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.2014.2365516, IEEE Transactions on Power Electronics
1

A Single-Stage PhotoVoltaic System for a Dual-

Inverter fed Open-End Winding Induction Motor
Drive for Pumping Applications
Sachin Jain, Athiesh Kumar T, Ramsha Karampuri, and V. T. Somasekhar, Member IEEE
Department of Electrical Engg., National Institute of Technology, Warangal – 506004 (INDIA)
jsachin@nitw.ac.in, athiesh.44@gmail.com, ramsha_k@nitw.ac.in, vtsomasekhar@rediffmail.com

Abstract — This paper presents an integrated solution for fans, submersible pumps etc. Such loads can have the option of
PhotoVoltaic (PV) fed water-pump drive system, which uses an optimally using PV power whenever Sun power is available
Open-End Winding Induction Motor (OEWIM). The dual- [2].
inverter fed OEWIM drive achieves the functionality of a three- Further, when such loads are used in the stand-alone system
level inverter and requires low value DC bus voltage. This helps like water pumping application in domestic, agricultural and
in an optimal arrangement of PV modules, which could avoid
large strings and helps in improving the PV performance with
industrial sectors, solar PV powered system could be a good
wide band-width of operating voltage. It also reduces the voltage solution. It could meet the requirement during critical situation
rating of the DC-link capacitors and switching devices used in the i.e., during summer especially in tropical countries like India.
system. The proposed control strategy achieves an integration of This encourages the use of electric motor-pump with better
both Maximum Power Point Tracking (MPPT) and V/f control performance and efficiency with the PV system [3].
for the efficient utilization of the PV panels and the motor. The Some possible solutions given for PV fed water-pump
proposed control scheme requires the sensing of PV voltage and were based on usage of DC motor either directly coupled [4]
current only. Thus, the system requires less number of sensors. or via a DC-DC converter [5] with PV source. However, the
All the analytical, simulation and experimental results of this requirement of continuous maintenance and higher cost
work under different environmental conditions are presented in
restricts the use of DC motors for their application in PV water
this paper.
pumping systems [3]. Thus, there is a need of such a solution
Index Terms— Centrifugal Pump, Maximum Power Point that uses the PV power effectively, while using a low cost, low
Tracking, Dual-Inverter, Open-end Winding Induction Motor, maintenance, reliable and robust motor for pumping
Photovoltaic cell. application. The most suitable motor for such an application is
an Induction Motor (IM).
I. INTRODUCTION Some of the initial proposals for a PV fed IM, based on
two-stage system were given by various authors in the past.
E LECTRICAL motors constitute more than 40% of total
electric power consumption [1]. Modernization of human
society and growing applications of electric motors, have
exponentially increased the demand for electrical energy. This
forces an increase in the power generation capacity. However,
due to ecological concerns, restriction and constraints are
imposed on increasing the generation capacity of conventional
sources. So, contemporary research is focused towards an
effective utilization of non-conventional energy sources.
Among the available non-conventional sources PhotoVoltaic
(PV) technology seems to be the most promising and attractive.
This can be attributed to declining cost of PV modules, free
energy source, zero maintenance, and noise free operation.
Thus, employment of PV source for powering electric motor
could be a good solution especially for water-pumps, electric

Fig. 1. Block diagram of PV powered centrifugal pump (a) Conventional

two-stage; (b) Conventional single-stage; (c) Proposed single-stage.

0885-8993 (c) 2013 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.2014.2365516, IEEE Transactions on Power Electronics
Most of them have used a DC-DC boost converter in the first
stage and the second-stage comprises of a DC-AC inverter.
Boost converter amplifies and operates the low value PV input
voltage near Maximum Power Point (MPP) while the inverter
gives the required AC voltage to the IM. Also, the control
techniques are based on either independent frequency control
[2], or a V/f control [6]. Recently, a proposal based on closed
loop speed control to improve the performance and efficiency
of the system was given by Montiê Alves Vitorino et. al. [7], in
which the authors have presented a sensor-less speed control
technique. However, two-stage power conversion, high voltage
rating of semi-conductor devices and more number of sensors
increase the power loss and cost of the system.
A typical PV pumping system with two power conditioning
stage is shown in Fig. 1(a). As discussed earlier, this system
results in poor performance and lesser efficiency. Therefore, a
single-stage system with simpler control as shown in Fig. 1(b),
could be a better choice [8-13]. One of such a system was
proposed by Eduard Muljadi [14], in which the author have
used a six-step quasi square wave inverter, which can take care
of DC-AC inversion as well as Maximum Power Point
Tracking (MPPT) for the PV source. It is a better proposition
since the author has given an integrated single-stage power
conversion solution. But the drawback of this system is that,
the six-step quasi-square inverter deteriorates the motor
performance and hence the efficiency. It also requires filters,
which are bulky and expensive. Further, this system requires a
higher voltage rating for the input DC-link capacitor and semi-
conductor devices. All these may increase the cost, weight, size
and power loss of the system.
Recently, Messaoud Makhlouf et. al. [15] have suggested
the closed loop vector control for PV pumping system using a
single-stage multi-level Pulse Width Modulation (PWM)
converter. While this system results in low motor losses, it is
more complex with more number of control variables and
sensors. Most recently, Tomás Perpétuo Corrêa et. al. [16]
have published a paper with a stand-alone PV pumping system.
It is also a single-stage system, wherein the authors have
proposed MPPT and Minimum Losses Point Tracking (MLPT)
methods of control. However, this system requires more
number of sensors, and results in low band-width for PV
operating voltage.
Thus, from the above discussion, it is evident there is a
requirement for a low-cost, low-voltage single-stage power
conversion PV water pumping system with wide bandwidth of
PV operating voltage. Also, the three-level/multi-level inverter
with a low DC bus voltage could be a better solution as the
performance of this system improves with the increased
number of levels/steps in the inverter output voltage. Also,
many investigations on Open-End Winding Induction Motor
(OEWIM) with multilevel inverters are documented in the
literature [17-21]. Further various PWM techniques and
control schemes [22-27] for OEWIM drive have been analyzed
by various researchers. In short, OEWIM promises to provide
effective solutions for drive application as compared to
common neutral IM [20]. Apart from high reliability and
redundancy [23, 27] OEWIM has many good features as
discussed in the next paragraph.
One of the most important feature of OEWIM is it uses two
two-level Voltage Source Inverters (VSIs) to achieve three-
0885-8993 (c) 2013 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.
2
level inversion [28]. It also lowers voltage rating of
semiconductor devices and input DC bus electrolytic capacitor
[17]. This can be attributed to the fact that DC bus voltage is
designed based on maximum value of phase voltage instead of
line to line voltage. The decoupled PWM technique increases
the apparent switching frequency of switching devices,
reducing the ripple in the motor phase current. Unlike the
conventional three-phase Neutral Point Clamped (NPC) three-
level inverter, wherein the DC-link capacitor is constrained to
carry load current leading to large fluctuations of voltage of the
DC-neutral point, the DC-link capacitor of OEWIM carries
only the ripple current, thus resulting a negligible voltage
fluctuations [23]. This increases the life and reliability of the
DC bus capacitor or in other words the reliability of the
inverter system, which is of paramount interest for systems
connected to PV source.
Thus, OEWIM coupled with PV source could be a good
proposition. This paper presents one such solution for a simple
water-pump application as depicted in Fig. 1(c). The proposed
system has the following features:
1. It is an economical system as it uses a single power
conditioning stage.
2. It has inherent low DC bus voltage requirement with V/f
control integrated with MPPT and uses three-level (DC-
AC) inversion for better performance of motor.
3. Low input DC bus voltage requirement reduces the voltage
rating of DC-link capacitor and increases bandwidth of PV
operating voltage. It also reduces the voltage rating of the
semi-conductor devices used in the inverter.
4. It optimally uses PV source for all environmental
conditions by operating at MPP. Also, it employs V/f
control integrated in MPPT algorithm which improves the
performance of the motor and requires less number of
sensors for its operation.
5. Three-level output with Decoupled Sample-Averaged Zero-
sequence Elimination (DSAZE) algorithm [24] further
improves the performance with reduced motor current
ripple.
Rest of the manuscript is divided into four sections. Second
section gives details of modeling of the proposed system. Third
section describes the operation and analysis of the proposed
system. Fourth section describes the control strategy and
algorithm proposed. Fifth section depicts the simulation and
experimental results obtained. It also gives the detailed cost
analysis and comparison for the proposed system with the
existing system(s).
II. MODELING OF THE PROPOSED SYSTEM
A proposed configuration of the solar PV powered pumping
system is shown in Fig. 2, which comprises of 1) solar PV
array; 2) dual-inverter namely inverter-I and inverter-II; 3)
three-phase open-end winding induction motor with pump load;
4) controller block which consists of MPPT and DSAZE PWM
algorithm. These components are described in detail in the
following sub-sections.
A. PV source Model
The PV source was modeled by using PV cell current-
voltage characteristic equation as furnished below [29]:
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.2014.2365516, IEEE Transactions on Power Electronics
3

V pv ⎡ 2 1 ⎤
v aa ' = ( S A − S A ' ) − (( S B − S B ' ) + ( S C − S C ' ) )⎥ (5)
2 ⎢⎣ 3 3 ⎦
Similarly the other phase voltages vbb’, vcc’ of the inverter
output can be derived for the system.
Further the input inverter current iinv can also be derived
using switching functions.
Current flowing through inverter-I is given by
1 1 1
iinv 1 = ( S A + 1)i aa ' + ( S B + 1)ibb ' + ( S C + 1) icc ' (6)
2 2 2
Current flowing through the inverter-II is given by
1 1 1
i inv 2 = ( S A ' + 1)( − i aa ' ) + ( S B ' + 1)( − i bb ' ) + ( S C ' + 1)( − i cc ' ) (7)
2 2 2
The net current flowing through the dual-inverter is
1 1 1
iinv = iaa ' ( S A − S A ' ) + ibb ' ( S B − S B ' ) + icc ' ( S C − S C ' ) (8)
Fig. 2. Schematic circuit diagram of proposed system. 2 2 2
Thus the above values of phase voltage and current can be
⎛ ⎞ (
q v pvcell + i pvcell R s ) used by Simulink model of OEWIM, which is discussed in the
i pvcell = i L − i0 ⎜ e nkT
− 1⎟ (1) next section.
⎜ ⎟
⎝ ⎠
where ipvcell : PV cell current; iL : photo-current; i0 : Diode C. Open-End Winding Induction Motor model
saturation current; n : Diode quality or ideality factor; k : An Open-End Winding Induction Motor [28] is obtained by
Boltzmann constant; q: electron charge; T : panel operating opening the neutral point of the star connected stator windings
temperature in Kelvin; Rs : PV cell series resistance; vpvcell : PV of a normal three-phase Induction motor. The winding diagram
cell voltage (V). of three-phase OEWIM is shown in Fig. 3.
The output of PV source is connected to inverter with DC For modeling and analysis the decoupled form of OEWIM
bus capacitance Cpv. By applying KCL at input of inverter is considered. For transforming the stator (ϕ=θ) and the rotor
(from Fig. 2) parameters (ϕ=β) to decoupled form, the transformation matrix
dv pv used is given in (9).
i pv = ic + iinv ⇒ i pv = C pv + iinv (2)
dt ⎡ cos φ cos(φ − 2π ) cos(φ + 2π ) ⎤
⎡ x qy ⎤ ⎢ 3 3 ⎥ ⎡ x aa ' ⎤
Integral solution of (2) is the voltage vpv across capacitance ⎢ ⎥ 2⎢
⎢ ⎥ 3 ⎢ sin φ
x = sin(φ − 2 π ) sin(φ + 2 π )⎥ ⎢x ⎥ (9)
Cpv, which is used by the PV model to calculate the PV source dy 3 3 ⎥ ⎢ bb ' ⎥
current. The inverter current iinv is the current drawn by ⎢ x 0y ⎥ ⎢ 1 ⎥ ⎢⎣ x cc ' ⎥⎦
⎣ ⎦ 1 1
⎢⎣ 2 2 2 ⎥⎦
inverters I and II. Further dual-inverter has two series
connected equal value capacitors across the DC-link. These where, θ is the angle between the stator as-axis and the
capacitors share equal voltage with respect to the common quadrature (q) axis, β is the angle between rotor ar-axis and the
point ‘o’ as shown in Fig. 2. q-axis, also β = θ − θ r , θr is the angle between rotor ar-axis and
stator as-axis (Fig. 3), parameter x can be either voltage ‘v’ or
B. Modular three-level dual-inverter model:
current ‘i’ or flux linkage ‘λ’ and subscript parameter y can be
Dual-inverter used in the proposed configuration is ‘s’ or ‘r’. The subscript ‘s’ denotes the parameters of stator and
modeled using switching functions [30]. To model dual- the subscript ‘r’ denotes the parameters of rotor.
inverter, switching function SW (where W can be A, B, C, A', B' The dynamic d-q model of an OEWIM is described by in
or C' depending on the phase and number of inverter) requires (10) to (13).
the logic generated from PWM controller. It has value 1 and
-1 which represents turn ON of top and bottom switch vqs = Rs iqs + ω ( Lls ids + Lm (ids + idr' )) + p λ qs (10)
respectively for the given leg or phase of the inverter. The
vds = Rs ids − ω ( Lls iqs + Lm (iqs + iqr' )) + p λds (11)
modular dual-inverter shown in Fig. 2 consists of six poles (a,
b, c, a', b', and c') and twelve switches (two switches per vqr' = Rr' iqr' + (ω − ω r )( L'lr idr' + Lm (ids + idr' )) + p λ'qr (12)
pole). The value of the pole voltages in a particular phase can
be ±Vpv/2 depending on the switch (whether top or bottom) is v dr' = Rr' idr' − (ω − ω r )( L'lr iqr' + Lm (iqs + iqr' )) + p λ'dr (13)
turned ON. If top switch of phase ‘a’, S1 is turned ON, the
pole voltage vao is +Vpv/2 and when bottom switch of phase where, Rr – rotor resistance, Rs – stator resistance, Lls – stator
‘a’, S4 is turned ON then the pole voltage vao is –Vpv/2. Thus, leakage inductance, L'lr – rotor leakage inductance, Lm –
pole voltage of inverter-I can be given as mutual inductance between stator and rotor winding, ω is the
V pv V pv V pv synchronous speed, ωr is the electrical speed of motor, p
v ao = S A ; v bo = S B ; v co = S C (3)
2 2 2 denotes the time derivative. Also, here vqr' = vdr' = 0 , since rotor
Similarly, pole voltage of inverter-II can be given as bars are short-circuited.
V pv V pv V pv The expression for the electromagnetic torque Tem is given by
v a 'o = S A ' ; v b 'o = S B ' ; v c 'o = S C ' (4)
2 2 2 3 P
Tem = L m [ iqs idr' − ids iqr' ] (14)
The motor phase voltage vaa’ is given by 2 2

0885-8993 (c) 2013 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.2014.2365516, IEEE Transactions on Power Electronics
4

bs-axis
b’s
Air gap
bs Stator
q-axis
br-axis
Rotor
ar-axis
β
θ
as as-axis
θr
a’s

c’s
cs
cr-axis d-axis
cs-axis

Fig. 3. Open-end winding induction motor model.

where, P is the number of poles.
The mechanical equation governing the OEWIM-pump
drive is expressed in (15) as
d ω rmech
Tem = J + B ω rmech
2
+ TL (15)
dt
where, J – motor inertia (kg-m2), B – centrifugal load torque
coefficient, TL – load torque (N-m), ωrmech is instantaneous
angular velocity of motor shaft (rad/sec).
III. OPERATION AND ANALYSIS OF THE PROPOSED SYSTEM
The proposed dual-inverter is operated by using the
decoupled PWM strategy. It incorporates simple V/f control
for the efficient operation of system below the rated speed.
The proposed PWM strategy requires the information of
magnitude and angle of the reference voltage vector. The
magnitude is calculated and controlled by the MPPT algorithm
and the angle ‘α’ is the function of time and fundamental
frequency of reference modulating waveform. The reference
voltage vector |vsr|∠α is further divided into two decoupled
components |vsr|/2∠ α and |vsr|/2∠(180+ α). The decoupled
components are then given as the reference vector for inverter-
I and inverter-II respectively as shown in Fig. 2, for generation
of required output voltage. Thus, using decoupled PWM
configuration has the benefit of double output voltage.
Low input DC bus voltage requirement of dual-inverter for
OEWIM-pump drive:
To analytically verify the low input DC bus voltage
requirement, a comparison between two-level and dual-
inverter fed OEWIM is done. Both the inverters are compared
for generating same output voltage vector with different
values of input PV source voltage. The low input voltage
requirement of dual-inverter for an OEWIM drive is
demonstrated in Fig. 4.
Let the PV source voltage required to generate the rated
instantaneous induction motor phase voltage, van is Vpv as
shown in Fig. 4(a). From Fig. 4(a) and 4(b) the inverter output
voltage, van, vbn and vcn can be obtained using the inverter pole
voltage, vao, vbo and vco; and switching functions SA, SB and SC
as follows:
2 1 V The solar PV powered fed dual-inverter connected to
v an = v ao − v no = ( S A − ( S B + S C )) pv
3 3 2
where vno is common-mode voltage given by
Fig. 4. Demonstration and comparison of DC bus voltage requirement for H-
bridge and dual-inverter systems. (a) Schematic circuit diagram of two-level
H-bridge inverter with input DC voltage (PV voltage) of ‘Vpv’ ; (b) Space
vector locations of voltage vector obtained from two-level inverter; (c)
Schematic circuit diagram of dual-inverter fed OEWIM drive with input DC
voltage (PV voltage) of ‘Vpv/2’; (d) Space vector locations of voltage vector
obtained from three-level dual-inverter scheme.
1 1 V
v no = ( vao + vbo + vco ) = ( S A + S B + S C ) pv (17)
3 3 2
Thus, the space vector location of reference voltage vector OA
(Fig. 4(b)) can be generated by the switching functions SA = 1,
SB = -1 and SC = -1. Substituting these values in (17) will
result in the phase voltage, van as 2Vpv/3.
Now, consider the 3-phase, three-level dual-inverter
connected to an OEWIM as shown in Fig. 4(c). Let the input
PV source voltage is Vpv/2, which is half of the voltage taken
for two-level H-bridge inverter. Now, the dual-inverter output
phase voltage OG (Fig. 4(d)) is given as
v aa ' = v ao − v a ' o ' − v oo ' (18)
where voo' is the common mode voltage (Fig. 4(c)) which is
given by
1 V pv
voo ' = [( S A − S A ' ) + ( S B − S B ' ) + ( S C − S C ' )] (19)
3 4
Therefore vaa’ is given by
⎛2 1 ⎞V
v aa ' = ⎜ (S A − S A ' ) − [( S B − S B ' ) + ( S C − S C ' )] ⎟ pv (20)
⎝3 3 ⎠ 4
So, to generate voltage vector OG (Fig. 4(d)), the
switching functions required are SA = 1, SB = -1 and SC = -1 ;
SA’ = -1, SB’ = 1 and SC’ = 1. Substituting these values in (20),
results in the phase voltage of magnitude 2Vpv/3 corresponding
to phase aa'.
Hence, to generate the phase voltage of 2Vpv/3 the PV
source voltage required in case of two-level inverter is Vpv and
in case of dual-inverter connected to OEWIM is Vpv/2.
However, DSAZE PWM technique [22] needs excess 15% of
DC-link voltage to generate the rated motor phase voltage.
IV. CONTROL STRATEGY AND MPPT ALGORITHM
(16) OEWIM-pump drive is operated using a simple control
strategy, which simultaneous accomplishes MPPT and
DSAZE PWM integrated together. This integrated algorithm
generates the required PWM control signals for the modular

0885-8993 (c) 2013 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.2014.2365516, IEEE Transactions on Power Electronics
Start
Read vpv , ipv
Initialize ma, vpv,prev, ppv,prev
Calculate ppv and carry vpv
Calculate Δppv = ppv – ppv,prev and
Δvpv = vpv – vpv,prev
Yes No
If Δppv / Δvpv > 0
ma
Decrease ma by Δma Increase ma by Δma
Calculate |vsr|∟α
Calculate vαs, vβs
Transform vαs, vβs into vjs where j = a, b, and c
Using SVPWM algorithm
Calculate Tjs where j = a, b, and c
Extract Tmax, Tmin and hence calculate Teff and Tzero
Using DSAZE PWM algorithm calculate Toffset = Ts/2
Yes No
If Tzero < 0
Update Tjs where j = a, b, and c; also
update Toffset
Calculate Tgj = Tjs + Toffset where j
= a, b, and c
Generate firing pulses for the switches in
inverter-I and inverter-II
Fig. 5. Flow chart of the MPPT, DSAZE algorithm used in the proposed
scheme.
dual-inverter. The flowchart of control algorithm is shown in
Fig. 5. The MPPT part of algorithm facilitates motor-pump
drive to extract maximum available power from the PV
source, thereby assuring effective utilization of the PV source.
The DSAZE PWM part of the algorithm incorporates V/f
control. It maintains constant rated flux in the motor which
retains the maximum torque capability of the machine for the
given PV power. Thus, in pump drive application where
torque is proportional to square of speed, maintaining the
maximum torque further helps in maximum utilization of the
input source.
A. MPPT algorithm:
One of the simplest, most popular and commercially used
methods of MPPT, namely, the Hill Climbing algorithm [31]
is employed in the proposed system. The algorithm first senses
the voltage (vpv) and current (ipv) of PV array for calculating
the power (ppv). The calculated and sensed PV power and
voltage are then compared with their previous value to
determine the slope of the operating point. The considered
slope then determines the correction for modulation index
0885-8993 (c) 2013 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.
5
(ma). With respect to sign of the slope (negative or positive),
the value of ma is modified (incremented or decremented) as
described in the flowchart shown in Fig. 5 until the operating
point reaches near MPP. The calculated value of ma is then
used by the DSAZE PWM algorithm.
B. DSAZE PWM algorithm:
The magnitude of the reference voltage vector generated
ppv,prev = ppv and
vpv,prev = vpv
by the MPPT output and angle ‘α’ is decomposed into the
instantaneous three-phase reference voltage vas, vbs and vcs for
inverter-I. The gating time Tgj (j=a, b, c) for top switches of
inverter-I (Fig. 2) is then obtained by the switching algorithm
[32]. This algorithm [32] requires instantaneous values of the
MPPT
reference phase voltage for calculating the effective time (time
for which all the active vectors are switched) or turned ON
DSAZE PWM time for top switches. The position of effective time period
can be adjusted in such a way that the offset time is equal to
Ts/2 within a switching period. This feature is exploited by
DSAZE PWM technique to eliminate the zero sequence
voltage within a sampling period.
As DSAZE is a center spaced PWM technique for the
dual-inverter system, the ripple content in current is less and
hence results in the improvement of developed torque. The
other advantage of the DSAZE PWM algorithm for OEWIM
drive is that, it requires less memory and computing time for
processing. As both the reference voltage vectors for inverter-I
and inverter-II are in the phase opposition, the gating time Tgj'
(j=a, b, c) for the top switches of inverter-II are directly
obtained using the gating time for inverter-I as:
T gj ' = Ts − Tgj (21)
where, j=a, b, c and Ts is the inverter switching time
period(s).
Complement of the respective gating signals for inverter-I
and inverter-II are generated for bottom switches for both
inverters.
V. SIMULATION AND EXPERIMENTAL RESULTS WITH COST
ANALYSIS
A. Simulation results:
The single-stage PV powered OEWIM drive for pump
application, shown in Fig. 2 is simulated using
MATLAB/Simulink. A 3.6 kW (20x3) PV array feeding
power to a 4 kW OEWIM-pump drive is considered for
simulation. The Solarex MSX60 PV panel parameters at
Standard Test Conditions (STC) are Voc=21V, Isc=3.74A,
Vm=17.1V, Im=3.5A, Pm=59.9W. A DC bus capacitor (Cpv) of
value 1000 μF is used at the output of PV source. A 4-pole,
400V, 1430rpm induction motor is used. Further the important
parameters of the motor are Rs=1.405Ω, Rr'=1.395Ω,
xls=xlr'=1.8344Ω, xm=54.0982Ω. Simulations are performed by
considering 96 samples per cycle of applied fundamental
voltage, irrespective of the modulation index. This means that
the switching frequency is a variable quantity. The switching
frequency varies from 1.28 kHz (corresponding to a
modulation index, ma = 0.2 with fundamental frequency f =
13.33Hz) to 4.8 kHz (corresponding to a modulation index, ma
>= 0.75 with fundamental frequency f = 50Hz).
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.2014.2365516, IEEE Transactions on Power Electronics
6

Fig. 6 shows the simulation results of the proposed system 500

at different environmental conditions for PV source side 0 Motor phase voltage, vaa' (V)
parameters. The increasing and decreasing nature of PV
-500
power with respect to insolation (G) and temperature (T) can
10
be observed with the waveforms of PV power and ma. This
0 Motor phase current, iaa' (A)
verifies the MPP tracking, further small oscillations in the -10
value of ma near MPP and small ripple content in a PV power
30
confirms the operation near optimum voltage. Also, another
20
useful observation in the simulation results is that, the 10 Torque, Tem (N-m)
operating voltage of PV array passes through optimum 0
voltage for every step increase in insolation and temperature. 1500
1000 Speed, (rpm)
This can be justified with the matching values of peak power
500 0.33xMechanical power output, (W)
value during transient tracking and steady-state near MPP as 0
given in the PV power subplot. Further, it can be noted that
PV voltage waveform shows a sudden rise and fall in the 0.05
Slip
value with the step increase in insolation and temperature. 0
This can be attributed to charging and discharging of PV 200
capacitor CPV with excess or deficit PV power respectively 100 Slip power (W)
during transient condition.
0
Fig. 7 shows the motor side parameters for different 100
environmental conditions. Variation on ma with respect to PV
50 Efficiency (%)
power can also be seen here in the waveform of torque, speed
20
and mechanical power output. It can be observed that torque, 0 5 10 15 20 25 30
Time (s)
speed and mechanical power output follows ma or the PV
Fig. 7. Simulation results showing waveforms at motor-pump side.
power indirectly. Also, it can be easily observed that slip
power too follows ma. Thus, slip power loss increases and improves with increase in PV power. This can be observed in
decreases with increase and decrease of PV power Table I, where the simulation results are summarized.
respectively. This may help in keeping small variation in
efficiency which can be observed from last subplot of Fig. 7. B. Experimental results:
Also, another important feature that relate results of Fig. 6 and To verify simulation results a prototype of the proposed
Fig. 7 is motor phase voltage plot. The transient variations in system is developed. To emulate PV characteristics, a DC
PV voltage Vpv during step increase in insolation can be power supply with fixed series resistance is used. The power
depicted with a peak value of motor phase voltage waveform. supply is operated in constant voltage mode with variable
Another typical observation between Figs. 7 and 8 is the current limit. The 3-phase OEWIM with specifications 440V,
ripple contents in the torque waveform and motor phase 1.5A, 1440 rpm and 1hp is used with fan type of centrifugal
current follows ma. It can be observed that torque ripple is load which guarantee the operation for water-pump load
more during lower insolation and decreases as insolation profile. The motor is fed from the dual-inverter which uses
increases. This can be attributed to the fact that the increase Semikron IGBT module (SKM75GB128DN). Gate pulses
and decrease of ripple content in the motor phase current will required by IGBT module is generated from dSPACE 2201
results in the increase and decrease of the torque ripple and is given via a driver IC (HCPL 316J). The control
respectively. Further it can be noted that the amplitude of algorithm implemented in dSPACE includes the MPPT and
motor phase current also follows ma and Total Harmonic V/f control with DSAZE PWM for the OEWIM coupled to a
Distortion (THD) of motor phase current decreases with centrifugal torque load. This control algorithm can also be
increase in insolation. Thus in short, system performance implemented using a low-cost Freescale semiconductor made
MC56F8013 signal controller. For tracking MPP of the
2
0.033xTemperature (ºC) source, a voltage sensor (LV25-P, 600V rms, 10mA/25mA) to
1 sense the terminal voltage and a current sensor (Tektronix
Insolation (Suns)
0 A622, 0-70A rms, 100mV/A) to sense the PV current are used.
15 The sensed voltage and current are given to the computing
10 0.04xPV voltage, vpv (V) system through the ADC’s of dSPACE. The computing
5 PV current, ipv (A)
0
system internally generates the value of ma required for the
4000 DSAZE PWM algorithm to calculate Tgj and Tgj' for inverter I
2000
Matching PV power
PV power, ppv (W) and II respectively (see Fig. 5).
Fig. 9 gives the waveform for Tga and Tga' obtained from
0
simulation and experimental set-up. The complementary
0.5 Oscillations near MPP nature of Tga and Tga' can be observed in the waveforms. This
Modulation index, ma
verifies the utilization of decoupled PWM strategy. Further, as
0
0 5 10 15 20 25 30 proved decoupled PWM also reduces the required value of DC
Time (s) bus voltage which can be depicted from Fig. 10. In this figure,
Fig. 6. Simulation results showing waveforms at PV source side.
the plots of phase voltage vector in α-β plane obtained from

0885-8993 (c) 2013 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.2014.2365516, IEEE Transactions on Power Electronics
7

10
Phase "a" [A]
5
0
-5
-10
3 3.05 3.1 3.15 3.2 3.25
Time (s)
6
Current spectrum

5
THD = 11.94% Fig. 9. DSAZE PWM modulating waveforms for leg a & a' of Inverter –I
(A peak)

4
3 and Inverter-II (a) from simulation (Vpv≈320V and ma=0.64) (b) from
2 experimental setup (Vpv≈110V and ma=0.5) (x-axis: simulation time (s); y-
1 axis: gate time (s)).
0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Frequency (Hz)
120
(a) 400
15 80
Phase "a" [A]

10 200 Change in ma
after completion 40
5
0 0 every cycle 0
-5
-40
-10 -200
-15 -80
13 13.01 13.02 13.03 13.04 13.05 13.06 13.07 13.08 -400
Time (s) -120
10
Current spectrum

-400 -200 0 200 400 -120 -80 -40 0 40 80 120

8 vαs (V) vαs (V)
THD = 5.06%
(A peak)

6 (a) (b)
4 Fig. 10. α-β plot of 3-phase vo1tage input to OEWIM obtained from (a)
2 simulation (Vpv≈320V and ma=0.64) (b) experimental setup (Vpv≈110V and
0 ma=0.5).
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Frequency (Hz)
(b) result in a poor performance of the system. This can be taken
15
10
care of by assuring proper synchronization between MPPT
Phase "a" [A]

5 and V/f control under all field conditions. The other drawback
0
-5
in the field may be the inverter failure. If an inverter fails, a
-10 rewiring would enable the healthy inverter to deliver partial
-15
18 18.01 18.02 18.03 18.04 18.05 18.06 load. In other words, the fault tolerance with OEWIM is more
Time (s)
12
compared to the conventional two-level VSI drive. Also, the
Current spectrum

10 power cable required to connect the OEWIM and inverter is

(A peak)

8 THD = 3.77%
6
double compared to the conventional two-level VSI drive.
4 However, this drawback in the field can be taken care by
2
0
minimizing the distance between inverter and the motor.
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Frequency (Hz) Fig. 11 gives the experimental results for the parameters at
(c) the input side of inverter. Increasing value of modulation
Fig. 8. Harmonic spectrum of motor phase current (iaa') at steady state index and input power confirms the MPP tracking which can
obtained from simulation at different environmental conditions; (a) at 0.1
insolation and 25ºC; (b) at 0.6 insolation and 35ºC; (c) at 1.0 insolation and be seen in Fig. 11(a). Oscillations near MPP can be observed
55ºC. in Fig. 11(b). More number of oscillations in ma near MPP can
be attributed to sharp power-voltage characteristics of input
TABLE I source and the algorithm used. However, operation near MPP
SUMMARY OF SIMULATION RESULTS FOR THE PROPOSED SYSTEM.
can be observed in the waveform of source power. Also, to
PV O/P Slip Current
G T
Power Power
η Speed Torque Slip
Power THD emulate various environmental conditions in PV system the
o
Suns ( C) (%) (rpm) (N-m) (%) current limit of programmable power supply is changed from
(W) (W) (W) (%)
0.1 25 329 240 72.95 352 6.5 4.74 12 11.94 0.3A to 0.5A and again back to 0.3A. This would nearly
emulate the increasing and decreasing insolation in PV array
0.3 30 1040 895 86.06 740 11.6 3.62 33.5 8.58 system. Fig. 11(c) shows experimental results obtained from
0.6 35 2133 1902 89.17 1026 17.7 3.8 75.25 5.06 developed test set-up for this condition. The increasing and
0.8 45 2763 2465 89.21 1139 20.65 4.24 109.25 4.28 decreasing value of ma and source power shows MPP tracking
0.9 50 3072 2735 89.03 1187 22 4.46 128 3.97 for various environmental conditions. As fan type load has
1 55 3365 2990 88.86 1229 23.25 4.7 147.5 3.77 approximately similar characteristics to centrifugal water
pump load so nearly same performance is accepted with actual
simulation and experimental results were shown. Requirement water pump load systems.
of low DC bus voltage can be easily verified with the given Fig. 12 shows the experimental results for the parameters
input values of ma and input PV voltage of the inverter as at the inverter output. Implementation of V/f control can be
mentioned in Fig. 10(a) and (b). The synchronization of verified by variable frequency operation at starting and
change in ma after one complete cycle can be depicted from running condition. At starting when the value of ma is low, the
zoom-in part of Fig. 10(a). If this synchronization between operating frequency is kept low and near MPP when ma value
MPPT and V/f control is disturbed in the field, then it may becomes high frequency of phase voltage also increases. The

0885-8993 (c) 2013 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.2014.2365516, IEEE Transactions on Power Electronics
8

Source voltage (V)

Source voltage (V) Source voltage (V)
Source current (A)
Source current (A)
Source current (A)

Source power (W)

This Source power (W) This
plot is plot is Source power (W) This plot
captured captured is
from the from the captured
Modulation index control control from the
desk desk control
layout Modulation index layout desk
of of layout of
dSPAC dSPACE
dSPAC
E 2201
Modulation index
E 2201 2201

(a) (b) (c)

Fig. 11. Experimental results for the proposed system on the DC source side i.e., source voltage, current, power and the modulation index (a) at starting
condition (b) at running condition and (c) for various emulated environmental conditions.

TABLE II
motor ‘aa'’ phase current waveform along with its harmonic
COST COMPARISON BETWEEN THE PROPOSED SYSTEM WITH TWO BASIC
spectrum at starting and running conditions are shown in Fig. TOPOLOGIES AVAILABLE IN THE LITERATURE FOR PUMP APPLICATION
13(a) and (b) respectively. The harmonic spectrum at starting Using conventional induction Using
and running conditions is comparative with the harmonic motor OEWIM
spectrum of current at low and high insolation (see Fig. 8) Item / part no. 1-stage 3-
2-stage 2-level 1-stage 2-
level
respectively. Another important thing is the typical pattern of [3] level [14]
[proposed]
ripple in the current waveform shown in Fig. 13 is matching PWMs PWMs PWMs
with the simulation results given in Fig. 8. 4 3 6
Controller ADCs ADCs ADCs
C. Cost Analysis: requirements 3 2 2
A brief cost analysis has been carried out (Table II) for the Flash / RAM Flash / RAM Flash / RAM
Medium Less Less
three topologies presented in Fig. 1. The cost values presented
MC56F8013 MC56F8013 MC56F8013
in the Table II are only for the inverter and converter. The Controller
$4 $4 $4
induction motor-pump and the PV panel costs are assumed to Voltage Sensors/ 2 1 1
be nearly similar (since the same power rating is considered) LV 25-P $ 120 $ 60 $ 60
for all the three topologies. When compared to a two-stage Current Sensors/ 1 1 1
LTSR 15-NP $ 23.68 $ 23.68 $ 23.68
system, the proposed one is a low cost because of the non-
Miscellaneous $ 20 $ 10 $ 20
requirement of an extra DC-DC converter. Also, the lower Sub-total $167.68 $97.68 $107.68
cost of the proposed system compared to the conventional INVERTER
single-stage two-level is because of the requisition of low- DC- link 4 4 2
voltage DC link capacitors and switching devices. capacitors
(2200uH, 400V)/
$ 155.84 $ 155.84 $ 77.92
ALS30A222MF4
VI. CONCLUSION 00
In this paper, an integrated single-stage solution of PV fed 6 (900V, 12 (400V,
6 (900V, 28A)/
Semiconductor 28A)/ 20A)/
pump drive is presented. The proposed system has the feature Switches*
IRG4PF50WPB
IRG4PF50WP IRGB14C40L
of low DC bus voltage requirement, MPPT integrated with V/f, F
(IGBT’s) BF PBF
$ 31.68 $ 31.68 $ 39.36
Driver circuit $ 26 $ 26 $ 52
Sub-total $213.52 $213.52 $169.28
FRONT-END CONVERTER
Inductor/ 1 NA NA
DLFL-0147-12D5 $ 25.6 --- ---
Capacitor/ 3 NA NA
ALS30A222MF4
$116.88 --- ---
00
Fig. 12. Experimental results for the proposed system at motor-pump side Semiconductor 1 NA NA
i.e., Motor phase voltage and current (a) at starting condition (b) at running switch*/
condition. IRG4PH40KDPB $ 5.44 --- ---
F
Motor phase Diode/ 1 NA NA
Motor phase
current (A) D22-25-12 $ 6.16 --- ---
current (A)
Driver circuit $5 NA NA
Sub-total $ 159.08 --- ---
Harmonic Harmonic TOTAL $540.28 $311.20 $276.96
spectrum spectrum * Switches are chosen with respect to the availability of the devices with nearest rating.

(a) (b)
Fig. 13. Motor aa' - phase current and its harmonic spectrum (a) at starting three-level inverter operation and low cost. Analytical proof
condition and (b) at running condition. for low DC bus voltage requirement in the proposed system is

0885-8993 (c) 2013 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.2014.2365516, IEEE Transactions on Power Electronics
presented. Implementation of V/f with MPPT can be verified
with simulation and experimental results. Thus, a high-
performance integrated solution is proposed. Low cost of the
proposed system can be attributed to the requirement of low
voltage DC-link bus capacitor, low voltage rating switches and
less number of sensors for the integrated control operation
which is presented in Table II. Thus in all, this paper presents
one of the effective and simple solutions for PV power fed
water-pump application.
REFERENCES
[1] D. Maheswaran, K. K. Jembu Kailas, V. Rangaraj, W. Adithya Kumar,
“Energy Efficiency in Electrical Systems”, Proc. of IEEE Inter. Conf. on
Power Elect., Drives and Energy sys., pp 1-6, Dec. 2012.
[2] Y. Yao, P. Bustamante, R. S. Ramshaw, “Improvement of induction
motor drive systems supplied by photovoltaic arrays with frequency
control”, IEEE Trans. Energy Conv., vol. 9, no. 2, pp. 256-262, Jun.
1994.
[3] João Victor Mapurunga Caracas, Guilherme de Carvalho Farias, Luis
Felipe Moreira Teixeira, Luiz Antonio de Souza Ribeiro,
“Implementation of a high-efficiency, high-lifetime and low-cost
converter for an autonomous photovoltaic water pumping system”, IEEE
Trans. Ind. App., vol. 50, no. 1, pp. 631-641, Jan./Feb. 2014.
[4] J. Appelbaum, “Starting and steady-state characteristics of dc-motors
powered by solar cell generators”, IEEE Trans. Energy Conv., vol. EC-
1, no. 1, pp. 17-25, Mar. 1986.
[5] S. M. Alghuwainem, “Steady-state performance of DC motors supplied
from photovoltaic generators with step-up converter”, IEEE Trans.
Energy Conv., vol. 7, no. 2, pp. 267-272, Jun. 1992.
[6] S. R. Bhat, Andre Pittet, B. S. Sonde, “Performance optimization of
induction motor-pump system using photovoltaic energy source”, IEEE
Trans. Ind. App., vol. IA-23, no. 6, pp. 995-1000, Nov./Dec. 1987.
[7] Montiê Alves Vitorino, Maurício Beltrão de Rossiter Corrêa, “An
effective induction motor control for photovoltaic pumping”, IEEE
Trans. Ind. Elect., vol. 58, no. 4, pp. 1162-1170, Apr. 2011.
[8] Sachin Jain, Vivek Agarwal, “A single-stage grid-connected inverter
topology for solar PV systems with maximum power point tracking”,
IEEE Trans. Power Elect., vol. 22, no. 5, pp. 1928-1940, Sep. 2007.
[9] Luiz Henrique S. C.Barreto, Paulo Peixoto Prac¸a, Demercil S. Oliveira
Jr., and Ranoyca N. A. L. Silva “High-voltage gain boost converter
based on three-state commutation cell for battery charging using PV
panels in a single conversion stage”, IEEE Trans. Power Elect., vol. 29,
no. 1, pp. 150-158, Jan. 2014.
[10] Mario A. Herrán, Jonatan R. Fischer, Sergio Alejandro González,
Marcos G. Judewicz and Daniel O. Carrica, “Adaptive dead-time
compensation for grid-connected PWM inverters of single-stage PV
systems”, IEEE Trans. Power Elect., vol. 28, no. 6, pp. 2816-2825, Jun.
2013.
[11] Bader N. Alajmi, Khaled H. Ahmed, Grain Philip Adam, and Barry W.
Williams “Single-phase single-stage transformer less grid-connected PV
system”, IEEE Trans. Power Elect., vol. 28, no. 6, pp. 2664-2676, Jun.
2013.
[12] Yufei Zhou, and Wenxin Huang “Single-stage boost inverter with
coupled inductor”, IEEE Trans. Power Elect., vol. 27, no. 4, pp. 1885-
1893, Apr. 2012.
[13] Saeed Danyali, Sayed Hossein Hosseini, and Gevorg B. Gharehpetian
“New extendable single-stage multi-input dc-dc/ac boost converter”,
IEEE Trans. Power Elect., vol. 29, no. 2, pp. 775-788, Feb. 2014.
[14] Eduard Muljadi, “PV water pumping with a peak-power tracker using a
simple six-step square-wave inverter”, IEEE Trans. Ind. App., vol. 33,
no. 3, pp. 714-721, May/Jun. 1997.
[15] Messaoud Makhlouf, Feyrouz Messai, Hocine Benalla, “Vectorial
command of induction motor pumping system supplied by a
photovoltaic generator”, J. Elect. Engg., vol. 62, no. 1, pp. 3-10, 2011.
[16] Tomás Perpétuo Corrêa, Seleme Isaac Seleme Jr., Selênio Rocha Silva,
“Efficiency optimization in stand-alone photovoltaic pumping system”,
J. Renew. Energy, 41, pp. 220-226, 2012.
[17] Anandarup Das and K. Gopakumar, “A voltage space vector diagram
formed by six concentric dodecagons for induction motor drives”, IEEE
Trans. Power Elect., vol. 25, no. 6, pp. 1480-1487, Jun. 2010.
0885-8993 (c) 2013 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.
9
[18] B. Venugopal Reddy, V. T. Somasekhar, Yenduri Kalyan, “Decoupled
space-vector PWM strategies for a four-level asymmetrical open-end
winding induction motor drive with waveform symmetries”, IEEE
Trans. Ind. Elect., vol. 58, no. 11, pp. 5130-5141, Nov. 2011.
[19] Jaison Mathew, P. P. Rajeevan, K. Mathew, Najath Abdul Azeez and K.
Gopakumar, “A multilevel inverter scheme with dodecagonal voltage
space vectors based on flying capacitor topology for induction motor
drives”, IEEE Trans. Power Elect., vol. 28, no. 1, pp. 516-525, Jan.
2013.
[20] Jaison Mathew, K. Mathew, Najath Abdul Azeez, P. P. Rajeevan and K.
Gopakumar, “A hybrid multilevel inverter system based on dodecagonal
space vectors for medium voltage IM drives”, IEEE Trans. Power
Elect., vol. 28, no. 8, pp. 3723-3732, Aug. 2013.
[21] V. T. Somasekhar, B. Venugopal Reddy, K. Sivakumar, “A four-level
inversion scheme for a six-n-pole open-end winding induction motor
drive for an improved DC-link utilization”, IEEE Trans. Ind. Elect., vol.
61, no. 9, pp. 4565-4572, Sep. 2014.
[22] M. R. Baiju, K. K. Mohapatra, R. S. Kanchan and K. Gopakumar, “A
dual two-level inverter scheme with common mode voltage elimination
for an induction motor drive”, IEEE Trans. Power Elect., vol. 19, no.3 ,
pp. 794-805, May 2004.
[23] R. S. Kanchan, P. N. Tekwani and K. Gopakumar, “Three-level inverter
scheme with common mode voltage elimination and DC link capacitor
voltage balancing for an open-end winding induction motor drive”,
IEEE Trans. Power Elect., vol. 21, no. 6, pp. 1676-1683, Nov. 2006.
[24] V. T. Somasekhar, S. Srinivas, K. K. Kumar, “Effect of zero-vector
placement in a dual inverter fed open-end winding induction motor drive
with a decoupled space vector PWM strategy”, IEEE Trans. Ind. Elect.,
vol. 55, no. 6, pp. 2497-2505, Jun. 2008.
[25] Chintan Patel, P. P. Rajeevan, Anubrata Dey, Rijil Ramchand, K.
Gopakumar and Marian P. Kazmierkowski, “Fast direct torque control
of an open-end induction motor drive using 12-sided polygonal voltage
space vectors”, IEEE Trans. Power Elect., vol. 27, no. 1, pp. 400-410,
Jan. 2012.
[26] Jeffrey Ewanchuk, John Salmon and Chris Chapelsky, “A method for
supply voltage boosting in an open-ended induction machine using a
dual inverter system with a floating capacitor bridge”, IEEE Trans.
Power Elect., vol. 28, no. 3, pp. 1348-1357, Mar. 2013.
[27] Yang Wang, Debiprasad Panda, Thomas A. Lipo and Di Pan, “Open-
winding power conversion systems fed by half-controlled converters”,
IEEE Trans. Power Elect., vol. 28, no. 5, pp. 2427-2436, May 2013.
[28] H. Stemmler and P. Guggenbach, “Configurations of high-power voltage
source inverter drives”, Proc. of EPE conf., pp. 7-12, 1993.
[29] G. Walker, “Evaluating MPPT converter topologies using a MATLAB
PV model,’’ J. Elect. Electron. Engg., vol. 21, no. 1, pp. 49---56, 2001.
[30] Byoung-kuk Lee, Ehsami. M, “A simplified functional simulation model
for three-phase voltage source inverter using switching function
concept,” IEEE Trans. Ind. Elect., vol. 48, no. 2, pp. 309-321, Apr.
2001.
[31] Sachin Jain, V. Agarwal, “Comparison of the performance of maximum
power point tracking schemes applied to single-stage grid-connected
photovoltaic systems”, IET Elect. Power App., vol. 1, no. 5, pp. 753-762,
2007.
[32] J. .-S. Kim and S.-K. Sul, “A novel voltage modulation technique of the
space vector PWM”, Conf. Proc. IPEC, pp. 742-747, 1995.