IEEE JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER ELECTRONICS, VOL. 10, NO.

1, FEBRUARY 2022 361

A Fixed Zone Perturb and Observe MPPT

Technique for a Standalone Distributed PV System
Niraja Swaminathan , Member, IEEE, N. Lakshminarasamma, Member, IEEE, and Yue Cao , Member, IEEE

Abstract— This article proposes a fixed zone perturb and maintenance [1]. The PV system is constructed in either a
observe (FZPO) technique to achieve improved steady-state centralized or a distributed architecture [2], [3], and the power
efficiency and a fast and drift-free maximum power point track- output depends on the solar irradiance and operating condi-
ing (MPPT) for photovoltaic (PV) systems without any additional
sensors. In this technique, the PV array’s power–voltage (P–V) tions [4]–[6]. In a centralized architecture, a single power con-
curves are divided into multiple zones with unique zone bound- verter is employed for the complete PV system, and therefore,
ary voltages for different irradiance conditions. This technique maximum power is harvested using a single central maximum
employs a combination of adaptive and fixed step sizes to improve power point tracking (MPPT) controller. This MPPT controller
the performance, where the adaptive step size is calculated must be capable of differentiating local maximum power
from simple mathematical equations, resulting in a reduced
computational burden. Besides, natural drift-free tracking is point (LMPP) and global maximum power point (GMPP)
achieved without any additional sampling or computation, over- as the power–voltage (P–V) characteristics of a solar panel
coming existing techniques’ limitations. The FZPO technique array exhibit multiple maximum power points (MPPs) in the
requires a PV panel’s information only at the initial design stage event of partial shading [2], [3], [7]–[11]. In contrast with the
but not during real-time tracking, making the implementation centralized architecture, the distributed architecture employs
possible using only low-cost processors. This article presents the
implementation and design of the proposed controller. The FZPO an individual dc–dc converter for each PV string, as shown in
technique’s performance is validated through a comprehensive Fig. 1, and therefore, the individual MPPT controller for each
set of hardware experiments on a buck–boost full-bridge (BBFB) PV string results in efficient tracking [2]. Besides, a simple
converter under various irradiance conditions per the EN50530 MPPT controller suffices for a distributed architecture as the
standard. For a step change in irradiance, the FZPO technique effect of partial shading is low due to individual MPP tracking
is experimentally verified to be 42% and 20% more efficient
than the conventional and VSS techniques, respectively. During of relatively small PV areas.
the irradiance varying conditions using the FZPO technique, the This work focuses on MPPT techniques for a standalone
peak power loss is one-sixth compared to the conventional and distributed solar PV system, demanding a low-cost and sim-
VSS techniques. Practical field-related considerations, such as ple MPPT controller. Hill climbing MPPT techniques are
PV panel temperature effects, are further investigated through well-known [4], [5], [12]–[18]. Among them, the conven-
experiments.
tional perturb and observe (P&O) technique is cost-effective
Index Terms— Adaptive step size, drift-free tracking,
dynamic performance, maximum power point tracking (MPPT), and simple [4], [5], [13]–[15], [18]–[20]. The conventional
model-based MPPT technique, perturb and observe (P&O), solar P&O technique tracks the MPP by perturbing the control
photovoltaics (PVs), steady-state oscillation. parameter based on the PV voltage and power changes. This
control parameter is usually the converter’s duty ratio in
I. I NTRODUCTION the direct-duty perturbation scheme or the PV voltage in

S OLAR photovoltaic (PV) sources are increasing in today’s

electricity generation due to several advantages, such as
eco-friendliness, fossil fuel-free, sound-free operation, and low
the reference-voltage scheme (later, Section III explains these
schemes in detail). The conventional MPPT controller uses a
fixed perturbation frequency and fixed step size. Perturbation
frequency defines how frequently the MPPT controller adjusts
Manuscript received November 23, 2020; revised January 27, 2021;
accepted February 26, 2021. Date of publication March 15, 2021; date of the control parameter, and step size is a magnitude change in
current version February 3, 2022. This work was supported in part by the the control parameter. With a large step size, MPP tracking
Oregon State University Foundation and in part by the Indian Institute of is fast but causes a large steady-state energy loss, while a
Technology Madras, India. Recommended for publication by Associate Editor
J. Marcos Alonso. (Corresponding author: Yue Cao.) small step size reduces the steady-state energy loss at the
Niraja Swaminathan was with the Department of Electrical Engineering, cost of sluggish tracking [1], [16], [21]. Therefore, there is
Indian Institute of Technology Madras, Chennai 600036, India. She is now a compromise between the steady-state energy loss and the
with the School of Electrical Engineering and Computer Science, Oregon State
University, Corvallis, OR 97331 USA (e-mail: [email protected]). tracking speed under the conventional P&O technique. The
N. Lakshminarasamma is with the Department of Electrical Engineering, incremental conductance-based P&O method is presented in
Indian Institute of Technology Madras, Chennai 600036, India (e-mail: [20] to improve tracking speed and reduce energy loss com-
[email protected]).
Yue Cao is with the School of Electrical Engineering and Computer pared to the conventional method. However, both techniques
Science, Oregon State University, Corvallis, OR 97331 USA (e-mail: are prone to drift in tracking due to their operating principle.
[email protected]). Drift in tracking is a condition when the MPPT controller
Color versions of one or more figures in this article are available at
https://doi.org/10.1109/JESTPE.2021.3065916. perturbs the system in the opposite direction of MPP, partic-
Digital Object Identifier 10.1109/JESTPE.2021.3065916 ularly when there is an increase in irradiance [4], [14], [17].
2168-6777 © 2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See https://www.ieee.org/publications/rights/index.html for more information.

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
362 IEEE JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER ELECTRONICS, VOL. 10, NO. 1, FEBRUARY 2022

Fig. 2. P–V curves of VSS P&O technique showing (a) slope on left and
right sides of the MPP and (b) the operating point moving toward left during
Fig. 1. Example of a distributed solar PV system. a continuous decrease in irradiance.

perfect constant. Santos et al. [13] present an MPPT technique

Although the MPPT controller eventually tracks MPP in the using PV power–conductance (P–G) characteristics, where the
subsequent perturbations, the drift will be predominant when control parameter is adjusted to match the load conductance
there is a continuous change in the irradiance, which is the case at maximum power. However, this method does not discuss
in the practical scenario. Besides, drift will also be prevailing drift.
with a large step size. An enhanced adaptive P&O (EA-P&O) in [21] overcomes
To avoid the drift, the MPPT controller in [15] perturbs the the abovementioned limitations, such as steady-state oscil-
system by considering the PV current changes together with lations and drift. This technique features an adaptive step
the PV voltage and power changes. Some MPPT techniques size for fast MPP tracking and reduced steady-state oscilla-
in the literature eliminate the drift by using multi-sampling tions. Besides, the technique is capable of tracking GMPP
techniques, in which the intermediate PV voltage and current under partial shading conditions with a reduced computational
are measured to track the climatic change by subjecting the burden. However, the EA-P&O controller tracks GMPP by
system to a small disturbance in between the usual MPPT comparing multiple samples of perturbation, and therefore, the
controller perturbations [14], [22]. As a consequence, this performance requires further validation for a fast irradiance
additional disturbance increases the implementation and con- change of 100 W/m2 /s, for example, as suggested by the
trol complexities. Besides, the response time of the PV system standard EN50530 [26].
limits the maximum perturbation frequency. Jain and Agarwal [27] present a Beta MPPT technique to
Adaptive step-size P&O techniques improve both track MPP based on the magnitude of a variable β instead of
steady-state efficiency and tracking speed [4], [5], [14], PV power. This technique uses a hybrid of adaptive and fixed
[19], [21]–[24]. One such is a variable step-size (VSS) step sizes for fast-tracking and reduced steady-state losses.
technique [23], in which the step size is adaptive as defined However, β and the scaling factor require manual tuning. This
by Dn = D(n−1) ± N ∗ |(Pn − P(n−1) /Vn − V(n−1) )|, where is overcome in [28], [29], where the parameters are autotuned
D is the duty of the converter, P and V are the solar panel to accommodate climatic changes. Besides, the modified Beta
power and voltage, and N is a scaling factor. Suffixes n and MPPT technique in [28] is capable of tracking the GMPP
(n − 1) represent the present and previous operating points. during partial shading conditions. Although Beta MPPT and
The scaling factor “N” is determined from an experiment its variants provide fast and efficient tracking, the β variable
or a complete PV system model. Therefore, this is partially computation includes a logarithmic calculation.
a model-based technique. In this technique, the step size is Apart from P&O, fractional short-circuit and open-circuit
proportional to the slope of the P–V curve, which is small techniques are other well-known MPPT techniques [14], [24],
and large in magnitude, on the left and right sides of the [30]. In these, drift does not exist; however, the MPPT
MPP point, respectively, as shown in Fig. 2(a). The VSS controller interrupts the PV system periodically by opening or
MPPT controller can be sluggish under a rapid decrease in shorting the PV terminals for a short duration. Therefore, these
irradiance due to smaller step size, as illustrated in Fig. 2(b). are not suitable for standalone (off-grid) solar applications due
Furthermore, the VSS technique exhibits a predominant drift to periodic service interruptions [23], [30], [31]. The use of a
in tracking during an increase in irradiance, as the step size secondary PV cell for the PV voltage and current measurement
is large when the operating point is on the right of MPP. overcomes these periodic interruptions of the primary PV
Tousi et al. [4] present another adaptive P&O technique, system but increases the implementation cost [14].
in which both the step size and frequency of the perturbation Some model-based MPPT techniques provide overall effi-
are adaptive and are determined from the Gaussian and arc tan- cient tracking. One such method tracks the MPP based on
gent functions involving trigonometric and exponential terms. an estimation from a PV array model [32]. Unlike the P&O
This technique is model-based with a complicated design and techniques, this one does not oscillate around the MPP, thereby
requires a high-end processor. The zero-oscillation adaptive improving the steady-state efficiency. Predictive control, fuzzy
MPPT (ZA-MPPT) technique in [25] not only provides an logic, particle swarm, and neural network [10], [11], [33]–[37]
adaptive step size but also restricts the MPPT controller to are some of the other model-based MPPT techniques. Model
idle when the irradiance is constant to reduce the energy predictive control-based MPPT approaches provide better
loss. However, the ZA-MPPT method is not effective in steady-state and dynamic performance [34]. A polynomial
practical implementation as the irradiance is not always a fuzzy logic-based technique in [36] requires a dc–dc converter

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
SWAMINATHAN et al.: FZPO MPPT TECHNIQUE FOR STANDALONE DISTRIBUTED PV SYSTEM 363

TABLE I
C OMPARISON OF MPPT T ECHNIQUES IN THE L ITERATURE

model, and designing the MPPT controller is complex. One 7) implementable even on low-end processors. Besides, the
such method predicts the irradiance to track MPP, but tem- proposed FZPO technique is not converter specific and can
perature measurement is needed [37]. Hartmann et al. [38] be implemented on any dc–dc converter. This new approach
present a model for MPP locus with temperature variations requires PV panel characteristics but only at the initialization
combining both the heuristic and model-based techniques. stage.
In general, most of the model-based approaches require a The rest of the article is organized as follows. Section II
high-end processor to compute sophisticated mathematical details the proposed FZPO theory. Section III explains the
equations. Some of them are certain dc–dc converter specific, implementation aspects and provides a design example of
while some require temperature sensors. the FZPO controller on a dc–dc converter. In Section IV,
As a summary of the literature review, Table I compares the aforementioned features are validated by experiments,
popular heuristic and model-based MPPT techniques based on in which a 200-W buck–boost full-bridge (BBFB) converter
the step size, complexity, drift in tracking, sensor requirement, is chosen. The performance is compared with selected other
and sampling count per perturbation. Overall, the heuristic MPPT techniques based on the start-up time, steady-state
and model-based MPPT techniques presented in the literature oscillations, drift occurrence, and all of these at constant, slow,
exhibit one or more of the following limitations: 1) large and fast varying irradiances per the EN50530 standard [26].
energy loss at steady state; 2) drift in tracking; 3) slow Section V concludes this work.
tracking; 4) complicated mathematical computations; 5) mul-
tiple sampling within one MPPT tracking period; 6) inability II. P ROPOSED FZPO T ECHNIQUE
to track fast-changing irradiance conditions; and 7) extra
The proposed FZPO technique aims to split the P–V curves
sensor/measurement. These limitations can be eliminated when
into several linear regions (zones) and employs a mix of fixed
the present operating point on the P–V curve is located
and adaptive step sizes to improve the performance at all
accurately. This leads to the proposal of a zonal-based MPPT
irradiance conditions. The FZPO technique provides natural
technique, where the P–V curve is partitioned into multiple
drift-free tracking without additional sensors or sampling,
zones to pinpoint the operating point’s location.
making the implementation simple.
With the above limitations in mind, this article proposes
Upcoming subsections discuss the concept, analysis, and
a fixed zone P&O (FZPO) technique for a distributed solar
operation of this FZPO technique.
PV system. In this technique, P–V curves are divided into
five zones, and the step size for four zones is adaptive, while
the step size is fixed for one remaining zone. The proposed A. Defining Zones
FZPO MPPT approach features: 1) improved steady-state In the FZPO technique, P–V curves are divided into multiple
performance, with reduced oscillations and energy loss; 2) fast zones. Three is the minimum number of zones; however,
natural drift-free tracking at all operating conditions; 3) sim- at lower irradiance conditions, the tracking speed can be slow,
ple mathematical equations for step-size calculation, i.e., and the linearity of the P–V curve in certain regions can be
less computational burden; 4) no additional sampling; 5) no lost. Therefore, in this work, the PV curves are divided into
extra sensor; 6) no dc–dc converter model required; and five zones, which is the next larger zone number, as shown in

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
364 IEEE JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER ELECTRONICS, VOL. 10, NO. 1, FEBRUARY 2022

Fig. 3. P–V characteristics curve of the proposed FZPO technique. Fig. 4. P–V curves of the solar PV array at different irradiances and
temperatures.

Fig. 3. More than five zones increase the complexity of the

algorithm. In Fig. 3, Zone 1 and Zone 5 are the non-MPP panel (K), respectively; G ref and G are reference irradiance
zones and are far from the MPP. Zone 3 contains the MPP, and actual irradiance (W/m2 ), respectively; Q d is the diode
and Zone 2 and Zone 4 are the transition zones. The challenge ideality factor; k is the Boltzmann’s constant (J/K); q is the
is a mathematical definition of these zone boundaries. B12 , coulomb constant (C); E g is band-energy gap (eV); and K Voc
B23 , B34 , and B45 lines represent the zone boundaries and are and K Isc are temperature coefficients of open-circuit voltage
functions of PV voltage and current. Boundary points, where (V/◦ C) and short-circuit current (A/◦ C), respectively.
the boundary lines intercept with the irradiance curve, are Although zones in this technique are divided based on solar
represented in terms of voltages. These boundary voltages are array P–V curves, small variations in the parameters will
unique to different irradiance curves to improve the proposed not affect the performance. This is because the perturbation
FZPO controller’s performance since the MPP voltage is not step size depends only on the operating PV voltage and
the same at various irradiance conditions and temperatures. current but not on the actual panel parameters, as presented
The zone boundaries are defined based on the change in next. Therefore, accurate experimental P–V curves are not
solar PV conductance (P/V 2 ) using the slope of the P–V mandatory; instead, the proposed FZPO controller can be
curve versus PV voltage characteristics for the two extreme designed using the theoretical P–V curves obtained using the
conditions, including a maximum irradiance (1000 W/m2 ) PV array datasheet parameters.
with a lower panel temperature (25 ◦C) and a minimum Using a linear curve fitting of the P–V curve slope versus
irradiance with a higher panel temperature (per designer’s PV voltage characteristics, four points are obtained on each
choice, such as 300 W/m2 and 55 ◦ C), as shown in Fig. 4. curve at which the change in conductance occurs, as illustrated
The solar PV conductance at the two extreme conditions is in Fig. 5(a) (at maximum irradiance) and (b) (at minimum
obtained by plotting the P–V curve slope (P/V) versus PV irradiance). The pink dashed lines shown in Fig. 5(a) and
voltage, as shown in Fig. 5(a) and (b), respectively. For this, (b) are used for curve fitting. The intersection points on the
the P–V curve slope [in other words, change in power (P) for P–V curves define the boundary voltages. VB12(max) , VB23(max) ,
a given change in the voltage (V)] is obtained from the solar VB34(max) , and VB45(max) represent the zone boundary voltages
PV model [23], [25], [39] given in the following equations. at the maximum irradiance, and VB12(min) , VB23(min) , VB34(min) ,
It is worth noting that these equations are also temperature- and VB45(min) represent the zone boundary voltages at the
dependent minimum irradiance. After defining the boundary voltages,
 
V
  boundaries B12 , B23 , B34 , and B45 , which are straight lines,
− oc_t
Iph1 − VPV1 VIscT N p e Ns VT Tpar + R1p are formed by joining the minimum and maximum boundary
IPV1 =     (1) voltages, as shown in Fig. 5(c). As a result, the four boundary
V
− oc_t equations are defined as
1 + VIscT N p Rs e Ns VT Tpar + RRsp
VB12 (g) = VB23 (g) + A1 (6)
P = VPV1 I + IPV1 V + V I (2)
VB23 (g) = m 23 In (g) + c23 (7)
where
VB34 (g) = m 34 In (g) + c34 (8)
G   
Iph1 = N p Isc + (T − Tref ) · K Isc (3) VB45 (g) = VB34 (g) + A2 (9)
G ref
 
T 3 1
−1
qEg where In (g) is the operating PV current at irradiance “g” (A);
Tpar = e Tref T k Qd (4) m 23 and m 34 are the slopes of boundary lines B23 and B34
Tref
Voc_t = (T − Tref ) · K Voc + Voc (5) (); c23 and c34 are the x-intercepts of boundary B23 and B34
(V); and A1 and A2 are the voltage differences between the
where VPV1 , Voc , and VT are PV voltage, open-circuit voltage, parallel boundaries B12 and B23 and B45 and B34 (V).
and thermal potential (V), respectively; IPV1 , Iph1 , and Isc Using (6)–(9), the FZPO controller determines boundary
are PV current, photon current and short-circuit current (A), voltages of any operating irradiance “g.” After determining the
respectively; N p and Ns are numbers of parallel strings and boundary voltages, the FZPO controller correctly identifies the
series connected panels, respectively; Rs and R p are series operating point zone from the present PV voltage value. For
and parallel resistances of the PV array (), respectively; Tref example, let VB12 = 19.3 V, VB23 = 23.9 V, VB34 = 25.8 V,
and T are reference temperature and actual temperature of the and VB45 = 27 V be boundary voltages at one given condition,

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
SWAMINATHAN et al.: FZPO MPPT TECHNIQUE FOR STANDALONE DISTRIBUTED PV SYSTEM 365

Fig. 5. (a) Slope of P–V curve versus PV voltage for case 1 (maximum irradiance and minimum PV panel temperature). (b) Slope of P–V curve versus PV
voltage for case 2 (minimum irradiance and maximum PV panel temperature). (c) P–V curves for cases 1 and 2.

and let the present operating PV voltage be 24 V. In this case,

the FZPO controller identifies the operating point to be in
Zone 3 since the present PV voltage (24 V) is between VB23
and VB34 .
After identifying the operating zone, the FZPO controller
computes the perturbation step size for the corresponding zone,
as discussed in Section II-B.

B. Mixed Step Size

The FZPO technique utilizes a combination of fixed and
adaptive step sizes, depending on which zone the operating
point lies in, while, in Zone 3, as the operating point is in
the MPP vicinity, the step size is kept small and constant to
reduce the steady-state energy loss. The working principle of
the FZPO in this zone is similar to that of the conventional
P&O. On the other hand, the step size (step) in the remaining
four zones is adaptive and computed using

step = m(Vn (g) − VB (g)) + c (10)

Fig. 6. Flowchart of the proposed FZPO technique.
where VB (g) is the boundary voltage of the adjacent zone
toward the MPP at a given irradiance “g” (V); m is the slope For MPP tracking, the FZPO controller measures the present
of the linear fitting curve (V −1 ); and c is the y-intercept. operating PV voltage Vn and current In , as the first step. The
This equation is linear as the P–V curve in these zones is FZPO controller then determines the zone boundary voltages
almost linear. for the present operating irradiance from (6) to (9), which
Here, the step size step is directly proportional to the is step 2. Once zone boundaries are determined, the FZPO
difference between the present operating voltage Vn (g) and controller identifies the present operating zone as step 3.
the respective boundary voltage VB (g) such that the operating Step 4 is to compute step. If the present operating point
point moves to the next zone toward the MPP faster. Besides, is already in Zone 3, step is fixed. If not in Zone 3,
the sign of [Vn (g) − VB (g)] in (10) decides the direction of step is calculated from (10). The resultant control parameter,
the perturbation, and hence, no separate logic is required. It is CP, which can be duty or phase shift, or any other feasible
also noted that the step size equation is independent of the PV parameter based on the converter, is obtained by adding (or
array parameters, and therefore, the MPPT performance is not subtracting) the step. As a next step, the FZPO controller
affected by the parameter variations. updates this resultant CP in the converter and waits for the next
perturbation period to measure Vn and In again and repeat the
C. FZPO Controller Operation process.
The flowchart in Fig. 6 presents step-by-step operations of
the proposed FZPO controller. In design stages 1 and 2, the D. Natural Drift-Free Operation
panel characteristics are fetched; zone boundaries and step size The FZPO controller operates with natural drift-free track-
(6)–(10) are defined, as described in Section II-A and II-B. ing. The term natural drift-free operation in this article
These are performed offline and are executed only once. Steps is coined as the drift is naturally eliminated due to the
1–6 are the real-time MPP tracking algorithms of the FZPO FZPO controller’s operation, without requiring any additional
technique. algorithm.

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
366 IEEE JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER ELECTRONICS, VOL. 10, NO. 1, FEBRUARY 2022

Fig. 7. Natural drift-free operation for step increase in irradiance from

300 to 1000 W/m2 .

Consider an operating point “A” at 300 W/m2 irradi- Fig. 8. Solar PV-fed buck–boost full-bridge converter.
ance, as shown in Fig. 7. When the irradiance increases to TABLE II
1000 W/m2 , “A” shifts automatically to “B,” where the load S AMPLE PARAMETERS FOR THE FZPO T ECHNIQUE
and new source curves meet (per the definition of an operating
point). At this condition, the FZPO controller first identifies
the “B” zone, which is Zone 5 (see Fig. 7). Then, the FZPO
controller computes the step size step, which is positive for
this scenario as (Vn |@B − VB45 ) in (10) is greater than zero.
The positive step increments the control parameter CP, thus
moving the operating point toward MPP to “D” avoiding drift,
as indicated by the green arrow in Fig. 7.
In some cases, the new shifted operating point can be
located in Zone 3. In such conditions, though the FZPO
controller behaves similar to the conventional controller in this
zone, the drift will still be negligible due to the smaller step
size. It is noted that the new operating point will be in Zone 3
mostly when there is a minimal increment in the irradiance,
and hence, the power loss due to the drift while in Zone 3
will be negligible.
The FZPO controller is implemented on a power converter
to verify the concept and analysis presented in this section. The
implementation aspects with the design example are presented
in Section III.

III. I MPLEMENTATION A SPECTS OF

THE FZPO T ECHNIQUE
Fig. 9. Block diagram of the decoupled control scheme.
To demonstrate any MPPT method, a power electronics
converter is a required platform. In this work, a BBFB compensation loop [40], as shown in Fig. 9. This compensation
converter shown in Fig. 8 is chosen for implementation loop mitigates the dc current in the transformer of the BBFB
with the specifications mentioned in Table II. A minimum converter by using an asymmetrical duty control (D ∗ ) on
irradiance of 300 W/m2 (which is approximately 50 W) and the leading leg MOSFETs S2 and S3 , while the duty (D)
a maximum panel temperature of 55 ◦ C are considered. These of the lagging leg MOSFETs S1 and S4 is 50%, as shown
numbers are just an example case for demonstration purposes. in Fig. 9. The dc-current compensation loop’s input is the
However, there is no constraint on the minimum irradiance transformer primary dc current (Im(dc) ), which is the difference
and maximum temperature choices for FZPO. Also, the FZPO between positive (I p ) and negative (I " p ) transformer peak
technique can be implemented on any power converter, for current magnitudes. The control parameter CP for the FZPO
which the analysis presented in this work still holds. MPPT loop is the phase shift (φ). Therefore, the duty reference
The BBFB converter features a high gain and continuous (D ∗ ) from the dc-current compensation loop and the phase
input and output currents with high power capability [40]. shift reference (φ) from the MPPT loop are together modulated
In this work, BBFB is connected to a resistive load for with the ramp (carrier) signal to generate gating signals for the
demonstration purposes. Besides, the FZPO performance is leading leg MOSFETs S2 and S3 , as shown in Fig. 9.
not affected by the load type because the MPPT controller A PI controller is sufficient for the dc-current compensation
only requires voltage and current measurements from the input loop since the plant transfer function is a first-order system
side. [40], as given in
BBFB Converter—Control Scheme: The BBFB con-
verter uses a decoupled control scheme consisting of the Iˆpri−dc (s) VPV
= (11)
proposed FZPO MPPT technique and a separate dc-current d̂(s) sLm + Ron

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
SWAMINATHAN et al.: FZPO MPPT TECHNIQUE FOR STANDALONE DISTRIBUTED PV SYSTEM 367

TABLE III
C ONSTANTS OF Z ONE B OUNDARY E QUATIONS

Design Example: From Fig. 5(a) and (b), boundary voltages

VB12(max) , VB12(min) , VB23(max) , and VB23(min) are obtained,
as given in the following equation. Similarly, PV currents at
VB23(max) and VB23(min) are obtained from Fig. 5 as given in
the following equation:

VB12(max) = 19.00 V, VB12(min) = 14.50 V (13)

VB23(max) = 23.00 V, VB23(min) = 18.50 V (14)
@ VB23(max) , In = 7.8 A (15)
Fig. 10. Methods of P&O implementations. (a) Reference voltage perturba- @ VB23(min) , In = 2.8 A. (16)
tion. (b) Direct-duty perturbation.
Substituting above values in (7), the following equations are
where Ipri−dc is the dc current in the transformer (A); d is obtained as follows:
the change in duty in the leading leg MOSFETs; L m is the
magnetizing inductance (H); and Ron is the on-state resistance 23 = 7.8m 23 + c23 (17)
of the MOSFETs (). 18.5 = 2.8m 23 + c23 . (18)
The dc voltage gain (M) of the BBFB converter with this
control scheme is given as Solving (17) and (18) results in the values of constants
m 23 and c23 as 0.9  and 15.98 V, respectively. Likewise,
2nφ substituting the values of VB12(max) and VB23(max) in (6) results
M= (12)
1+ 4n 2 f s L lk in constant A1 to be −4 V. Now, (6) and (7) constants are
Ro
fully determined.
where n is the transformer turns ratio; φ is the phase shift Following the same procedure, constants of (8) and (9)
of the control signal; L lk is the leakage inductance of the are obtained. Table III presents the summary of all the zone
transformer (H); f s is the converter switching frequency (Hz); boundary equation constants.
and Ro is the load resistance (). For a real-time operating point, using (6)–(9) with the design
FZPO Controller—Implementation Techniques: The parameters given in Table III, the proposed FZPO controller
P&O implementation is classified into reference voltage and first computes all the boundary voltages based on the PV
direct-duty perturbation [5], [13], [14]. In the reference voltage operating current and then locates the actual zone given the
perturbation scheme shown in Fig. 10(a), the MPPT controller operating voltage. Even though the operating current may be
perturbs the reference voltage (v ref ) based on the change in the same across different irradiances and temperatures, the
PV voltage VPV and current IPV . The PI controller adjusts the FZPO controller still identifies the operating zone accurately.
control parameter to match the PV voltage to the reference To understand and validate this, consider six operating points
voltage Vref [5], [13]. In this method, improved transient “A” to “F” with the same operating current of 2.8 A at different
stability is observed, but the controller design is based on the irradiances and temperatures, as shown in Fig. 11.
converter model. For the assumed PV current of 2.8 A, the boundary voltages
In the direct-duty perturbation method shown in Fig. 10(b), VB12 to VB45 are obtained from (6) to (9) as 14.5, 18.5,
the duty (or any other control parameter) of the converter is 25.62, and 26.87 V. These boundary voltages are the same
perturbed directly by the MPPT controller without requiring for all the considered operating points “A” to “F,” as they
a PI controller [14]. Therefore, the MPPT controller design is only depend on the PV current. The FZPO controller then
not converter specific. This implementation method is consid- identifies the operating zone for each point by comparing the
ered in this article. The design of the FZPO controller parame- present operating voltage with the calculated zone boundaries.
ters, based on the theory presented in Section II, is discussed For example, the FZPO identifies the operating zone of “A”
next. as Zone 2 by comparing the operating voltage of 17.79 V
FZPO Controller—Zone Boundary Equation Design: Under with the corresponding calculated zone boundary voltages, i.e.,
the overall conditions from Table I, the boundary voltages between 14.5 and 18.5 V. This is observed to be accurate
at maximum and minimum irradiance conditions referring to from Fig. 11(b). Table IV provides the zones of “A” to “F” as
Fig. 5(a) and (b) are obtained. Similarly, PV currents at these identified by the FZPO controller (termed “Zone–theoretical”)
boundary voltages are obtained from Fig. 5(c) to determine and the actual zones identified from Fig. 11(b). It is evident
the zone boundary equation constants. The design example of that, even if the PV current of all the considered points is the
determining zone boundary (6) and (7) constants is presented same, the proposed FZPO controller still identifies the zones
below. correctly.

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
368 IEEE JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER ELECTRONICS, VOL. 10, NO. 1, FEBRUARY 2022

Fig. 12. Hardware experimental setup of 200-W BBFB converter.

TABLE IV
VALIDATION OF THE Z ONE I DENTIFICATION

Fig. 11. (a) I –V curves at different irradiances and panel temperatures TABLE V
showing the operating points at the same current of 2.8 A. (b) P–V curves S TEP -S IZE E QUATION C ONSTANTS FOR E ACH Z ONE
at different irradiances and panel temperatures showing the corresponding
operating points.

FZPO Controller—Step-Size Equation Design: The adap-

tive step sizes for Zones 1, 2, 4, and 5 are computed from
(10). Minimum and maximum step sizes for each zone define
the step-size equation constants, as discussed in the design chosen for hardware implementation and side-by-side com-
example below. parison. The MPPT perturbs every 1 s for all the techniques
Design Example: Based on the intuitive approach, the implemented in this work. The conventional P&O step-size
maximum and minimum step sizes are defined in following is chosen as 4.5%. For VSS, the scaling factor N is tuned
equations: as 4 based on the design guidelines provided in [23]. For
tuning N in VSS, a maximum step considered is 8%, also
Zones 1, 5 : stepmin = 6%, stepmax = 8% (19) equivalent to that of the proposed FZPO technique.
Zones 2, 4 : stepmin = 2%, stepmax = 6%. (20) The performance of the abovementioned techniques is com-
pared based on their tracking speed, steady-state oscillation,
Referring to (19), for Zone 1, when the operating point is MPPT efficiency, and drifting. Additional tests are conducted
at 0 V, the step size is 8%, and when it is at VB12(max) , the at various conditions exhibiting the practical scenario, per
step size is 6%. It is worth noting that the step size reduces standard EN50530 [26].
as the operating point moves toward MPP. Substituting these The steady-state MPPT efficiency is calculated using [26]
conditions in (10), the following equations are obtained
PPV(avg)
ηsteady−state = (23)
−8 = m(0 − VB12(max) ) + c (21) PMPP
−6 = m(VB12(max) − VB12(max) ) + c. (22) where PPV(avg) and PMPP are the average power supplied by
the PV array and the actual MPP power, respectively (W ).
Solving (21) and (22) results in m and c values to be 0.105 In general, the MPPT efficiency for any given window is
and −6, respectively. Similarly, constants for the remaining computed based on the energy ratio as [26]
zones are obtained in Table V.  t2
With the above-designed parameters, the FZPO controller t PPV (t).dt
ηwindow =  t21 (24)
t1 PMPP (t).dt
is ready to be implemented on the BBFB converter hardware.

where PPV (t) and PMPP (t) are the instantaneous power sup-
IV. H ARDWARE E XPERIMENTAL VALIDATION plied by the PV array and the instantaneous MPP power,
The proposed FZPO technique is verified experimentally respectively (W ); t1 and t2 are the lower and upper limits of
on a 200-W BBFB converter prototype. Experimental setup the window considered (s).
with the specification given in Table II is shown in Fig. 12. Constant Irradiance: The proposed FZPO technique is
A Chroma 62050H-600S simulator is used as the solar PV tested at constant irradiance of 300, 500, and 1000 W/m2 . The
source. results are shown in Fig. 13(a), (b), and (c), respectively. The
Other MPPT techniques, such as VSS [23] and conventional PV voltage and current oscillations are higher at 300 W/m2
P&O with similar complexity levels as that of FZPO, are also than those at 1000 W/m2 due to the constant step size used

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
SWAMINATHAN et al.: FZPO MPPT TECHNIQUE FOR STANDALONE DISTRIBUTED PV SYSTEM 369

Fig. 13. Hardware experimental results of the proposed FZPO technique at constant irradiance. (a) 300 W/m2 . (b) 500 W/m2 . (c) 1000 W/m2 .

Fig. 14. Hardware experimental results: performance with step change in irradiance from 300 to 1000 W/m2 at time t = 15 s. (a) Test Waveform.
(b) Conventional P&O. (c) VSS P&O. (d) Proposed FZPO.

Fig. 15. Hardware experimental results: performance with a step change in irradiance from 1000 to 300 W/m2 at time t = 15 s. (a) Test Waveform.
(b) Conventional P&O. (c) VSS P&O. (d) Proposed FZPO.

for Zone 3. The results validate the MPP tracking at different conventional and VSS techniques to reach steady state is 7 and
irradiance conditions. 5 s, respectively, with drift in tracking, as shown in Fig. 14(b)
The FZPO technique is further tested at varying and (c), respectively. However, the proposed FZPO technique
irradiance conditions as suggested in the standard reaches the steady state in 4 s with drift-free tracking, as shown
EN50530, and the results are discussed further in this in Fig. 14(d).
section. Step Decrease in Irradiance: For a similar experiment when
Starting and Steady-State Behaviors: Two metrics, the a step down in irradiance is from 1000 to 300 W/m2 at t =
response time to reach the steady-state at starting and the 15 s, as shown in Fig. 15(a), the VSS performance, as shown
steady-state energy loss, are experimentally measured for in Fig. 15(b), is observed to take 13 s to reach steady state as
conventional, VSS, and FZPO controllers at a constant its step size is small at this transition, while the conventional
1000-W/m2 irradiance. The results are given in Table VI. For technique takes only 3 s, as shown in Fig. 15(c). However, the
starting, the FZPO tracking speed is faster than the conven- FZPO technique tracks the steady-state in only 2 s, as shown
tional and VSS techniques by 50% and 22%, respectively. For in Fig. 15(d).
steady state, the energy loss from the proposed FZPO is on par According to Standard EN50530 [26], the irradiance can
with the VSS approach while significantly reduced compared vary at the rate of 0.5 to 100 W/m2 /s in the practi-
to the conventional method, as illustrated in the steady-state cal scenario. A slow irradiance change (say 0.5 W/m2 /s)
average power and MPPT efficiency values from the usually does not affect MPPT’s performance as the irra-
table. diance change between the tracking periods is negligible.
Although the FZPO steady-state performance is not signif- However, a fast change in irradiance may affect MPPT’s
icantly better than VSS at a constant irradiance, the main performance significantly. Therefore, to validate the perfor-
advantage is observed under a dynamic irradiance change, mance at close-to-reality conditions, ramp sequence tests at
which is discussed next. 20 and 100 W/m2 /s are presented. In some test sequences,
Step Increase in Irradiance: The Chroma PV emulator is the dwell time between rise and fall of the irradiance is
programmed to provide a step change in irradiance from 300 to provided for MPPT algorithms to stabilize, as recommended
1000 W/m2 at t = 15 s, as shown in Fig. 14(a). Time taken for in EN50530 [26].

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
370 IEEE JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER ELECTRONICS, VOL. 10, NO. 1, FEBRUARY 2022

TABLE VI
MPPT T ECHNIQUE P ERFORMANCE D URING S TEP C HANGE IN I RRADIANCE F ROM 300 TO 1000 W/m2

Fig. 16. Hardware experimental results: performance with ramp test sequence (medium–high irradiance) as per EN50530 standard [26], having the slope of
20 W/m2 /s and a dwell time of t = 30 s. (a) Test waveform. (b) Conventional P&O. (c) VSS P&O. (d) Proposed FZPO.

Fig. 17. Hardware experimental results: performance with ramp test sequence (medium–high irradiance) as per EN50530 standard [26], having the slope of
100 W/m2 /s and dwell time of t = 30 s. (a) Test waveform. (b) Conventional P&O. (c) VSS P&O. (d) Proposed FZPO.

Ramp Sequence at 20 W/m2 /s With 30-s Dwell Time: Ramp Sequence Varying Continuously at 20 W/m2 /s: In the
Fig. 16(a) shows the ramping irradiance with a slope of practical scenario, the irradiance tends to change continuously,
20 W/m2 /s, as recommended by EN50530 [26]. With the particularly on partial-cloudy days. Therefore, the MPPT
conventional and VSS techniques, drift is observed, particu- techniques are tested under continuously varying irradiance
larly during the rise in irradiance, as shown in Fig. 16(b) and as recommended by EN50530 [26] with a triangular profile
(c), respectively. Furthermore, drift is significant in VSS as with a positive or negative 20-W/m2 /s slope, as shown in
the step size is large during this test. However, the proposed Fig. 18(a). The conventional and VSS techniques’ response
FZPO operation, as shown in Fig. 16(d), does not show drift shown in Fig. 18(b) and (c) show drift in tracking during the
in tracking. Therefore, the experimental MPPT efficiency for rise in irradiance. However, the proposed FZPO technique’s
the 20–120-s window with the proposed FZPO technique is performance in Fig. 18(d) reveals that the tracking is smooth
as high as 98.2%, while, with the VSS and conventional even under continuously varying irradiance conditions.
techniques, it is 94.3% and 97.95%, respectively. Ramp Sequence Varying Continuously at 100 W/m2 /s:
Ramp Sequence at 100 W/m2 /s With 30-s Dwell Time: Furthermore, the same MPPT techniques are tested under
Fig. 17(a) shows the ramping irradiance with a slope of continuously varying irradiance as recommended by EN50530
100 W/m2 /s, as recommended by EN50530 [26]. The con- [26] with a triangular profile having a positive or negative
ventional and VSS techniques fail to track the MPP during the 100-W/m2 /s slope, as shown in Fig. 19(a). The perfor-
fast increase in irradiance and only track after the irradiance mance of the conventional and VSS techniques, as shown
settles (after t = 25 s), as shown in Fig. 17(b) and (c), in Fig. 19(b) and (c), shows that they fail to track during
respectively. Meanwhile, the FZPO technique tracks the MPP continuously fast-changing irradiance. In contrast, the pro-
even during a fast irradiance change, as shown in Fig. 17(d). posed FZPO tracks the MPP, as shown in Fig. 19(d). The
The measured MPPT efficiencies of the conventional, VSS, power losses during (10.5 − 24.5 s) are plotted in Fig. 20(b).
and FZPO techniques during the period of (18 − 62 s) are It is observed that the loss is less than 10% for the FZPO
86.7%, 89.5%, and 97.85%, respectively. The power loss (p.u.) technique, while it reaches as high as 70% for the conventional
is shown in Fig. 20(a), which indicates the minimum loss with and VSS techniques. This particular test reveals a significant
the FZPO technique. This experimental exercise unveils a good performance improvement with the FZPO technique compared
dynamic performance for the FZPO technique. to the others.

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
SWAMINATHAN et al.: FZPO MPPT TECHNIQUE FOR STANDALONE DISTRIBUTED PV SYSTEM 371

Fig. 18. Hardware experimental results: performance with the continuous ramp test sequence (medium–high irradiance) per EN50530 standard [26], with a
slope of ± 20 W/m2 /s. (a) Test waveform. (b) Conventional P&O. (c) VSS P&O. (d) Proposed FZPO.

Fig. 19. Hardware experimental results: performance with the continuous ramp test sequence (medium–high irradiance) as per EN50530 standard [26], with
a slope of ±100 W/m2 /s. (a) Test waveform. (b) Conventional P&O. (c) VSS P&O. (d) Proposed FZPO.

Fig. 20. Hardware experimental results. (a) Power loss during the window Fig. 21. P–V curves at different irradiances and temperatures indicating the
of (18–62 s) for the test waveform in Fig. 17(a). (b) Power loss during the
MPP points and zones.
window of (10.5–24.5 s) for the test waveform in Fig. 19(a).

The experimental performance of MPPT techniques at

several test conditions discussed above is summarized in
Table VII. It is observed that the MPPT efficiency of the
FZPO technique is about 4%, 9%, and 47% higher than the
VSS techniques during slow, fast, and continuously varying
irradiances, respectively.
Effect of PV Panel Temperature Variations: In a practical
Fig. 22. Hardware experimental results of the proposed FZPO technique.
field-related scenario, panel temperature changes constantly, (a) Step change from 1000 W/m2 25 ◦ C to 300 W/m2 55 ◦ C. (b) Step change
though not as fast as the irradiance does. The change in from 300 W/m2 55 ◦ C to 1000 W/m2 25 ◦ C.
panel temperature affects the PV voltage due to a temperature
coefficient of −0.35%/◦ C, resulting in the P–V curve shifting
horizontally. Therefore, validating the proposed FZPO tech- MPP tracking of the FZPO technique when subjected to a
nique for temperature variations is necessary. step fall in irradiance from 1000 W/m2 25 ◦ C to 300 W/m2
The FZPO zones are designed using (1)–(5) following the 55 ◦ C. The tracking speed at these two conditions is similar
guidelines given in Section II. Fig. 21 shows the P–V curves to the step change in irradiance under the uniform 25 ◦ C case.
with zones at different irradiances and panel temperatures, for Besides, a difference of about 3.5 V in the MPP voltage at
the parameters given in Table II. As (1)–(5) are the functions of 55 ◦ C is observed due to the negative temperature coefficient
irradiance and panel temperature, their effects are taken care of the PV panel material, verifying the effects of temperature
of at the design stage, to ensure FZPO tracking. To verify change.
this, the step changes between two extreme conditions, i.e., Significant Features and Contributions of the FZPO
1000 W/m2 25 ◦ C and 300 W/m2 55 ◦ C, are considered. Technique: From the hardware experimental verification
Fig. 22(a) shows the MPP tracking of the FZPO tech- and implementation, several features, advantages, and con-
nique when subjected to a step rise in irradiance from tributions of the proposed FZPO are summarized as
300 W/m2 55 ◦ C to 1000 W/m2 25 ◦ C. Fig. 22(b) shows the follows.

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
372 IEEE JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER ELECTRONICS, VOL. 10, NO. 1, FEBRUARY 2022

TABLE VII
MPPT S CHEME P ERFORMANCE D URING DYNAMIC I RRADIANCE C ONDITIONS

low. Under a significant partial shading condition, the FZPO

technique may track the GMPP with limitations. In particular,
when the GMPP exhibits in Zone 3 as in Case 1 in Fig. 23,
FZPO will track the point as normal. However, FZPO fails
when the GMPP lies outside Zone 3, as in Case 2, when
it instead tracks the LMPP in Zone 3. Besides, the FZPO
algorithm may become confused when the regular GMPP and
an extra LMPP both exist in Zone 3.
Fig. 23. Two different P–V curves during partial shading. Case 1: FZPO
successfully tracks as GMPP is located in Zone 3. Case 2: FZPO fails to track
GMPP, instead tracks LMPP that is in Zone 3. V. C ONCLUSION
This article proposes a fixed zone P&O (FZPO) MPPT tech-
1) The proposed zonal-based technique locates the operat- nique to overcome the limitations, including high steady-state
ing point irrespective of the irradiance and panel tem- loss, poor dynamic performance under varying irradiance
perature variations, while most of the heuristic MPPT conditions, drift in tracking, multiple samplings per perturba-
techniques fail to do so. tion, need for a high-end computing processor, and need for
2) Though the proposed FZPO technique is model-based, additional temperature and irradiance sensors. In the FZPO
the design parameters are computed only once in an technique, the P–V curves are divided into multiple zones,
off-line mode. At the same time, online step-size com- with the boundary voltages varying for different irradiance
putations involve only linear equations. conditions. A mixed step size with simple linear equations is
3) Despite being a model-based technique, the proposed employed, making the computations more efficient. Besides,
FZPO does not require high-end processors due to the FZPO technique achieves natural drift-free tracking with-
simple computations, thereby making it cost-effective out an additional sampling/algorithm. The FZPO technique
and efficient. requires a one-time computation of the controller parameters
4) The perturbation step size in this technique is inde- from the PV panel information at the initial design stage.
pendent of the PV array, and therefore, the parameter However, it does not require any complex computation during
variations in the array due to aging or other factors do real-time tracking. The proposed FZPO converter is best suited
not affect the performance. for the standalone distributed PV systems where a low-cost,
5) The zone boundary voltages are unique for different simple MPPT controller is the key requirement.
irradiance levels, improving the tracking speed even at The proposed FZPO scheme is validated on a 200-W
a lower irradiance level. experimental prototype of a BBFB converter with a decou-
6) The mixed perturbation step size improves both pled control scheme. The design and implementation of the
steady-state and dynamic performances. MPPT controller are discussed. Several experiments present
7) The perturbation step size and the zone boundaries are a performance comparison of the conventional, VSS, and
defined based on linear equations, which reduces the FZPO techniques under constant/slow/fast varying irradiance
complexity of the implementation. conditions. Per the guidelines of Standard EN50530, the ramp
8) The proposed FZPO technique exhibits natural drift-free change in irradiance at two different rates with and without
operation as the direction of perturbation is embedded a dwell time is tested. It is verified that, even during a fast
in the step-size computation. irradiance change, the proposed FZPO technique offers strong
9) This technique exhibits efficient tracking during slow dynamic performance. The FZPO tracks the MPP with an
and fast varying irradiance conditions. efficiency of 42% and 20% more than the conventional and
10) No additional sensors apart from the voltage and current VSS techniques for a step rise in irradiance. Similarly, the
sensors are required. proposed FZPO technique is 24% and 47% more efficient than
Limitation During Partial Shading: The FZPO technique the conventional and VSS techniques during a continuously
is generally reliable for the distributed PV architecture, as the varying irradiance condition. Besides, under all test conditions,
scope of this article, where partial shading effect is expected the FZPO provides drift-free tracking. The drift-free operating

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
SWAMINATHAN et al.: FZPO MPPT TECHNIQUE FOR STANDALONE DISTRIBUTED PV SYSTEM 373

nature results in reduced power loss, particularly during the [19] H. Cai, J. Xiang, and W. Wei, “Decentralized coordination control of
dynamic condition. The power loss plots unveil the FZPO multiple photovoltaic sources for DC bus voltage regulating and power
sharing,” IEEE Trans. Ind. Electron., vol. 65, no. 7, pp. 5601–5610,
technique having 0.1-p.u. power losses while the conventional Jul. 2018.
and VSS techniques having as high as 0.6-p.u. power losses. [20] Y.-C. Kuo, T.-J. Liang, and J.-F. Chen, “Novel maximum-power-point-
Also, practical field-related issues, such as PV panel tempera- tracking controller for photovoltaic energy conversion system,” IEEE
Trans. Ind. Electron., vol. 48, no. 3, pp. 594–601, Jun. 2001.
ture variations, were investigated and experimentally validated [21] J. Ahmed and Z. Salam, “An enhanced adaptive P&O MPPT for fast and
not to affect the proposed FZPO performance. efficient tracking under varying environmental conditions,” IEEE Trans.
Sustain. Energy, vol. 9, no. 3, pp. 1487–1496, Jul. 2018.
R EFERENCES [22] A. Pandey, N. Dasgupta, and A. K. Mukerjee, “High-performance
algorithms for drift avoidance and fast tracking in solar MPPT sys-
[1] H. A. Sher, K. E. Addoweesh, and K. Al-Haddad, “An efficient and cost- tem,” IEEE Trans. Energy Convers., vol. 23, no. 2, pp. 681–689,
effective hybrid MPPT method for a photovoltaic flyback microinverter,” Jun. 2008.
IEEE Trans. Sustain. Energy, vol. 9, no. 3, pp. 1137–1144, Jul. 2018. [23] F. Liu, S. Duan, F. Liu, B. Liu, and Y. Kang, “A variable step size INC
[2] G. Velasco-Quesada, F. Guinjoan-Gispert, R. Pique-Lopez, MPPT method for PV systems,” IEEE Trans. Ind. Electron., vol. 55,
M. Roman-Lumbreras, and A. Conesa-Roca, “Electrical PV array no. 7, pp. 2622–2628, Jul. 2008.
reconfiguration strategy for energy extraction improvement in grid- [24] S. K. Kollimalla and M. K. Mishra, “Variable perturbation size adaptive
connected PV systems,” IEEE Trans. Ind. Electron., vol. 56, no. 11, P&O MPPT algorithm for sudden changes in irradiance,” IEEE Trans.
pp. 4319–4331, Nov. 2009. Sustain. Energy, vol. 5, no. 3, pp. 718–728, Jul. 2014.
[3] O. Khan and W. Xiao, “An efficient modeling technique to simulate [25] F. Paz and M. Ordonez, “Zero oscillation and irradiance slope tracking
and control submodule-integrated PV system for single-phase grid for photovoltaic MPPT,” IEEE Trans. Ind. Electron., vol. 61, no. 11,
connection,” IEEE Trans. Sustain. Energy, vol. 7, no. 1, pp. 96–107, pp. 6138–6147, Nov. 2014.
Jan. 2016. [26] Overall Efficiency of Grid Connected Photovoltaic Inverters, European
[4] S. M. R. Tousi, M. H. Moradi, N. S. Basir, and M. Nemati, “A function- Standard EN-50530, Apr. 2010.
based maximum power point tracking method for photovoltaic systems,” [27] S. Jain and V. Agarwal, “A new algorithm for rapid tracking of
IEEE Trans. Power Electron., vol. 31, no. 3, pp. 2120–2128, Mar. 2016. approximate maximum power point in photovoltaic systems,” IEEE
[5] J. Kivimäki, S. Kolesnik, M. Sitbon, T. Suntio, and A. Kuperman, Power Electron Lett., vol. 2, no. 1, pp. 16–19, Mar. 2004.
“Design guidelines for multiloop perturbative maximum power point [28] X. Li, H. Wen, Y. Hu, L. Jiang, and W. Xiao, “Modified beta algorithm
tracking algorithms,” IEEE Trans. Power Electron., vol. 33, no. 2, for GMPPT and partial shading detection in photovoltaic systems,” IEEE
pp. 1284–1293, Feb. 2018. Trans. Power Electron., vol. 33, no. 3, pp. 2172–2186, Mar. 2018.
[6] A. F. Murtaza, M. Chiaberge, F. Spertino, J. Ahmad, and A. Ciocia, [29] X. Li, H. Wen, L. Jiang, Y. Hu, and C. Zhao, “An improved beta method
“A direct PWM voltage controller of MPPT & sizing of DC loads with autoscaling factor for photovoltaic system,” IEEE Trans. Ind. Appl.,
for photovoltaic system,” IEEE Trans. Energy Convers., vol. 33, no. 3, vol. 52, no. 5, pp. 4281–4291, Sep. 2016.
pp. 991–1001, Sep. 2018. [30] T.-W. Hsu, H.-H. Wu, D.-L. Tsai, and C.-L. Wei, “Photovoltaic energy
[7] S. Selvakumar, M. Madhusmita, C. Koodalsamy, S. P. Simon, and harvester with fractional open-circuit voltage based maximum power
Y. R. Sood, “High-speed maximum power point tracking module for point tracking circuit,” IEEE Trans. Circuits Syst. II, Exp. Briefs, vol. 66,
PV systems,” IEEE Trans. Ind. Electron., vol. 66, no. 2, pp. 1119–1129, no. 2, pp. 257–261, Feb. 2019.
Feb. 2019. [31] N. Swaminathan, N. L. Narasamma, M. Kumaravel, and
[8] H. Li, D. Yang, W. Su, J. Lü, and X. Yu, “An overall distribution particle A. Jhunjhunwala, “A novel zonal based MPPT control scheme for a
swarm optimization MPPT algorithm for photovoltaic system under full bridge series resonant converter,” in Proc. IEEE 43rd Photovoltaic
partial shading,” IEEE Trans. Ind. Electron., vol. 66, no. 1, pp. 265–275, Spec. Conf. (PVSC), Jun. 2016, pp. 3263–3268.
Jan. 2019. [32] M. J. Z. Zadeh and S. H. Fathi, “A new approach for photovoltaic
[9] J. Ahmed and Z. Salam, “An improved method to predict the position arrays modeling and maximum power point estimation in real operating
of maximum power point during partial shading for PV arrays,” IEEE conditions,” IEEE Trans. Ind. Electron., vol. 64, no. 12, pp. 9334–9343,
Trans. Ind. Informat., vol. 11, no. 6, pp. 1378–1387, Dec. 2015. Dec. 2017.
[10] I. Shams, D. S. Mekhilef, and T. K. Soon, “Maximum power point [33] E. Karatepe and T. Hiyama, “Artificial neural network-polar coordinated
tracking using modified butterfly optimization algorithm for partial fuzzy controller based maximum power point tracking control under
shading, uniform shading and fast varying load conditions,” IEEE Trans. partially shaded conditions,” IET Renew. Power Generat., vol. 3, no. 2,
Power Electron., vol. 36, no. 5, pp. 5569–5581, May 2020. pp. 239–253, Jun. 2009.
[11] K. Ishaque, Z. Salam, M. Amjad, and S. Mekhilef, “An improved [34] A. Lashab, D. Sera, and J. M. Guerrero, “A dual-discrete model
particle swarm optimization (PSO)–based MPPT for PV with reduced predictive control-based MPPT for PV systems,” IEEE Trans. Power
steady-state oscillation,” IEEE Trans. Power Electron., vol. 27, no. 8, Electron., vol. 34, no. 10, pp. 9686–9697, Oct. 2019.
pp. 3627–3638, Aug. 2012. [35] P. E. Kakosimos, A. G. Kladas, and S. N. Manias, “Fast photovoltaic-
[12] D. Sera, R. Teodorescu, J. Hantschel, and M. Knoll, “Optimized max- system voltage- or current-oriented MPPT employing a predictive digital
imum power point tracker for fast-changing environmental conditions,” current-controlled converter,” IEEE Trans. Ind. Electron., vol. 60, no. 12,
IEEE Trans. Ind. Electron., vol. 55, no. 7, pp. 2629–2637, Jul. 2008. pp. 5673–5685, Dec. 2013.
[13] O. Lopez-Santos et al., “Analysis, design, and implementation of a [36] M. Rakhshan, N. Vafamand, M.-H. Khooban, and F. Blaabjerg, “Max-
static conductance-based MPPT method,” IEEE Trans. Power Electron., imum power point tracking control of photovoltaic systems: A polyno-
vol. 34, no. 2, pp. 1960–1979, Feb. 2019. mial fuzzy model-based approach,” IEEE J. Emerg. Sel. Topics Power
[14] G. Escobar, S. Pettersson, C. N. M. Ho, and R. Rico-Camacho, “Mul- Electron., vol. 6, no. 1, pp. 292–299, Mar. 2018.
tisampling maximum power point tracker (MS-MPPT) to compensate [37] L. Cristaldi, M. Faifer, M. Rossi, and S. Toscani, “MPPT definition and
irradiance and temperature changes,” IEEE Trans. Sustain. Energy, validation: A new model-based approach,” in Proc. IEEE Int. Instrum.
vol. 8, no. 3, pp. 1096–1105, Jul. 2017. Meas. Technol. Conf., May 2012, pp. 594–599.
[15] M. Killi and S. Samanta, “Modified perturb and observe MPPT algo- [38] L. V. Hartmann, M. A. Vitorino, M. B. D. R. Correa, and
rithm for drift avoidance in photovoltaic systems,” IEEE Trans. Ind. A. M. N. Lima, “Combining model-based and heuristic techniques for
Electron., vol. 62, no. 9, pp. 5549–5559, Sep. 2015. fast tracking the maximum-power point of photovoltaic systems,” IEEE
[16] M. Lasheen, A. K. A. Rahman, M. Abdel-Salam, and S. Ookawara, Trans. Power Electron., vol. 28, no. 6, pp. 2875–2885, Jun. 2013.
“Adaptive reference voltage-based MPPT technique for PV applica- [39] H. Tian, F. Mancilla-David, K. Ellis, E. Muljadi, and
tions,” IET Renew. Power Gener., vol. 11, no. 5, pp. 715–722, Apr. 2017. P. Jenkins, “Detailed performance model for photovoltaic systems,”
[17] I. G. Zurbriggen and M. Ordonez, “PV energy harvesting under Nat. Renew. Energy Lab. (NREL), Golden, CO, USA, Tech.
extremely fast changing irradiance: State-plane direct MPPT,” IEEE Rep. NREL/JA-5500-54601, Jul. 2012. [Online]. Available:
Trans. Ind. Electron., vol. 66, no. 3, pp. 1852–1861, Mar. 2019. https://www.nrel.gov/docs/fy12osti/54601.pdf
[18] M. T. Azary, M. Sabahi, E. Babaei, and F. A. A. Meinagh, “Modified [40] N. Swaminathan and L. Narasamma, “Hybrid control scheme for miti-
single-phase single-stage grid-tied flying inductor inverter with MPPT gating the inherent DC-current in the transformer in buck-boost full-
and suppressed leakage current,” IEEE Trans. Ind. Electron., vol. 65, bridge converter for an all-electric motor drive system,” IET Power
no. 1, pp. 221–231, Jan. 2018. Electron., vol. 11, no. 8, pp. 1452–1462, Jul. 2018.

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.
374 IEEE JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER ELECTRONICS, VOL. 10, NO. 1, FEBRUARY 2022

Niraja Swaminathan (Member, IEEE) received the Yue Cao (Member, IEEE) received the B.S. degree
B.E. degree in electrical and electronics engineering (Hons.) in electrical engineering with a second
from Anna University, Chennai, India, in 2012, major in mathematics from The University of
and the dual M.S. and Ph.D. degrees in electri- Tennessee, Knoxville, TN, USA, in 2011, and the
cal engineering from the Indian Institute of Tech- M.S. and Ph.D. degrees in electrical engineering
nology Madras (IIT Madras), Chennai, in 2019, from the University of Illinois at Urbana–Champaign
respectively. (UIUC), Champaign, IL, USA, in 2013 and 2017,
Before joining her Ph.D., she was a Project Asso- respectively.
ciate with the Decentralized Solar PV System Team, He is currently an Assistant Professor with the
IIT Madras. She is currently a Post-Doctoral Scholar Energy Systems Group, Oregon State University
with the Energy Systems Group, Oregon State Uni- (OSU), Corvallis, OR, USA. Before joining OSU,
versity (OSU), Corvallis, OR, USA. Her current research interests include he was a Research Scientist with the Propulsions Team, Amazon Prime
power converters for renewable energy and aviation applications, digital Air, Seattle, WA, USA. He was a Power Electronics Engineer Intern with
control for power converters, maximum power point tracking for solar PV, the Special Projects Group, Apple Inc., Cupertino, CA, USA; Halliburton
and energy-efficient green buildings. Company, Houston, TX, USA; Flanders Electric, Evansville, IN, USA; and
Dr. Swaminathan was a recipient of the SERIIUS—MAGEEP Award to pur- Oak Ridge National Laboratory, Oak Ridge, TN, USA. In 2016, he was a
sue collaborative work with a SERIIUS partner organization in 2014, and was Sundaram Seshu Fellow at UIUC, where he was also a James M. Henderson
awarded the 13th rank among 5418 candidates who graduated B.E. in 2012, Fellow in 2012. His research interests include power electronics, motor drives,
for her academic excellence. She has been an invited reviewer of several IEEE and energy storage with applications in renewable energy integration and
and IET transactions and was the Session Chair of ECCE 2020 conference. transportation electrification.
She helped to establish a joint IEEE PES/PELS Chapter at OSU in 2020. Dr. Cao is a Board Member and the Award Chair of IEEE Power Electronics
Society (PELS) TC11–Aerospace Power. He received the Myron Zucker
N. Lakshminarasamma (Member, IEEE) received Award from the IEEE Industry Applications Society (IAS) in 2010. He was
the Ph.D. degree in electrical engineering from the a recipient of the Oregon State Learning Innovation Award for transformative
Indian Institute of Science, Bengaluru, India, in education in 2020. He served as the Corresponding Technical Programs
2007. Chair of the 2016 IEEE Power and Energy Conference at Illinois (PECI).
Since 2009, she has been a faculty with the Depart- He is currently the Tutorials Chair of the 2021 IEEE Energy Conversion
ment of Electrical Engineering, Indian Institute of Congress Expo (ECCE) and the Special Sessions Chair of 2022 ECCE.
Technology Madras, Chennai, India. Prior to this, In 2020, he helped to establish an IEEE PELS Chapter at OSU. He is
she has put in four years in academics; she has currently an Associate Editor of IEEE T RANSACTIONS ON T RANSPORTATION
worked as a Software Engineer at I2 Technologies E LECTRIFICATION.
India Pvt., Ltd., Bengaluru, and as a Systems Spe-
cialist at GE Healthcare India Ltd., Bengaluru. She
has coauthored several journal articles in peer-reviewed journals, including
the IEEE T RANSACTIONS ON P OWER E LECTRONICS and several premier
conferences. Her areas of interest include power electronics, switched-mode
power conversion, and renewable energy systems.

Authorized licensed use limited to: Institut Teknologi Sepuluh Nopember. Downloaded on March 28,2024 at 18:24:44 UTC from IEEE Xplore. Restrictions apply.