Solar Energy 178 (2019) 79–89

Contents lists available at ScienceDirect

Solar Energy
journal homepage: www.elsevier.com/locate/solener

Design of battery charging circuit through intelligent MPPT using SPV T

system
⁎
Pawan Kumar Pathaka, , Anil Kumar Yadavb
a
School of Automation, Banasthali Vidyapith, Rajasthan 304022, India
b
Department of Electrical Engineering, National Institute of Technology, Hamirpur, HP 177005, India

A R T I C LE I N FO A B S T R A C T

Keywords: In a solar photovoltaic (SPV) based hybrid renewable energy system, batteries are used as a power reservoir. SPV
Fuzzy logic based DPID (FL-DPID) MPPT system provides energy under steady operating condition whereas SPV along with batteries serve as the source of
algorithm energy under transient operating condition. This paper puts forward the design of a battery charging circuit
Battery charging circuit through an intelligent fuzzy logic based discrete proportional-integral-derivative (FL-DPID) maximum power
Optimal PID (O-PID) controlled buck converter
point tracking (MPPT) algorithm. SPV system in conjunction with FL-DPID MPPT technique driven DC-DC boost
converter enhances the output voltage besides tracking maximum power point (MPP) under varying irradiance
in between 400–1000 W/m2 and a constant temperature of 25 °C. The output voltage of the boost converter
drives the optimal PID (O-PID) controlled buck converter to behave as a battery charging circuit under non-
deterministic atmospheric conditions. The objective of this study is to operate the designed SPV system at MPP
under varying environmental condition in order to achieve higher efficacy, minimize overall system cost and
obtain apropos voltage and current for effective charging of battery thereby reducing battery losses and en-
hancing life cycle. A 200 W prototype of an SPV panel has been designed, simulated and investigated in the
MATLAB/Simulink environment.

1. Introduction whereas it is being widely employed in the battery-powered electric

vehicle (Larminie and Lowry, 2003; Wirasingha and Pihef, 2011) and as
Electricity act as a paramount factor in the commercial growth of a a backup source in the hybrid renewable energy system (Fakham et al.,
nation. The transition from traditional to the mechanized world has 2011).
created a black hole of energy with nearly 1.3 billion sapiens round the The SPV system is characterized by nonlinear P-V and I-V curves
globe not even having regular access to the electricity (World energy (Cheng et al., 2015) and it greatly relies on atmospheric conditions
outlook, 2013; Universal access to modern energy for poor). Around 2.7 namely solar irradiance and temperature. Change in irradiance and
billion people still rely on long-established bio-energy like wood, crop temperature alter the voltage and current thereby constantly changing
residues, animal dung etc. as their prime energy source for cooking and the maximum available power of SPV panel. Hence the conversion ef-
heating (Ekouevi and Tuntivate, 2012; Weimann et al., 2013). In de- ficacy of the system becomes as low as 7–19% (Haq et al., 2013) and
veloping countries, the marginal population faces adverse effects of fabrication cost shoots up. To overcome this setback SPV system is
power crunch most significantly on their health and economic growth. made to operate at MPP under changing atmospheric conditions
Many such regions have an abundant source of renewable energy such (Agarwal and Patel (2010); Metin et al., 2007; Soto et al., 2006; Graditi
as solar, wind, and bio-mass. In recent years due to the energy crisis and and Adinolfi, 2011; Graditi et al., 2014). To operate SPV panel at MPP
environmental concern countries around the world is focusing on green various MPPT techniques have been studied and developed so far such
technology where solar energy has grown as one of the most potent as perturb and observe (P&O) (Elgendy et al., 2012), incremental
form of renewable energy. SPV system is used in conjunction with conductance (IC) (Elgendy et al., 2013) and fuzzy logic (FL) based
batteries which act as energy soak pit and are utilized under the tran- MPPT (Algazar et al., 2012; EI-Khatib et al., 2017; Chen et al., 2016,
sient condition when the SPV system alone fails to generate the re- 2015; Rezk and Eltamaly, 2015; Al-Majidi et al., 2018). MPPT algo-
quired power. In remote and rural areas batteries are considered as rithm utilizes the output of SPV system in terms of voltage and current
most potent and pocket-friendly source for home lightning (Pode, 2015) to generate PWM gating signal for power switch of the boost converter

⁎
Corresponding author.
E-mail addresses: ppathak999@gmail.com (P.K. Pathak), anilei007@gmail.com (A.K. Yadav).

https://doi.org/10.1016/j.solener.2018.12.018
Received 15 July 2018; Received in revised form 29 November 2018; Accepted 7 December 2018
0038-092X/ © 2018 Elsevier Ltd. All rights reserved.
P.K. Pathak, A.K. Yadav Solar Energy 178 (2019) 79–89

to modulate its output (Fathabadi, 2016; Singh et al., 2018; Babaei ZN-PI and ZN-PID controllers are to improve the performances indices
et al., 2017; Hart, 2011; Adinolfi et al., 2015; Graditi et al., 2010). such as rise time (RT), settling time (ST), overshoot (OS), integral of the
Constant voltage and constant currents are the mostly used methods to absolute error (IAE) and integral of the square of error (ISE) that re-
charge a battery (Borage et al., 2006; Cho et al., 2013; Chen and Lai, duces the charging time of the battery thereby enhancing its life. The
2012). To render invariable current and voltage to the connected load key contributions of the paper are summarized as follows: (1) The de-
PID controlled buck converter is employed as the charging circuit. The sign and implementation of intelligent i.e. FL-DPID MPPPT technique
PID controller is widely used in various industrial applications such as for battery charging circuit under varying solar irradiance i.e.
electric vehicles, process control, and power system because of its in- 400–1000 W/m2 and the results are compared with existing P&O and IC
expensiveness, ease in designing and excellent performance (Yadav and MPPT techniques. (2) A detailed comparative analysis among designed
Gaur, 2016a,b; Ang et al., 2005; Neath et al., 2014). Yilmaz et al. ZN-PI, ZN-PID and O-PID controllers of buck converter for charging of a
(2018) proposed FL based SPV system for battery charging circuit under 18 V battery. The results are compared with the literature (Yilmaz et al.,
different atmospheric conditions and analyzed the system response. 2018). The transient response and ripples in output voltage and current
Eldahab et al. (2016) develop a novel MPPT technique for SPV based of the boost converter are considered as the key factors for a compre-
battery charging controller. The prominent feature of this novel MPPT hensive comparison of designed P&O, IC, and FL-DPID MPPT techni-
technique is remote monitoring and controlling of various system ele- ques.
ments. Mirzaei et al. (2017) designed a control topology for power The rest of the work is organized as follows. Section 2 deals with
management in a standalone hybrid system. Pode (2015) details battery design considerations of one diode model of a crystalline solar cell and
charging stations which are quite welcomed in lonesome which are yet SPV module along with its characteristics under varying atmospheric
not connected to the grid. The implementation of digital control condition. In Section 3 the design methodology of P&O, IC, and FL-
strategy through DSP has been done by Lopez et al. (2016) for buck DPID MPPT techniques are presented. The design of DC-DC power
converter charging a battery through solar generated power under converter covers the power conversion system including both DC-DC
various atmospheric conditions. In Rajani and Pandya et al. (2016) boost and buck converter employing control strategies such as ZN-PI,
MPPT based SPV system is utilized for charging battery and ultra-ca- ZN-PID, and O-PID controllers are discussed in section 4. The obtained
pacitor individually, and it depicts the enhancement in the charging results have been vividly discussed in Section 5 i.e. results and dis-
rate as compared to non MPPT charging. cussion, and the concluding remarks of this research are presented in
The structure of FL-DPID MPPT technique is proposed and im- Section 6.
plemented for battery charging circuit under varying solar irradiance in
between 400–1000 W/m2. The proposed MPPT technique has been
designed with a lesser number of rules i.e. nine as compared to existing 2. Description of SPV system
literature (Yilmaz et al., 2018). An FL based system with the least
number of rules reduces the computational time and complexity of the The output and efficacy of the SPV system completely rely on dif-
system. The output voltage of FL-DPID MPPT driven boost converter ferent array configuration as well as various atmospheric conditions
having least ripple drives the O-PID controlled buck converter to deliver such as non-uniform solar insolation and varying the environmental
regulated power to the battery as a load with constant voltage and temperature. The P-V and I-V characteristics of an SPV system for a
constant current irrespective of change in solar irradiance as shown in constant environmental temperature of 25 °C and varying solar insola-
Fig. 1. The parameters of the proposed O-PID controller are obtained tion have been depicted in Fig. 2(a) and (b), whilst Fig. 3(a) and (b)
utilizing a Genetic Algorithm (GA) with suitable objective function. The represents the P-V and I-V characteristics of an SPV subjected to varying
results of O-PID controlled buck converter are compared with Ziegle- temperature keeping insolation constant at 1000 W/m2 respectively.
r–Nichols (ZN) tuned PI and PID controllers and named as ZN-PI and The effect of a change in temperature as seen from Figs. 2 and 3 has
ZN-PID respectively. The expected advantages of O-PID controller over a lesser influence on the characteristic of the SPV system as compared
to the change in solar insolation. Therefore, in this paper, the work is

D u Optimal e Ref. Value

PID +
Controller -
Actual Value

PWM
Boost Converter Generator 2

Ipv
PV Array

Vpv Battery

Vdc Vdcb

Buck Converter

Ipv U D

Vpv
FL-DPID MPPT PWM
Generator 1
Fig. 1. SPV system based battery charging circuit.

80
P.K. Pathak, A.K. Yadav Solar Energy 178 (2019) 79–89

200 9
1000 W/m2
8
800 W/m2
150 7
600 W/m2

Current (A)
Power (W) 400 W/m2 6

100 5
4
3
50 1000 W/m2
2 800 W/m2
600 W/m2
1
400 W/m2
0 5 10 15 20 25 30 34
0 5 10 15 20 25 30 34
Voltage (V) Voltage (V)
(a) (b)
Fig. 2. (a) P-V curve and (b) I-V curve on different solar irradiance.

focused on constant solar temperature i.e. 25 °C and varying solar in- Rs I

solation in range of 400–1000 W/m2 for the SPV system based battery
charging circuit.
Id Ish
2.1. Mathematical equation of a solar cell

D Rsh V
The simplified electrical circuit of each solar cell can be modeled in
various ways out of which one diode model is considered in the pre-
sented work. Iph
A simplified circuit for the same is represented in the Fig. 4 and the
I–V relation is given below:
V + Rs I
{
I = Iph − I0 e
q (V + Rs I )
AKT −1 − } Rsh (1) Fig. 4. Equivalent circuit of a solar cell.

where Iph denotes photocurrent, Id is the current through diode, theIsh

{ }
q (V + Rs I )
denotes shunt current, Rsh denotes shunt resistance, Rs denotes series I = ISC − IO e Ns AKT −1
connected resistance, I denotes the net output current of SPV, V de- (3)
notes the voltage across SPV, I0 denotes diode reverse saturation cur- Under the open-circuit condition of a PV module, I = 0 and hence
rent, q denotes charge of the electron, A is curve fitting factor andK is the term q in (3) will be written as follows:
the Boltzmann constant (1.38 × 10−23 J/K).
Ns AKT

q
=
ln ISC + 1
O ( I
)
2.2. Modeling of an SPV module Ns AKT VOC (4)

The last term of (1) can be omitted due to Rsh is assumed to be where VOC denotes module open-circuit voltage. The current can be
infinite and the slope of I-V curve is zero at short circuit condition. obtained from (3) and (4) as:
Further replacing Iph by ISC (short-circuit current): I (V + RS I )
I ln ⎛ SC + 1⎞
I = ISC ⎧1 − O ⎜⎛e ⎝ IO ⎠ VOC − 1⎟⎞ ⎫
I = ISC − IO { q (V + Rs I )
e AKT −1 } (2)
⎨
⎩ ISC ⎝ ⎠⎬
⎭ (5)

q (V + R s I ) where k = ISC / IO and after solving (5), obtained (6):

In a PV module the appropriate application of (2) is mod-
AKT
ified and substituted by q (V + Rs I ) , where Ns denotes total solar cells 1 (V + RS I ) 1
Ns AKT I = ISC ⎧1 − (k + 1) VOC + ⎫
connected back to back in a crystalline type SPV module. Yields (3); ⎨
⎩ k k⎬⎭ (6)

9
200 8
25 degC
50 degC
150 6
75 degC
Current (A)
Power (W)

100 4

50 25 degC
2
50 degC
75 degC
0
0 5 10 15 20 25 30 34 0 5 10 15 20 25 30 34
Voltage (V) Voltage (V)
(a) (b)
Fig. 3. (a) P-V curve and (b) I-V curve on different temperature.

81
P.K. Pathak, A.K. Yadav Solar Energy 178 (2019) 79–89

P&O algorithm:
Step 3: Calculate power: P (n) = V (n) ∗ I (n) .
PMPP Step 4: Call previous values of P and V from the memory i.e.P (n − 1) andV (n − 1) .
Step 5: Calculate the change in power ‘dP ’ and change in voltage‘dV ’ using:
dV = V (n) − V (n − 1) and dP = P (n) − P (n − 1) .
Step 6: If dP = 0 , Then no change in duty ratio is required and GOTO Step 7.
else If (dP ∗ dV ) > 0 , Then increase the duty ratio by ΔD and GOTO Step 7.
Current Power else decrease the duty ratio by ΔD and GOTO Step 7.
I SC Step 7: Return.
I MP
Maximum Power Point 3.2. Incremental conductance (IC)
Current (A)

(MPP)
The P&O technique fails under rapidly changing environmental
conditions; this can be overcome by using the IC technique. The slope of
the P-V curve of an SPV array is the basis of the IC algorithm. The
0 derivative of the output power of the SPV array is written as follows:
0 Voltage (V) VMPP VOC dP d (IV ) dI ΔI
= =I+V =I+V
dV dV dV ΔV (8)
Fig. 5. P-V and I-V curve for MPP.
The entire IC algorithm is depicted as follows:

Usually, k is quite high as ISC is much greater than IO . Few terms in IC algorithm:
(6) can be ignored for simplified I–V relation which is represented as Step 1: Start
Step 2: Read variables V (n) andI (n) .
follows:
Step 3: Call previous values of I and V from the memory i.e. I (n − 1) and V (n − 1) .
(V + RS I )
− 1⎞
Step 4: Calculate the change in current‘dI ’ and change in voltage‘dV ’ using:
I = ISC ⎛1 − k ⎜ VOC ⎟
dV = V (n) − V (n − 1) and dI = I (n) − I (n − 1) .
⎝ ⎠ (7) Step 5: IfdV = 0 , Then GOTO Step 6.
else GOTO Step 7.
Step 6: If dI = 0 , Then GOTO Step 8
3. Maximum power point tracking (MPPT) techniques else If dI > 0 , Then increase duty ratio by ΔD and GOTO Step 8.
else decrease the duty ratio by ΔD and GOTO Step 8.
The major setback of a commercial SPV system is less conversion Step 7: If
dI
+
I
< ε , Then GOTO Step 8.
dV V
efficiency. Therefore, to enhance the efficacy of the system MPPT al-
dI I
gorithm is employed. The maximum efficiency is expected from an SPV else If >− , Then increase duty ratio by ΔD and GOTO Step 8.
dV V
system when it operates at MPP and as represented in Fig. 5. else decrease the duty ratio by ΔD and GOTO Step 8.
Step 8: Update V (n) andI (n) .
Step 9: Return.
3.1. Perturb & observe (P&O)

P&O is based on perturbation in array voltage. From the P-V curve 3.3. FL-DPID MPPT technique
of an SPV array as shown in Fig. 6 it can be observed that while op-
erating on the left side of MPP increasing/decreasing with increasing/ The proposed FL-DPID is a powerful MPPT algorithm to operate at
decreasing the voltage ΔV , whereas it decreases/increases on the right MPP for SPV system. The advantages of FL-DPID MPPT over P&O and
of MPP. For minimizing the oscillations, the step size of perturbation IC MPPTs are a minimum ripples in the output, the capability to handle
can be reduced. nonlinearity, work with inexplicit inputs, and non-requirement of the
The entire P&O algorithm is depicted as follows: exact mathematical model whereas when compared to FL based MPPT
it reduces implementation and computation complexity. The flowchart
P&O algorithm: and proposed structure of FL-DPID MPPT technique are depicted in
Step 1: Start Fig. 7(a) and (b) respectively. The symbols K e, K ce, Kp and Ki are the
Step 2: Read variables V (n) andI (n) . scaling factors of FL-DPID MPPT controller and are obtained based on
intuition through ZN tuning method (Yadav and Gaur, 2016b).
The error E and change in error CE acts as the input variables of FL-
DPID MPPT and are computed using the output power and voltage of
SPV system as represented by (9) and (10).
ΔP P (n) − P (n − 1)
E (n) = =
ΔV V (n) − V (n − 1) (9)

CE (n) = E (n) − E (n − 1) (10)

The output variable of the FL system is UPD and the accumulated

output of FL-DPID MPPT is UPID as shown in Fig. 7(b). The UPID i.e. U is
acts as a reference and provided as an input to PWM genrator−1 that
modifies the duty cycle D as shown in Fig. 1. From Fig. 7(b), the control
law of FL-DPID MPPT i.e. UPID is written as follows:
1
UPID = Kp·UPD + Ki· ·UPD
1 − Z −1 (11)

The relation between input variable E and output variable UPD of

product-sum crisp type fuzzy is approximated and described as follows
Fig. 6. P-V curve of an SPV array. (Yadav and Gaur, 2016a,b):

82
P.K. Pathak, A.K. Yadav Solar Energy 178 (2019) 79–89

Start

Initialize duty
cycle (D)

Measure V (n)
and I (n)

Calculate
P (n) = V (n)*I (n),
dP and dV

Calculate
Error (E) & Change in
error (CE)

Fuzzification

Rule base Inference Fuzzy set

Defuzzification

Update D

(a)

E UPID
Ke Kp +
UPD +

+ Kce + Ki
- +
Fuzzy Logic
Z-1 Z-1

(b)
Fig. 7. (a) Flowchart and (b) Proposed structure of FL-DPID MPPT technique.

UPD = K e·E + K ce·(1 − Z −1)·E (12) 1

UPID = (Kp K e + K ce Ki )·E + (K e ×Ki )· ·E + (Kp ×K ce ) ·(1 − Z −1)·E
⏟ ⏟
Integralgain
1 − Z −1 ⏟
Proportionalgain Derivativegain
From (11) and (12)
(14)
Ki The final control law (14) represents the generalized equation of
UPID = Kp [K e·E + K ce·(1 − Z −1)·E ] + [K e·E + K ce·(1 − Z −1)·E ]
1 − Z −1 three term PID controller in the discrete domain. Therefore, the pro-
(13) posed structure of FL based MPPT is named as FL-DPID MPPT tech-
nique. The membership functions (MFs) forE and CE , and UPD are
After solving (13), yields (14)
shown in Fig. 8(a), and (b) respectively. The linguistic names of these
MFs are as follows: NB (Negative Big), N (Negative), Z (Zero), P (Po-
sitive) and PB (Positive Big).

83
P.K. Pathak, A.K. Yadav Solar Energy 178 (2019) 79–89

N Z P
1

0.5

-0.1 -0.05 0 0.05 0.1

(a)

NB N Z P PB
1

0.5

-1 -0.33 -0.66 0 0.33 0.66 1

(b)
Fig. 8. MFs for (a) E and CE , and (b)UPD

Table 1 charged with the voltage VL =Vin . During time period Dt < t < T , the
Rule base matrix. MOSFET goes into off state and the diode is in on state due to forward
E P Z N
biased. In this condition the output voltage Vout =Vdc = Vin +VL . In steady
CE state condition the total energy stored by, L must be equal to the energy
released by L in a period of switching. The values of boost converter
P PB (1) P (4) Z (7) parameters L and C are calculated using formulas as follows: (Hart,
Z P (2) Z (5) N (8)
2011)
N Z (3) N (6) NB (9)
Vin
Vout =Vdc =
(1 − D) (15)
The E and CE after being calculated and transmuted to the semantic
variables depending on the MFs generates the output of FL-DPID con- Iout =(1 − D) Iin (16)
troller in terms of UPID which in turn alter the duty cycle ΔD of the boost
converter. The ΔD is deducted from previous value of D and the new Vin ∗ D
L=
value acts as the gating signal that controls the boost converter to track fs ∗ ΔI (17)
the MPP of the SPV system. The crisp output UPD can be obtained from
the rule base matrix as given in Table 1. Iout ∗ D
C=
The 9 rules of Table 1 can be expressed as follows: Rule 6: ‘IF E is N fs ∗ ΔV (18)
and CE is Z THEN UPD is N’. Similarly, the other rules are described and
where C is caapacitor, Iout is the current output of boost converter, ΔI
used for evaluation of final output of FL-DPID controller (14) and ap-
denotes ripple in inductor current, ΔV denotes ripple in output voltage,
propriate ΔD is obtained.
fs denotes switching frequency of the power device and Iin is the input
current of boost converter.
4. Design of DC-DC power converters
4.2. DC-DC buck converter
For maximized power output SPV is made to operate at MPP. To
trace the MPP of SPV the power converter is operated with the corre-
Buck converter has the property of generating an output voltage
sponding D. With the change in solar insolation the D must vary ac-
which is lower in magnitude than the input supply. During 0 < t < DT ,
cordingly in order to track MPP. Various configuration of the DC-DC
the MOSFET turns on whereas diode is in off state and charging in-
converter studied so far out of which boost converter is widely chosen
ductor with a voltage VL = Vdc − Vdcb . MOSFET turns off during
and considered in the presented work due to less complexity and higher
DT < t < T , diode comes in conduction with voltage developed across
reliability as shown in Fig. 1.
the inductor changing to VL = −Vdcb . To render invariable current and
voltage for charging a battery ZN-PI, ZN-PID and O-PID controllers are
4.1. DC-DC boost converter designed to generate a controlled reference signal u for PWM generator-
2 that gives the updated D to the buck converter as shown in Fig. 1. The
Metal-oxide-semiconductor field-effect transistor (MOSFET) is em- parameters of DC-DC buck converter L and C are calculated using
ployed as a switching device for the operation of the boost converter. formulas as given follows: (Hart, 2011)
During time period 0 < t < DT , MOSFET turns on and diode enters into
reverse biased condition. During this time interval inductor L gets Vdcb =D ∗ Vdc (19)

84
P.K. Pathak, A.K. Yadav Solar Energy 178 (2019) 79–89

Vdcb ∗ (1 − D) Table 2
L=
fs ∗ ΔIb (20) Numerical values for simulation.
System Symbols Values
Vdcb ∗ (1 − D)
C=
8 ∗ L ∗ fs ∗ ΔVdcb (21) SPV VOC /module 10.908 V
ISC /module 8.21 A
where Vdcb is the output voltage generated by the buck converter, Vdc Cells in a module 18
denotes input dc supply voltage of the buck converter, ΔIb denotes Ns 3
ripple in inductor current and ΔVdcb is the ripple in output voltage. NP 1
A 1.36
Boost fs 150 kHz
4.3. Control strategy L 0.1075 mH
C 10.41 µF
In this section, the design methodology of classical ZN-PI and ZN- Buck fs 150 kHz
PID controllers, and proposed O-PID controller for voltage regulation L 0.1613 mH
using buck converter is presented. The differential equation of a gen- C 1.744 µF

eralized PID controller is usually represented in ‘parallel form’ or ‘ideal

form’ stated by (22) or (23) respectively.
corresponding to MPP and voltage and current of the boost converter
u (t ) = KP e (t ) + KI ∫ e (t ) dt + KD e ̇ (t ) (22)
under rapidly varying solar irradiation has been done.

1 5.1. Comparative analysis of P&O, IC, and FL-DPID MPPT techniques

u (t ) = KP ⎡e (t ) +
⎢ T
∫ e (t ) dt + TD e ̇ (t )⎤⎥ (23) under varying solar irradiance
⎣ I ⎦
where KP , KI andKD denotes proportional, integral and derivative gains The profile of solar radiation taken into account for the study has
respectively, TI and TD denotes integral and derivative time constants been illustrated in Fig. 9 i.e. varying in between 400–1000 W/m2 (Liu,
respectively, and u and e are output and input i.e. error signal of the et al., 2008). The change in solar irradiance for carrying out the de-
controller respectively. The performance parameters of the PID con- tailed analysis has been considered as trapezoidal in nature divided into
troller are quantified using ZN tuning method (Yadav and Gaur, five states as 600 W/m2 at 25 °C, 800 W/m2 at 25 °C, 1000 W/m2 at
2016b); using relations as given below: 25 °C, 800 W/m2 at 25 °C, and 400 W/m2 at 25 °C. According to the
For PI control: considered solar irradiance profile state 2 and state 4 have been taken
KP = 0.75Ku, TI = Tu/1.2, KI = KP / TI (24) at same irradiance level, thus only state 2 has been considered for the
analysis. The output parameters of boost converter i.e. voltage and
For PID control:
current and maximum obtainable SPV power implementing P&O, IC,
KP = 0.6Ku, TI = Tu/2, TD = Tu/8, KI = KP / TI and KD = KP ·TD (25) and FL-DPID MPPTs have been depicted in Fig. 10(a)–(c) respectively.
The values of scaling factors K e, K ce, Kp and Ki of FL-DPID MPPT
where Ku and Tu are the ultimate gain and period of the system re-
technique are 0.095, 0.007, 0.7 and 0.3 respectively. The output vol-
spectively. Sometimes, the ZN tuned PI and PID controllers give un-
tage corresponding to MPP of SPV system incorporating FL-DPID MPPT
desirable performance in terms of OS, RT, ST, IAE and ISE (Yadav and
algorithm along with the voltage after being boosted by the DC-DC
Gaur, 2016a; Neath et al., 2014). Therefore, the O-PID controller is
boost converter has been represented in Fig. 11. A vivid comparison of
proposed in this paper.
all the three MPPT techniques in terms of efficiency, dynamic perfor-
In O-PID controller, the KP, KI, and KD are obtained using GA for
mance like ST and the effect of ripples on the performance of boost
which the objective function j is formulated using IAE and ISE, and
converter are summarized in Table 3.
given as follows:
From the obtained results as shown in Fig. 10(a)–(c) and the cal-
∞ ∞
j = w1· ∫0 |e (t )| dt + w2· ∫0 e 2 (t ) dt (26)
culated values as given in Table 3 it can be noted that with the rapid
variation in the solar insolation the designed SPV system is efficiently
∞
where w1 and w2 are weights to IAE = ∫ |e (t )| dt and 0 able to trace the MPP in all the considered states. It can be spotted from
∞
ISE = ∫0 e 2 (t ) dt respectively, and equal weights for both IAE and ISE Fig. 10(c) that under steady state condition P&O technique gives higher
are considered in this paper. The objective is to find out the optimal ripple caused due to oscillation around MPP which is substantially re-
values of KP, KI and KD which gives excellent performance in terms of duced in IC technique and almost negligible in FL-DPID MPPT algo-
OS, RT, ST, IAE and ISE. The initial generation of GA is arbitrary; rithm. Higher ripple in the P&O algorithm reduces the average power
therefore the PID controller parameters at the preliminary stage could output and thereby reduces the efficacy of the SPV system. The max-
introduce instability in the system. Hence the lower and upper limit of imum obtained efficacy under all the considered states of solar insola-
the controller parameters is chosen such that the system retains stability tion is 97.00% for P&O, 98.60% for IC and 99.80% for FL-DPID MPPT
in this limit. The preliminary values of PID controller parameters are technique.
taken from ZN-PID controller that is given in Table 4. In terms of dynamic response, FL-DPID MPPT strategy shows better
performance in comparison to P&O and IC techniques. From the
5. Results and discussion Fig. 10(c) it can be observed that FL-DPID MPPT technique gives least
ST of 20 ms for power whereas in P&O and IC have 110 ms and 340 ms
The entire system represented schematically in Fig. 1 has been respectively. Considerable ripple in power leads to high ripple in the
formulated and simulated using MATLAB/Simulink software. The nu- output of the boost converter whose duty cycle ‘D’ in turn is decided by
merical values used in the simulation are given in Table 2. The designed the implemented MPPT technique. The ripple content in the output
system consists of an SPV array for a peak power of 200 W, a DC-DC voltage, as well as the current of the boost converter, is highest for P&O
boost converter and a DC-DC buck converter acting as a charge con- as shown in Fig. 10(a) and (b) with response remaining unsettled for the
troller for charging a 18 V battery and operating at a fs of 150 kHz. To entire simulation time and FL-DPID MPPT technique has almost negli-
operate the SPV system at MPP three distinct MPPT techniques are gible ripple with ST of 23 ms for voltage as well as current. From the
employed and a thorough comparative transient analysis of the dy- comparative analysis, FL-DPID MPPT algorithm surpasses P&O and IC
namic response of SPV system in terms of the output power of SPV algorithms in respect of steady state as well as dynamic response i.e.

85
P.K. Pathak, A.K. Yadav Solar Energy 178 (2019) 79–89

1000

800 800

Irradiance (W/m2)
600

400

State 1 State 2 State 3 State 4 State 5

0.0 0.4 0.5 0.9 1 1.4 1.5 1.9 2 2.4

Time (sec.)
Fig. 9. Profile of solar irradiance.

47
P&O
IC
40
FL-DPID
Voltage (V)

30

20

10
35
24
1.05 1.051 1.052 1.053
0
0 0.5 1 1.5 2 2.4
Time (sec.)
(a)

7 P&O
IC
6 FL-DPID

5
Current (A)

2
4.3
1
3.6
1.05 1.051 1.052 1.053
0
0 0.5 1 1.5 2 2.4
Time (sec.)
(b)
200
P&O
IC
150 FL-DPID
200
Power (W)

100
202 198
1.06 1.08 1.1
202
50 190
1.05 1.07 1.09 1.1 190
1.05 1.07 1.09 1.1
0
0 0.5 1 1.5 2 2.4
Time (sec.)
(c)
Fig. 10. (a) Output voltage, (b) Output current of boost converter and (c) Maximum power using P&O, IC and FL-DPID MPPT techniques.

86
P.K. Pathak, A.K. Yadav Solar Energy 178 (2019) 79–89

40 20

30 15
Voltage (V)

Voltage (V)
20 10 20
SV
10 ZN-PI
10 Output voltage of SPV 5 0 ZN-PID
Output voltage of FL-DPID based Boost Converter 0 0.01 0.02 0.03 0.04
O-PID
0 0
0 0.5 1 1.5 2 2.4 0 0.5 1 1.5 2 2.4
Time (sec.)
Time (sec.)
Fig. 11. Output voltage of SPV system and boosted output voltage using FL-
Fig. 12. Output voltage of buck converter.
DPID MPPT algorithm.

8
Table 3
Comparative performance analysis using P&O, IC and FL-DPID MPPT techni- 6

Current (A)
ques. 9
4
Parameter States P&O IC FL-DPID 5
ZN-PI
2 0 ZN-PID
Efficiency State 1 93.05% 93.40% 93.96% 0 0.01 0.02 0.03 0.04
State 2 95.18% 97.45% 96.30% O-PID
0
State 3 97.00% 98.60% 99.80% 0 0.5 1 1.5 2 2.4
State 5 96.70% 98.30% 82.43% Time (sec.)
Settling Time Power 0.11 sec 0.34 sec 0.02 sec
Voltage Not settled 0.35 sec 0.023 sec Fig. 13. Output current of buck converter.
Current Not settled 0.35 sec 0.023 sec
Ripple in Voltage State 1 High Less Negligible
150
(output of State 2 High Less Negligible
boost State 3 High Less Negligible
converter) State 5 High Less Negligible
100
Power (W)
Average Voltage State 2 i.e. 26.00 V 24.05 V 31 V
(Desired value (800 W/ 150
31 V) m2)
50 120 ZN-PI
Range of voltage – 10.80–41.20 23.30–24.80 constant
variation (V) 31 V 90 ZN-PID
Ripple in Current State 1 High Less Negligible 0 0.01 0.02 0.03 0.04
O-PID
(output of State 2 High Less Negligible 0
boost State 3 High Less Negligible
0 0.5 1 1.5 2 2.4
converter) State 5 High Less Negligible Time (sec. )
Average current State 2 i.e. 4.1A 3.785A 3.19A
Fig. 14. Output power of buck converter.
(Desired value (800 W/
3.19A) m2)
Range of current 1.7–6.5 3.66–3.91 constant Fig. 12, SV is set value i.e. 18 V. A comprehensive steady-state analysis
variation (A) 3.19
Control Strategy Sampling Sampling Intelligent
of ZN-PI, ZN-PID, and O-PID controlled buck converter under four
Method Method Control different states of solar insolation incorporating FL-DPID MPPT algo-
rithm for 200 W SPV system has been summarized in Table 5.
The output voltage and current of buck converter as depicted from
Table 4 Figs. 12 and 13, and Table 5 are maintained constant irrespective of the
Parameters of ZN-PI, ZN-PID and O-PID controllers. varying solar insolation thereby providing constant voltage and con-
Controller KP KI KD Ku andTu stant current to the load i.e. 18 V battery. The transient analysis of the
obtained responses infers that the ST of voltage, current, and power of
ZN-PI 0.01875 225 – Ku = 0.025&Tu = 0.1 × 10−3 ZN-PI controlled buck converter is 25 ms, 24 ms, and 22 ms respectively
ZN-PID 0.015 300 1.875 × 10−7 whereas the same for ZN-PID controlled buck converter is 23 ms, 22 ms,
O-PID 0.3157 293.6741 0.40723
and 19 ms respectively, and for O-PID is 1.6 ms, 1.5 ms, and 1.6 ms
respectively. The transient analysis thereby concludes that O-PID con-
efficiently tracking MPP under varying irradiance with least ST and trolled buck converter shows better performance as compared to ZN-PI
bearing least ripple in output voltage and current of the boost con- and ZN-PID controlled buck converter under dynamic condition. The
verter. The output voltage of the boost converter as seen in Fig. 11, is performance indices of ZN-PI, ZN-PID and O-PID controllers for the
varying in accordance with solar insolation that is undesirable for voltage of buck converter are shown in Table 6. The plot of IAE and ISE
battery charging application. Thus, O-PID controlled buck converter is for ZN-PI, ZN-PID, and O-PID tuned buck converter for voltage is shown
proposed in this paper. in Fig. 15.
From the values of Table 6 and Fig. 15, it is concluded that the ZN-PI
and ZN-PID controllers give almost similar performance whereas O-PID
5.2. Response of ZN-PI, ZN-PID, and O-PID controlled buck converter
controller gives significant improvement in terms of RT, ST, IAE, and
under varying solar irradiance
ISE i.e. 93.00%, 93.04%, 68.98%, and 90.49% respectively as compared
to ZN-PID controller. Therefore, the proposed O-PID controller is su-
The parameters of ZN-PI, ZN-PID and O-PID controllers of buck
perior among all three designed controllers for charging of a 18 V
converter are given in Table 4. The parameters of ZN-PI and ZN-PID
battery.
controllers are calculated using (24) and (25) respectively, whereas the
parameters of the O-PID controller is obtained using (26) after 65
generations of GA. The response of output voltage, current, and power 6. Conclusion
of ZN-PI, ZN-PID, and O-PID controlled buck converter under all four
discussed states are shown in Figs. 12, 13 and 14 respectively. In In this paper, a 200 W SPV system has been designed and its

87
P.K. Pathak, A.K. Yadav Solar Energy 178 (2019) 79–89

Table 5
Steady state analysis of ZN-PI, ZN-PID and O-PID controlled buck converter.
States ZN-PI ZN-PID O-PID

Ibuck(A) Vbuck(V) Ibuck(A) Vbuck(V) Ibuck(A) Vbuck (V)

2
State 1: (600 W/m and 25 °C) 7.57 18.01 7.567 18.02 7.56 18.03
State 2: (800 W/m2 and 25 °C) 7.585 17.93 7.57 18.01 7.56 18.03
State 3: (1000 W/m2 and 25 °C) 7.58 17.97 7.584 17.98 7.563 18.028
State 5: (400 W/m2 and 25 °C) 7.565 18.02 7.57 18.01 7.564 18.02

Table 6 Agarwal, V., Patel, H., 2010. MATLAB-based modeling to study the effects of partial
Performance index of ZN-PI, ZN-PID and O-PID controllers. shading on PV array characteristics. IEEE Trans. Energy Convers. 23, 302–310.
Al-Majidi, S.D., Abbod, M.F., Al-Raweshidy, H.S., 2018. A novel maximum power point
Controller Performance Parameters tracking technique based on fuzzy logic for photovoltaic systems. Int. J. Hydro.
Energy 43, 14158–14171.
OS (%) RT(ms) ST (ms) IAE ISE Algazar, M.M., AL-monier, H., EL-halim, H.A., Salem, M.E., EI, K., 2012. Maximum power
point tracking using fuzzy logic control. Elec. Pow. Energy Syst. 39, 21–28.
ZN-PI ≈0 22 25 0.3763 1weZ Ang, K.H., Chong, G., Li, Y., 2005. PID control system analysis, design, and technology.
ZN-PID ≈0 20 23 0.3585 1.187 IEEE Transactions on Control Systems Technology. 13 (4), 559–576.
O-PID 0.53 1.4 1.6 0.1112 0.1128 Babaei, E., Abbasnezhad, A., Sabahi, M., Hosseini, S.H., 2017. Analysis and design of a
soft switching boost DC/DC converter. IET J. Pow. Elec. 10, 1353–1362.
Improvement (%) – 93.00% 93.04% 68.98% 90.49%
Borage, M., Tiwari, S., Kotaiah, S., 2006. Constant-current, constant-voltage half-bridge
resonant power supply for capacitor charging. Inst. Eng. Tech. 153, 343–347.
Chen, Y.-T., Jhang, Y.-C., Liang, R.-H., 2016. A fuzzy-logic based auto-scaling variable
step-size MPPT method for PV systems. Sol. Energy. 126, 53–63.
Chen, P.-C., Chen, P.-Y., Liu, Y.-H., Chen, J.-H., Luo, Y.-F., 2015. A comparative study on
maximum power point tracking techniques for photovoltaic generation systems op-
erating under fast changing environments. Sol. Energy. 119, 261–276.
Chen, B.-Y., Lai, Y.-S., 2012. New digital controlled technique for battery charger with
constant current and voltage control without current feedback. IEEE Trans. Ind. Elec.
59, 1545–1553.
Cheng, P.-C., Peng, B.-C., Liu, Y.-H., Cheng, Y.-S., Huang, J.-W., 2015. Optimization of a
fuzzy logic control based MPPT algorithm using the particle swarm optimization
technique. Energies. 5339–5345.
Cho, S.-Y., Lee, Q., Moon, S.C., Moon, G.-W., Kim, B.-C., Young, K.K., 2013. Constant
current charging in series-series compensated non-radiative wireless power link. IEEE
2792–2795.
Fig.15. Plot of IAE and ISE. EI-Khatib, M.F., Shaaban, S., Mohamed, A.E., 2017. A proposed advanced maximum
power point tracking control for a photovoltaic-solar pump system. Sol Energy. 158,
321–331.
performances under four varying states of solar insolation have been Ekouevi, K., Tuntivate, V., 2012. Household energy access for cooking and heating.
observed. To operate the SPV system at MPP a novel FL-DPID MPPT Lesson learned and the way forward, The world bank report, pp. 70276.
technique with a lesser number of rules has been implemented and Eldahab, Y.A., Saad, N.H., Zekry, A., 2016. Enhancing the design of battery charging
controllers for photovoltaic systems. Renew. Sustain. Energy Rev. 58, 646–655.
compared with the existing P&O and IC techniques for the battery Elgendy, M.A., Zahawi, B., Atkinson, D.J., 2012. Assessment of perturb and observe MPPT
charging circuit. In comparison with P&O and IC techniques, the ob- algorithm implementation techniques for PV pumping application. IEEE Trans.
tained results depict that FL-DPID MPPT technique has higher max- Sustainable Energy. 3, 21–33.
Elgendy, M.A., Zahawi, B., Atkinson, D.J., 2013. Assessment of the incremental con-
imum efficiency, effective tracking speed and no deviation from the ductance maximum power point tracking algorithm. IEEE Trans. Sustain. Energy. 4,
MPP under varying atmospheric conditions. Therefore, the FL-DPID 108–117.
MPPT technique gives an excellent performance as compared to P&O Fakham, H., Lu, Di, Francois, B., 2011. Power control design of a battery charger in a
hybrid active PV generator for load-following applications. IEEE Trans. on Ind. Ele.
and IC techniques in terms of the output voltage and current of the 58 (1), 85–94.
boost converter, which is the main area of concern. In divergence with Fathabadi, H., 2016. Novel high-efficiency DC/DC boost converter for using in PV sys-
the existing literature which employs PI controller for charging a bat- tems. Sol. Energy. 125, 22–31.
Graditi, G., Colonnese, D., Femia, N., 2010. Efficiency and reliability comparison of DC-
tery, in this paper a comparative analysis of ZN-PI, ZN-PID and O-PID
DC converters for single phase grid connected photovoltaic inverters. SPEEDAM
controlled buck converter as a battery charging circuit has been done. 140–147.
The proposed O-PID controller gives excellent performance in terms of Graditi, G., Adinolfi, G., 2011. Comparative analysis of synchronous rectification boost
and diode rectification boost converter for DMPPT applications. ISIE 1000–1005.
RT, ST, IAE and ISE with the improvement of 93.00%, 93.04%, 68.98%,
Graditi, G., Adinolfi, G., Tina, G.M., 2014. Photovoltaic optimizer boost converters:
and 90.49% respectively as compared to ZN-PID controller. Hence the temperature influence and electro-thermal design. App. Energy 115, 140–150.
O-PID controller is superior among all designed controllers for charging Hart, D.W., 2011. Power electronics. McGraw Hill Edu.
of a 18 V battery. Thus it can be concluded that FL-DPID MPPT based Haq, S.S., Wilson, S.S.B., Zameer, M.G., 2013. Design and simulation of MPPT algorithm
of photovoltaic system using intelligent controller. Int J Adv Sci Tech Res. 3,
SPV system used for charging a battery through O-PID control as charge 2249–2254.
controller is highly efficient for fast and effective battery charging Larminie, J., Lowry, J., 2003. Electric vehicle technology explained. Wiley, Newyork.
thereby reducing losses and enhancing the battery life cycle. The future Liu, F., Duan, S., Liu, F., Liu, B., Kang, Y., 2008. A variable step size INC MPPT method for
PV system. IEEE Trans. Ind. Elec. 81 (7), 2622–2628.
scope of the research work will predominately focus on the perfor- Lopez, J., Seleme Jr, S.I., Donoso, P.F., Cortizo, P.C., Severo, M.A., 2016. Digital control
mance of O-PID controlled charge controller for charging a battery fed strategy for a buck converter operating as a battery charger for stand-alone photo-
through a partially shaded SPV system. If there is a relevant aid to the voltaic systems. Sol. Energy 140, 171–187.
Metin, C., Boztepe, M., Karatepe, E., 2007. Development of a suitable model for char-
research material, the designed system can be practically realized and acterizing photovoltaic arrays with shaded solar cells. Sol. Energy. 81, 977–992.
implemented. Mirzaei, A., Forooghi, M., Ghadimi, A.A., Abolmasoumi, A.H., Riahi, M.R., 2017. Design
and construction of a charge controller for stand-alone PV/battery hybrid system by
using a new control strategy and power management. Sol. Energy 149, 132–144.
References
Neath, M.J., Swain, A.K., Madawalaand, U.K., Thrimawithana, D.J., 2014. An optimal PID
controller for a bidirectional inductive power transfer system using multiobjective
Adinolfi, G., Graditi, G., Siano, P., Piccolo, A., 2015. Multi-Objective optimal design of genetic algorithm. IEEE Trans. Pow. Elec. 29 (3), 1523–1531.
photovoltaic synchronous boost converters assessing efficiency, reliability and cost Pode, R., 2015. Battery charging stations for home lighting in mekon region countries.
savings. IEEE Trans. Ind. Info. 11 (5), 1038–1048. Renew. Sustain. Energy Rev. 44, 543–560.

88
P.K. Pathak, A.K. Yadav Solar Energy 178 (2019) 79–89

Rajani, S.V., Pandya, V.J., 2016. Experimental verification of the rate of charge im- rural electrification, Third ed. Alliance for rural electrification. http://www.
provement using photovoltaic MPPT hardware for the battery and ultracapacitor ruralelec.org.
storage devices. Sol. Energy 139, 142–148. Wirasingha, S.G., Pihef, A.E., 2011. Plug in hybrid electric factor. IEEE Trans Veh
Rezk, H., Eltamaly, A.M., 2015. A comprehensive comparision of different MPPT tech- Technol. 60, 1279–1284.
niques for photovoltaic systems. Sol. Energy. 112, 1–11. World energy outlook 2013. International Energy Agency (IEA), Paris, 2013.
Singh, A.K., Hussain, I., Singh, B., 2018. Double- stage three-phase grid-integrated solar Yadav, A.K., Gaur, P., 2016a. Improved self-tuning fuzzy proportional–integral derivative
PV system with fast zero attracting normalized least mean fourth based adaptive versus fuzzy-adaptive proportional–integral–derivative for speed control of nonlinear
control. IEEE Trns. Indst. Elec. 65, 3921–3931. hybrid electric vehicles. J. Comput. Nonlinear Dynamics 11 (6), 061013–61017.
De Soto, W., Klein, S.A., Beckman, W.A., 2006. Improvement and validation of a model Yadav, A.K., Gaur, P., 2016b. An optimized and improved STF-PID speed control of
for photovoltaic array performance. Solar Energy 80, 78–88. throttle controlled HEV. The Arabian Journal for Science and Engineering. 41 (9),
Universal access to modern energy for poor. http://www.undp.org/content/undp/en/ 3749–3760.
home/outwork/environmentandenergy/focus_areas/sustainable-energy/universal- Yilmaz, U., Kircay, A., Borekci, S., 2018. PV system fuzzy logic MPPT method and PI
access.html. control as a charge controller. Renew. Sustain. Energy Rev. 81, 994–1001.
Weimann, M., Cantos Gomez, E.M., MiroBaz, L.-C., 2013. Best practices of the alliance for

89