Page 1 of 8
Journal for Studies in Management and Planning
Available at http://edupediapublications.org/journals/index.php/JSMaP/
ISSN: 2395-0463
Volume 04 Issue 03
March 2018
Available online: http://edupediapublications.org/journals/index.php/JSMaP/ P a g e | 173
Fuzzy-Logic-Controller-Based SEPIC Converter for MPPT in
Standalone PV systems
Challa Venkatesh1,P. Raghava Rani 2,
1 M.Tech, Department of EEE, sai tirumala NVR engineering college, A.P, India
2.Assoc.professor, Department of EEE, sai tirumala NVR engineering college, A.P, India
Abstract - This paper presents a fuzzy controller (FC)-
based single-ended primary-inductor converter (SEPIC)
for maximum power point tracking (MPPT) operation of a
photovoltaic (PV) system along with battery. The FLC
proposed scheme uses the convergent distribution of the
membership function. The fuzzy controller for the SEPIC
MPPT scheme shows the voltage without any changes in
different load conditions at the inverter output (load) side.
The behaviour of the converter is tested in simulation at
different operating conditions. The load is fed from the
battery storage continuously with constant voltage. The
battery will be charge with the help of PV module and the
SEPIC converter, which is controlled by FLC-based MPPT.
The proposed FLC-based MPPT with battery will supply
more power to the load than the without battery system.
Key Words: SEPIC converter , fuzzy controller, photovoltaic(PV)
modules, battery ,inverter.
I.INTRODUCTION
The single-ended primary inductor converter (SEPIC) acts as a buck–boost dc–dc converter, where it changes its output
voltage according to its duty cycle.The selection of a proper dc–dc converter plays an important role for maximum power
point tracking (MPPT) operation. Due to its output gain flexibility. Among known converters, the SEPIC, conventional buck–
boost, and Cuk converters have the ability to step up and step down the input voltage. Hence, this converter can transfer
energy for all irradiation levels. Another desirable feature is continuous output current, which allows converter output
parallel connection, or conversion to a voltage source with minimal capacitance. The buck or boost converters are not
preferable, due to the lack of output voltage flexibility.
The SEPIC is chosen because the output voltage can be higher or lower than the input voltage. Also the input and output
voltages are dc isolated. The isolation is provided by the
series capacitor c, which blocks the dc from the supply side to the output side[1]. An auxiliary switch and a clamp capacitor
are connected. A coupled inductor and an auxiliary inductor are utilized to obtain ripple-free input current. The voltage
multiplier technique and active clamp technique are applied to the conventional SEPIC converter to increase the voltage
gain, reduce the voltage stresses of the power switches and diode. Moreover, by utilizing the resonance between the
resonant inductor and the capacitor in the voltage multiplier circuit, the zero-current-switching operation of the output
diode is achieved and its reverse-recovery loss is significantly reduced. Both the SEPIC and the Cuk converter provide the
choice to have either higher or lower output voltage compared to the input voltage. The MPPT algorithm represents
optimal load for PV array, producing opportune voltage for the load. SEPIC converters can have a low input current ripple,
which is one of the advantages of SEPIC converters. However, a bulk inductor should be used to minimize the current
ripple. Input current ripple becomes one of important requirements due to the wide use of low voltage sources such as
batteries, super capacitors, and fuel cells. The PV panel yields exponential curves for current and voltage, where the
maximum power occurs at the curve’s mutual knee. The applied MPPT uses a type of control and logic to look for the
knee, which in turn allows the SEPIC converter to extract the maximum power from the PV array. The tracking method
used, i.e., perturb and observe (P&O). A tracking method based on parabolic function is proposed to perform the
photovoltaic maximum power point tracking. With the proposed method, the maximum power calculation is made from a
parabolic convex function. Then a systematic scheme is developed to adjust the concavity and optimal region of the
approximate parabola for ensuring the iterative convergence of the proposed method. In order to confirm the effectiveness of
this proposed design, the approach has been applied to investigate different atmospheric scenarios. Among different
intelligent controllers, fuzzy logic is the simplest to integrate with the system. Recently, the fuzzy logic controller (FLC) has
Page 2 of 8
Journal for Studies in Management and Planning
Available at http://edupediapublications.org/journals/index.php/JSMaP/
ISSN: 2395-0463
Volume 04 Issue 03
March 2018
Available online: http://edupediapublications.org/journals/index.php/JSMaP/ P a g e | 174
received an increasing attention to researchers for converter control, motor drives, and other
process control because it provides better responses than other conventional controllers. The imprecision of the weather
variations that can be reflected by PV arrays can be addressed accurately using a fuzzy controller. In order to take the
advantages of the fuzzy logic algorithm, the MPPT algorithm is integrated with the FLC so that the overall control system
can always provide maximum power transfer from the PV array to the inverter side, in spite of the unpredictable weather
conditions.
II. PROPOSED SYSTEM
Fig:2 overall control scheme for the proposed FLC based MPPT scheme for the SEPIC converter with battery.
In this paper, the voltage level increases or decreases depending on the maximum power. Furthermore, the controller
changes the voltage level by changing the duty cycle of the pulsewidth-modulated (PWM) signal, which tracks the reference
signal. A sinusoidal reference signal is compared with the output signal to produce a supposedly zero error signal. Another
reference signal is used to compare the SEPIC’s output, to achieve the maximum power. This reference signal is adaptive,
changing its shape according to weather conditions.
Fig. 1. Circuit diagram of the SEPIC converter for the FLC-based MPPT with battery scheme.
Fig. 1 is the circuit diagram of the SEPIC dc–dc converter together with the MPPT and the fuzzy controller with battery. The
design of the fuzzy controller was done using Mamdani’s method for the converter. The maximum power point can be achieved
in case of a grid-connected system, a full-load condition, or using battery charging in case of a standalone system. However,
if the load need is lower than PV capacity, the PV voltage will move right in the PV curve, achieving the opportune power. This
case happens even if the batteries of the standalone system are full and the load is lower than PV power. In grid-connected
systems, the load is always there due to the huge number of clients. Therefore, the maximum power point can always be
achieved subject to the load need.
III. FLC ALGORITHM
In FLC design, one should identify the main control variables and determine the sets that describe the values of each
linguistic variable. The input variables of the FLC are the output voltage error e(n) and the change of this error e_(n). The
output of the FLC is the duty cycle of d(n) of the PWM signal, which regulates the output voltage.
Fig. 3. Unsymmetrical focused membership function of the proposed FLC:
(a) e(n), (b) e_(n), and (c) d(n).
Fig.3 show the membership functions of the inputs and the outputs of the SEPIC-side FLCs. The triangular membership
Page 3 of 8
Journal for Studies in Management and Planning
Available at http://edupediapublications.org/journals/index.php/JSMaP/
ISSN: 2395-0463
Volume 04 Issue 03
March 2018
Available online: http://edupediapublications.org/journals/index.php/JSMaP/ P a g e | 175
functions are used for the FLC for easier computation. A five- term fuzzy set, i.e., negative big (N-II), negative small (N-I),
zero (Z), positive small (P-I), and positive big (P-II), is defined to describe each linguistic variable.
TABLE-I
FUZZYRULE-BASED
The fuzzy rules of the proposed PV SEPIC dc–dc converter can be represented in
a symmetric form, as shown in Table I. Moreover, as in Fig.5, the membership
functions of the output variables are nine term fuzzy sets with classical
triangular shapes, i.e., negative very big (N4), negative big (N3), negative small
(N2), negative very small (N1), zero (Z), positive very small (P1), positive small
(P2), positive big (P3), and positive very big (P4). The Mamdani fuzzy inference
method is used for the proposed FLC, where the maximum of minimum
composition technique is used for the inference and the center-of-gravity
method is used for the defuzzification process. Fig. 3 illustrates a focused
membership function, where the sets go toward zero. the membership functions in Fig. 3 are guaranteed to produce the
stable output signal. The design of the focused membership function values depends on the nature of the signal.
Fig. 4. Three-dimensional surface corresponding to the memberships in Fig. 3 and the rules in Table I.
The control signal value is confined between −1 and 1,owing to the PWM carrier wave. The input signal values are between
−100 and 100 because of the error signal, which is resultant from the difference between the output signal and the desired
reference signal. In addition, most of error values are centered from −20 to 20. The sharpness of the control signal is very
essential for minimizing the error signal to zero in short time; wherefore, the pulse membership function is used to configure
the control signal fuzzy sets. The FLC performance changes with unsymmetrical distribution of membership functions, where
both convergent and divergent types of asymmetry will be considered with varying degrees of the unsymmetrical membership
functions.
IV. PROPOSED MPPT-BASED SEPIC CONVERTER
The fuzzy controller is applied to the SEPIC converter to mimic the new reference signal coming from the MPPT. The
new duty cycle δ(k) of the SEPIC converter switch was adjusted either by adding or by subtracting the previous duty cycle
δ(k−1) with the duty cycle’s perturbation step size. Equation (1) presents the relation between the present and previous
duty cycles, i.e.,
δ(k) = δ(k−1) Δδ (1)
e'/e N-II N-I Z P-I P-II
N-II Z4 Z4 Z4 Z3 Z
N-I Z4 Z2 Z1 Z3 P3
Z Z4 Z1 Z P1 P4
P-I Z3 Z P1 P2 P4
P-II Z P3 P3 P3 P4
