D
PR-DPrevails
Prevailsover
overTwoTwoDegrees
DegreesandandCurrent
CurrentFeedback
FeedbackController
Controller
From the Bode plot in Figure 11, it is apparent that this filter introduces resonance to the system
From
Fromthe
theBode
Bodeplot
plotininFigure
Figure11,
11,ititisisapparent
apparentthat
thatthis
thisfilter
filterintroduces
introducesresonance
resonancetotothe
thesystem
system
that causes generation of harmonics and distortion in voltage and current fed to grid.
that
thatcauses
causesgeneration
generationofofharmonics
harmonicsandanddistortion
distortionininvoltage
voltageand andcurrent
currentfed
fedtotogrid.
grid.
Using Figure 12, two degrees of freedom controller is designed and the simulated output voltage
Using
UsingFigure
Figure12,
12,two
twodegrees
degreesof
offreedom
freedom controller
controller is
is designed and the simulated
simulated output
outputvoltage
voltage
is shown in Figure 13. We can say that the resonance peak is observed between −6 and −8 db at 1305
isisshown
shownininFigure
Figure13.13.WeWe can say
can that
say that thethe
resonance
resonance peak
peak is observed
is observed between
between −6 −
and −8atdb
−8 db
6 and 1305
at
Hz.
Hz.
1305 Hz.

IRef V Ig
IRef + Vg g + Inv+LCL I
Kp+Ki/s + + Inv+LCL g
+ - Kp+Ki/s - + -
- - -
Ilg K d std
Ilg Vm K d std
td s+ 1
Vm td s+ 1

Figure 12. Block Diagram of two degrees of freedom controller.

Figure 12.12.
Figure Block Diagram
Block ofof
Diagram two degrees
two ofoffreedom
degrees freedomcontroller.
controller.

The block diagram of injected grid current feedback controller [22] is shown in Figure 14a and the
relevant frequency response is shown in Figure 14b, where kg is the factor of the feed forward grid
voltage (here, kg = 1), iref is the reference current and Gc(s) is the gain of the current regulator, which is
used to track iref .

Figure 13. Bode plot for PI-D control of grid connected inverter with LCL filter.
Figure 13. Bode plot for PI-D control of grid connected inverter with LCL filter.
 Ilg K d std
Vm td s+ 1

Energies 2018, 11, 1024 10 of 15

Figure 12. Block Diagram of two degrees of freedom controller.

Energies 2018, 11, x FOR PEER REVIEW 10 of 14

The block diagram of injected grid current feedback controller [22] is shown in Figure 14a and the
relevant frequency response is shown in Figure 14b, where kg is the factor of the feed forward grid
voltage (here, Figure irefBode
kg = 1),13. is theplot for PI-Dcurrent
reference control of grid
and connected
Gc(s) is the inverter
gain of with filter.
LCL filter.
the current regulator, which is
used to track iref.

kg
ug
u
Iref 1 1 1
Gc(s)
L1S C1S L 2S
uinv

-Kcs
S+Wh

(a)

(b)
Figure 14. (a) Block diagram of grid injected feedback current controller. (b) Bode plot for grid
Figure 14. (a) Block diagram of grid injected feedback current controller. (b) Bode plot for grid injected
injected current feedback controller with LCL filter.
current feedback controller with LCL filter.

It can be observed in Figure 15 that the injected grid current has a resonance peak of at 1305 Hz.
Using PR-D controller, the resonant peak is reduced to −0.15 db. The block diagram of proposed
PR-Controller is shown in Figure 16. In this section, the prevailing two degrees of freedom controller
and grid injected current controller schemes are compared with the suggested PR-D controller.
 Energies 2018, 11, 1024 11 of 15

It can be observed in Figure 15 that the injected grid current has a resonance peak of at 1305 Hz.
Using PR-D controller, the resonant peak is reduced to −0.15 db. The block diagram of proposed
PR-Controller is shown in Figure 16. In this section, the prevailing two degrees of freedom controller
Energies
and grid 2018, 11, x FOR
injected PEER REVIEW
current controller schemes are compared with the suggested PR-D controller.11 of 14
Energies 2018, 11, x FOR PEER REVIEW 11 of 14

Figure 15. Bode

Bodeplot
plot for feedback current controller with LCL filter.
Figure
Figure 15.15.
Bode plot forfor feedback
feedback current
current controller
controller with
with LCL
LCL filter.
filter.

IRef Vg I
IRef GPR Vg + Inv+LCL Ig g
+ GPR + Inv+LCL
+ - + - + -
- - -
Ilg K d s td
Ilg Vm K dtds std+ 1
Vm td s + 1

Figure 16. Block diagram of PR-D controller.

Figure 16.16.
Figure Block diagram
Block of of
diagram PR-D controller.
PR-D controller.
PR controller was incorporated in this study to enhance the accuracy of the control system. The
PR controller was incorporated in this study to enhance the accuracy of the control system. The
important characteristic
PR controller feature is at resonance
was incorporated in this studyfrequency to contribute
to enhance infiniteofgain.
the accuracy This controller
the control system.
important characteristic feature is at resonance frequency to contribute infinite gain. This controller
has the ability to compensate known and fixed frequency. Signals that are
The important characteristic feature is at resonance frequency to contribute infinite gain. This controllercontrolled by PRD
has the ability to compensate known and fixed frequency. Signals that are controlled by PRD
controller are given
has the ability to the PWM.
to compensate known and fixed frequency. Signals that are controlled by PRD controller
controller are given to the PWM.
G is the transfer
are given to the PWM.
PR function of PR (proportional resonant) controller, which is given as
GPR is the transfer function of PR (proportional resonant) controller, which is given as
GPR is the transfer function of PR (proportional resonant) controller, which is given as
2 K ωc s
GPR = K p +  2 K ihih
ωc s (20)
GPR = K p + i =1,3,5.   s 2 2K + 2 ω ωsc s+ ω 2 (20)(20)
GPR = K p + ∑ s 2 + 2ω s + ω 2h
i =1,3,5. 2 ih
c 2
i =1,3,5.... s + 2ωcc s + ω hh
where Kp is the proportional gain tuned in the same manner as performed for a PI controller, and it
where
whereKpKdecides
is the proportional
p is the the
gain tuned
proportional in the same manner as performed for a PIacontroller, andandit
mainly dynamics gain tuned
of the in the
system in same
terms manner
of bandwidth,as performed
phase, for PI controller,
and gain margin. The
mainly decides
it mainly decides the dynamics
thespectrumof
dynamics the system
of the of in
system terms of
in voltage bandwidth,
terms ofand bandwidth,phase, and
phase, gain margin.
and gain The
margin.
simulated harmonic analysis output grid injected current is shown in
simulated
The harmonic
simulated spectrum
harmonic analysis
spectrum of output
analysis of voltage
output and grid
voltage and injected
grid current
injected is shown
current is in
shown
Figures 17 and 18, respectively. The THD values comply to the industrial standard and IEEE
Figures 17 and
in Figures 18, 18,
respectively. TheThe THD values comply to thethe industrial standard and
andIEEE
519-1992 as 17
theand
voltagerespectively.
harmonics are below THD 3%values
and the comply
currenttoharmonics industrial standard
are also less at PCC. IEEE
519-1992 as the voltage harmonics are below 3% and the current harmonics
519-1992 as the voltage harmonics are below 3% and the current harmonics are also less at PCC. are also less at PCC.

Figure 17. Grid output voltage harmonic profile.

Figure 17. Grid output voltage harmonic profile.
where Kp is the proportional gain tuned in the same manner as performed for a PI controller, and it
mainly decides the dynamics of the system in terms of bandwidth, phase, and gain margin. The
simulated harmonic spectrum analysis of output voltage and grid injected current is shown in
Figures 17 and 18, respectively. The THD values comply to the industrial standard and 12IEEE
Energies 2018, 11, 1024 of 15
519-1992 as the voltage harmonics are below 3% and the current harmonics are also less at PCC.

Figure
Figure 17.
17. Grid
Grid output
output voltage
voltage harmonic
harmonic profile.
profile.
Energies 2018, 11, x FOR PEER REVIEW 12 of 14

Figure
Figure 18.
18. Grid
Gridinjected
injected current
current harmonic
harmonic profile.
profile.

Bode plot is drawn with PR-D controller fed LCL filter (Figure 19), and we can distinguish that
Bode plot is drawn with PR-D controller fed LCL filter (Figure 19), and we can distinguish that
the resonant peak is reduced, which is observed between −0.15 and 0 db at 1305 Hz. This shows the
the resonant peak is reduced, which is observed between −0.15 and 0 db at 1305 Hz. This shows the
improved performance of the system with PR controller when compared to the two degrees of
improved performance of the system with PR controller when compared to the two degrees of freedom
freedom controller and grid injected feedback current controllers.
controller and grid injected feedback current controllers.

Bode Diagram
100

50
B))
(dB
e(d

0
de
itud
nitu

-50
M
Magn
ag

-100

-150
-90

-135
(de
e(d g))
eg

-180
ha
Ph
P se
as

-225

-270
1 2 3 4 5 6
10 10 10 10 10 10
Frequency (rad/s)

Figure
Figure19.
19.Bode
Bodeplot
plotfor
forPR-D
PR-Dcontrol
controlof
of grid
grid connected
connected inverter
inverter by
by LCL
LCL filter.
filter.

The hardware output for seven-level MLI is shown in Figure 20a and the whole prototype setup
The hardware output for seven-level MLI is shown in Figure 20a and the whole prototype setup
is shown in Figure 20b. The experimental hardware platform presents the MLI output. The grid
is shown in Figure 20b. The experimental hardware platform presents the MLI output. The grid
synchronization in real time is achieved through zero crossing detectors (ZCD) with phase locked
synchronization in real time is achieved through zero crossing detectors (ZCD) with phase locked loop
loop (PLL) circuitry.
(PLL) circuitry.

(a)
 Figure 19. Bode plot for PR-D control of grid connected inverter by LCL filter.

The hardware output for seven-level MLI is shown in Figure 20a and the whole prototype setup
is shown in Figure 20b. The experimental hardware platform presents the MLI output. The grid
synchronization in real time is achieved through zero crossing detectors (ZCD) with phase locked
Energies 2018, 11, 1024 13 of 15
loop (PLL) circuitry.

Energies 2018, 11, x FOR PEER REVIEW 13 of 14

(a)

(b)
Figure
Figure 20.
20. (a)
(a) Hardware
Hardware output
output of
of MLI
MLI unit
unit and
and (b)
(b) Hardware prototype setup.
Hardware prototype setup.

4. Conclusions
4. Conclusions
A PR-D controller-based resonance damping technique is projected in this paper to enhance the
A PR-D controller-based resonance damping technique is projected in this paper to enhance the
dynamic response of overall system and effectively reduce the current component. In a grid linked
dynamic response of overall system and effectively reduce the current component. In a grid linked
inverter, although the third order low pass LCL filter can filter out the output harmonics, it
inverter, although the third order low pass LCL filter can filter out the output harmonics, it introduces
introduces resonance at a particular frequency. Active damping is more flexible and does not cause
resonance at a particular frequency. Active damping is more flexible and does not cause power loss
power loss such as in passive damping. The proposed PR-D controller combines the effect of
such as in passive damping. The proposed PR-D controller combines the effect of proportional resonant
proportional resonant (PR) and derivative (D) controller to dampen the resonance.
(PR) and derivative (D) controller to dampen the resonance.
It has been found that, although 2DOF (PI-D) controller and grid injected feedback current
It has been found that, although 2DOF (PI-D) controller and grid injected feedback current
controllers can dampen the resonance, they fail to remove the resonance completely. The proposed
controllers can dampen the resonance, they fail to remove the resonance completely. The proposed
PR-D controller dampens the resonance to the maximum possible limit and hence is better when
PR-D controller dampens the resonance to the maximum possible limit and hence is better when
compared to its counterparts. Simulation results and mathematical analysis confirmed the
compared to its counterparts. Simulation results and mathematical analysis confirmed the effectiveness
effectiveness of the PR-D controller. In addition, a hardware prototype was realized, accordingly.
of the PR-D controller. In addition, a hardware prototype was realized, accordingly.
Author Contributions:
Author Contributions: All
Allauthors
authorscontributed
contributed equally
equally to the
to the finalfinal dissemination
dissemination ofproposed
of the the proposed research
research work
work as full article.
as full article.
Conflicts
Conflicts of Interest: The
Theauthors
authorsdeclare
declarethat
thatthere is is
there nono
conflict of interests
conflict regarding
of interests the the
regarding publication of this
publication of
this paper.
paper.

References
1. Azary, M.T.; Sabahi, M.; Babaei E.; Meinagh, F.A.A. Modified Single-Phase Single-Stage Grid-Tied Flying
Inductor Inverter with MPPT and Suppressed Leakage Current. IEEE Trans. Ind. Electr. 2018, 65, 221–231.
2. Onar, O.C.; Kobayashi, J.; Erb, D.C.; Khaligh, A. A Bidirectional High-Power-Quality Grid Interface with a
Novel Bidirectional Noninverted Buck–Boost Converter for PHEVs. IEEE Trans. Veh. Technol. 2012, 61,
2018–2032.
3. Hanif, M.; Khadkikar, V.; Xiao, W.; Kirtley, J.L. Two Degrees of Freedom Active Damping Technique for
LCL Filter-Based Grid Connected PV Systems. IEEE Trans. Ind. Electr. 2014, 61, 2795–2803.
4. Jones, M.; Vukosavic, S.N.; Dujic, D.; Levi, E.; Wright, P. Five-leg inverter PWM technique for reduced
Energies 2018, 11, 1024 14 of 15

References
1. Azary, M.T.; Sabahi, M.; Babaei, E.; Meinagh, F.A.A. Modified Single-Phase Single-Stage Grid-Tied Flying
Inductor Inverter with MPPT and Suppressed Leakage Current. IEEE Trans. Ind. Electr. 2018, 65, 221–231.
[CrossRef]
2. Onar, O.C.; Kobayashi, J.; Erb, D.C.; Khaligh, A. A Bidirectional High-Power-Quality Grid Interface with
a Novel Bidirectional Noninverted Buck–Boost Converter for PHEVs. IEEE Trans. Veh. Technol. 2012, 61,
2018–2032. [CrossRef]
3. Hanif, M.; Khadkikar, V.; Xiao, W.; Kirtley, J.L. Two Degrees of Freedom Active Damping Technique for LCL
Filter-Based Grid Connected PV Systems. IEEE Trans. Ind. Electr. 2014, 61, 2795–2803. [CrossRef]
4. Jones, M.; Vukosavic, S.N.; Dujic, D.; Levi, E.; Wright, P. Five-leg inverter PWM technique for reduced switch
count two-motor constant power applications. IET Electr. Power Appl. 2008, 2, 275–287. [CrossRef]
5. Sharma, R.K.; Mishra, S. Dynamic Power Management and Control of a PV PEM Fuel-Cell-Based Standalone
Ac/Dc Microgrid Using Hybrid Energy Storage. IEEE Trans. Ind. Appl. 2018, 54, 526–538. [CrossRef]
6. Yang, F.; Xia, N.; Han, Q.L. Event-Based Networked Islanding Detection for Distributed Solar PV Generation
Systems. IEEE Trans. Ind. Inform. 2017, 13, 322–329. [CrossRef]
7. Jeong, H.-G.; Kim, G.-S.; Lee, K.-B. Second-Order Harmonic Reduction Technique for Photovoltaic Power
Conditioning Systems Using a Proportional-Resonant Controller. Energies 2013, 6, 79–96. [CrossRef]
8. Hu, G.; Chen, C.; Shanxu, D. New Active damping strategy for LCL filter based grid connected inverters
with harmonics compensation. J. Power Electr. 2013, 13, 287–295. [CrossRef]
9. Yao, W.; Chen, M.; Matas, J.; Guerrero, J.M.; Qian, Z.M. Design and Analysis of the Droop Control Method
for Parallel Inverters Considering the Impact of the Complex Impedance on the Power Sharing. IEEE Trans.
Ind. Electr. 2011, 58, 576–588. [CrossRef]
10. Wu, W.; He, Y.; Tang, T.; Blaabjerg, F. A New Design Method for the Passive Damped LCL and LLCL
Filter-Based Single-Phase Grid-Tied Inverter. IEEE Trans. Ind. Electr. 2012, 60, 4339–4350. [CrossRef]
11. Hanif, M.; Khadkikar, V.; Xiao, W.; James, K.L. Two degrees of freedom active damping technique for LCL
filter-based grid connected PV systems. IEEE Trans. Ind. Electr. 2014, 61, 2795–2803. [CrossRef]
12. Zhang, J.; Zheng, W.X. Design of Adaptive Sliding Mode Controllers for Linear Systems via Output Feedback.
IEEE Trans. Ind. Electr. 2014, 61, 3553–3562. [CrossRef]
13. Mehrasa, M.; Pouresmaeil, E.; Catalao, J.P.S. Direct Lyapunov Control Technique for the Stable Operation of
Multilevel Converter Based Distributed Generation in Power Grid. IEEE J. Emerg. Sel. Top. Power Electr. 2014,
2, 931–941. [CrossRef]
14. Mehrasa, M.; Pouresmaeil, E.; Akorede, M.F.; Jorgensen, B.N.; Catalao, J.P.S. Multilevel converter control
approach of active power filter for harmonics elimination in electric grids. Energy 2015, 84, 722–731.
[CrossRef]
15. Mehrasa, M.; Pouresmaeil, E.; Akorede, M.F.; Zabihi, S.; Catalao, J.P.S. Function-based modulation control
for modular multilevel converters under varying loading and parameters conditions. IET Gener. Transm.
Distrib. 2017, 11, 3222–3230. [CrossRef]
16. Mehrasa, M.; Pouresmaeil, E.; Taheri, S.; Vechiu, I.; Catalao, J.P.S. Novel Control Strategy for Modular
Multilevel Converters Based on Differential Flatness Theory. IEEE J. Emerg. Sel. Top. Power Electr. 2017, 99.
[CrossRef]
17. Mehrasa, M.; Pouresmaeil, E.; Akorede, M.F.; Zabihi, S.; Catalao, J.P.S. A multi-loop control technique for the
stable operation of modular multilevel converters in HVDC transmission systems. Int. J. Electr. Power Energy
Syst. 2018, 96, 194–207. [CrossRef]
18. Zhang, N.; Tang, H.; Yao, C. A Systematic Method for Designing a PR Controller and Active Damping of the
LCL Filter for Single-Phase Grid-Connected PV Inverters. Energies 2014, 7, 3934–3954. [CrossRef]
19. Lorzadeh, I.; Abyaneh, A.H.; Savaghebi, M.; Bakhshai, A.; Guerrero, J.M. Capacitor Current Feedback-Based
Active Resonance Damping Strategies for Digitally-Controlled Inductive-Capacitive-Inductive-Filtered
Grid-Connected Inverters. Energies 2016, 9, 642. [CrossRef]
20. Zhang, W.; Remon, D.; Cantarellas, A.M.; Rodriguez, P. A Unified Current Loop Tuning Approach for
Grid-Connected Photovoltaic Inverters. Energies 2016, 9, 723. [CrossRef]
Energies 2018, 11, 1024 15 of 15

21. Espi, J.M.; Garcia-Gil, R.; Castello, J. Capacitive Emulation for LCL-Filtered Grid-Connected Converters.
Energies 2017, 10, 930. [CrossRef]
22. Xu, J.; Xie, S.; Tang, T. Active damping-based control for grid-connected LCL-filtered inverter with injected
grid current feedback only. IEEE Trans. Ind. Electr. 2014, 61, 4756–4758. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).