Modeling and Analysis of Maximum Power Point

Tracking Algorithms Using MATLAB/Simulink

Md. Rashed Hassan Bipu*, Syed Mohammad Sifat Morshed Chowdhury, Manik Dautta, Md. Zulkar Nain and
Shahidul Islam Khan
*rashed.hassan.bipu@gmail.com

Abstract— Maximum power point tracking (MPPT) is an open-circuit voltage and fractional short-circuit current
important part of solar photovoltaic (PV) system. It increases the algorithms are relatively easier and cheaper to implement. In
efficiency of a solar panel by tracking the maximum power point. recent time, some artificial intelligence based algorithms have
There are several MPPT control algorithms in use. In this paper, emerged such as fuzzy logic based algorithm, stimulated
four control algorithms are analyzed comparatively [1]. Using annealing, particle swarm optimization etc. In this paper, four
MATLAB/Simulink a solar PV system with MPPT controlled algorithms are analyzed namely perturb-and-observation,
buck-boost dc-dc converter is modeled. Then the efficiency of incremental conductance, fractional open-circuit voltage and
each algorithm is calculated using typical daily insulation and fuzzy logic based method. The efficiency of the algorithms is
temperature variation. Finally, a comparative analysis of the
determined from (1).
algorithms is presented.
𝑡
0 𝑃 𝑎𝑐𝑡𝑢𝑎𝑙 (𝑡)𝑑𝑡
Keywords—Solar photovoltaic; Maximum power point 𝜂𝑀𝑃𝑃𝑇 = 𝑡  (1)
0 𝑃𝑚𝑎𝑥 (𝑡)𝑑𝑡
tracking; P&O ; Incremental Conductance; Fuzzy logic; Simulink
Where 𝑃𝑎𝑐𝑡𝑢𝑎𝑙 is the actual (measured) power produced by
I. INTRODUCTION the PV array under the control of the MPPT, and 𝑃𝑚𝑎𝑥 is the
When a PV panel is directly connected to a load, the true maximum power the array could produce under a given
operating point is rarely at maximum power point. The temperature and irradiance.
maximum power point tracker maintains the PV array
operating point at maximum power point. Fig. 1 shows the I-V
and P-V characteristic curve and maximum power point of a
PV panel. To track the maximum power point a dc-dc
converter is placed between the panel and the load. The
converter’s duty cycle is set such that the power flow from the
panel is maximum and the load voltage is as per requirement.
The I-V characteristic of solar panel depends on the irradiance
and cell temperature. So, the maximum power point changes
with varying conditions and the duty cycle also needs to be
changed accordingly. The tracker uses some control algorithm
to do this task. Fig. 2 shows a block diagram of solar PV
system with MPPT. Fig. 2 Block diagram of solar PV system with MPPT.

II. DESCRIPTION OF ANALYSED MPPT ALGORITHMS

A. Perturb and observe (P&O)
The Perturb and observe algorithm is the most commonly
used in practice because of its ease of implementation. In the
P&O algorithm, the operating voltage of the PV array is
perturbed by a small increment, and the resulting change in
power, P, is measured. If P is positive, then the perturbation of
the operating voltage moved the PV array’s operating point
closer to the MPP. Thus, further voltage perturbations in the
same direction (that is, with the same algebraic sign) should
move the operating point toward the MPP. If P is negative, the
system operating point has moved away from the MPP, and
the algebraic sign of the perturbation should be reversed to
Fig. 1 I-V and P-V curve of PV panel. move back toward the MPP. These can be summarized using
the Table I [3].
A number of MPPT control algorithms have been
proposed [3]. The perturb-and-observe (P&O) algorithm is by B. Incremental Conductance (IC)
far the most commonly used in commercial MPPTs. Another The incremental conductance method is based on the fact
algorithm of choice is incremental conductance. Fractional that the slope of the PV array power curve is zero at the MPP,
positive on the left of the MPP, and negative on the right, as 𝑉𝑀𝑃𝑃 𝑉𝑂𝐶 ≈ 𝐾 < 1 (3)
seen in the Fig. 2 [3]. The constant voltage algorithm can be implemented using
the flowchart shown in Fig. 4. The solar array is temporarily
TABLE I. SUMMARY OF P&O ALGORITHM isolated from the MPPT, and a 𝑉𝑂𝐶 measurement is taken.
Perturbation Change in Power Next perturbation
Next, the MPPT calculates the correct operating point using
Equation and the pre set value of K, and adjusts the array’s
Positive Positive Positive voltage until the calculated 𝑉𝑀𝑃𝑃 is reached. This operation is
Positive Negative Negative repeated periodically to track the position of the MPP [2].
Negative Positive Negative Although this method is extremely simple, it is difficult to
choose the optimal value of the constant K. The literature
Negative Negative Positive
reports success with K values ranging from .71 to .80. In this
paper, K=.71 was used as it gives the best efficiency.
dP /dV =0, at MPP,
dP /dV >0, left of MPP,
dP /dV <0, right of MPP.
Since,
𝑑𝑃 𝑑(𝐼𝑉) 𝑑𝐼 Δ𝑉
𝑑𝑉
= 𝑑𝑉
= 𝐼 + 𝑉 𝑑𝑉 ≅ 𝐼 + 𝑉 Δ𝑉 (2) Fig. 4 Flowchart of fractional open circuit voltage algorithm.

D. Fuzzy Logic based algorithm

∆I/ΔV = -I/V, at MPP. Recently fuzzy logic controllers have been widely used for
industrial processes owing to their simplicity and effectiveness
∆I/ΔV > -I/V, left of MPP. for both linear and nonlinear systems. A MPP search based on
∆I/ΔV < -I/V, right of MPP. fuzzy logic does not need any parameter information. It
consists of a step wise adaptive search as well as leads to fast
convergence. The fuzzy controller consists of three functional
blocks: 1) Fuzzification, 2) Fuzzy rule algorithm and 3)
Defuzzification [5]. Also the fuzzy logic controller should be
modified and optimized base on the system it is controlling. In
this thesis, we devised a fuzzy logic controlled MPPT which is
optimized for this MPPT system.
These functions are described as follows:
1) Fuzzification: The fuzzy logic based MPPT method has
two input variables, namely dP(k) and dV(k), at a sampling
instant k. The output variable is dD(k+1) which is duty cycle
change that causes voltage to change at next sampling instant
(k+1). The variable dP(k) and dV(k) are expressed as follows:

dP(k)=P(k)-P(k-1) (4)
dV(k)=V(k)-V(k-1) (5)

In Fig. 5(a), the membership functions of the input

variable dP(k) is assigned five fuzzy sets, including positive
big (PB), positive small (PS), zero (ZE), negative small (NS),
and negative big (NB). The membership functions are denser
at the center in order to provide more sensitivity against
variation in the PV array terminal voltage. In Fig. 5(b), the
membership functions of the input variable dV(k) is assigned
three fuzzy sets, including positive (P), zero (Z), and
Fig. 3 Flow chart of Incremental Conductance algorithm. negative(N). Fig. 5(c) shows the membership functions of the
C. Fractional Open Circuit Voltage output variable dD(k+1) which is assigned fuzzy sets,
including positive big (PB), positive middle (PM), zero (ZE),
The basis for the fractional open circuit voltage algorithm negative middle (NM), and negative big (NB).
is the observation that the ratio of the array’s maximum power
2) Fuzzy rule algorithm: The rule base that associates the
voltage to its open-circuit voltage is approximately constant.
fuzzy output to the fuzzy inputs is derived by understanding
the system behavior. So P-V curve was observed and rules Using the steps mentioned above, the fuzzy controller was
were determined which are summerized in Table II. implemented in real time for MPPT.
III. MODELING OF SOLAR PV SYSTEM WITH MPPT
To analyze the efficiency of the MPPT algorithms a solar
PV system with MPPT is modeled in MATLAB/Simulink.
First Solar PV panel, DC-DC converter with battery and
MPPT block were modeled and then they were connected to
make the system.
A. PV panel modeling
(a)
A general mathematical description of I-V output
characteristics for a PV cell has been studied for over the past
four decades. Such an equivalent circuit-based model is
mainly used for the MPPT technologies [4]. The equivalent
circuit of the general model which consists of a photo current,
a diode, a parallel resistor expressing a leakage current, and a
series resistor describing an internal resistance to the current
flow, is shown in Fig. 6. The voltage-current characteristic
(b)
equation of a solar cell is given as
q V+IR S V+IR S
I = IPH − IS exp KT C A
−1 − R SH
(8)
Where IPH is a light-generated current or photocurrent, IS
is the cell saturation of dark current, q (=1.6x10-19 C) is an
electron charge, k (=1.38x10-23 J/K) is a Boltzmann’s
constant, TC is the cell’s working temperature, A is an ideal
factor, R SH is a shunt resistance, and R S is a series resistance.
The photocurrent mainly depends on the solar insolation and
(c) cell’s working temperature, which is described as,
Fig. 5 Membership function of (a) dP(k), (b) dV(k) and (c) dD(k+1) 𝐼𝑃𝐻 = 𝐼𝑆𝐶 + 𝐾𝐼 𝑇𝐶 − 𝑇𝑅𝐸𝐹 𝜆 (9)
Where 𝐼𝑆𝐶 is the cell’s short-circuit current at a 25° C and
TABLE II. SUMMARY OF P&O ALGORITHM
1kW/𝑚2 , 𝐾𝐼 is the cell’s short-circuit current temperature
dP(k)→ coefficient, 𝑇𝑅𝐸𝐹 is the cell’s reference temperature, and λ is
NB NS ZE PS PB the solar irradiance in kW/𝑚2 . On the other hand, the cell’s
dV(k)↓ dD(k+1)↘
saturation current varies with the cell temperature, which is
N NB NM ZE PM PB described as
Z ZE ZE ZE ZE ZE 1 1
𝑞𝐸 𝐺 −
𝑇 𝑅𝐸𝐹 𝑇 𝐶
P PB PM ZE NM NB 𝐼𝑆 = 𝐼𝑅𝑆 (𝑇𝐶 − 𝑇𝑅𝐸𝐹 )3 exp (10)
𝑘𝐴

3) Defuzzification: After the rules have been evaluated, the Where 𝐼𝑅𝑆 is the cell’s reverse saturation current at a
last step to complete the fuzzy control algorithm is to calculate reference temperature and a solar irradiation, 𝐸𝐺 is the bang-
the crisp output of the fuzzy control with the process of gap energy of the semiconductor used in the cell. The ideal
defuzzification. The well known center of gravity method for factor A is dependent on PV technology.
defuzzification is used here. It computes the center of gravity
The shunt resistance R SH is inversely related with shunt
from the final fuzzy space, and yields a result which is highly leakage current to the ground and it can be assumed as infinity
related to all of the elements in the same fuzzy set. The crisp without leakage current to ground. The appropriate model of
value of control output dD(k+1) is computed by the following PV solar cell with suitable complexity is shown in (11).
equation: Equation (8) can be rewritten to be
𝑛
𝑖=1 𝑤 𝑖 𝑑𝐷 𝑖
𝑑𝐷 = 𝑛 𝑤 (6) 𝑞 𝑉+𝐼𝑅𝑆
𝑖=1 𝑖 𝐼 = 𝐼𝑃𝐻 − 𝐼𝑆 exp −1 (11)
𝐾𝑇𝐶 𝐴
Where n is the maximum number of effective rules, 𝑤𝑖 is
the weighting factor, and 𝑑𝐷𝑖 is the value corresponding to the Since a typical PV cell produces less than 2W at 0.5V
membership function of dD. Then, the final control voltage is approximately, the cells must be connected in series-parallel
obtained by adding this change to the previous value of the configuration on a module to produce enough high power. The
control voltage: equivalent circuit for the solar module arranged in 𝑁𝑃 parallel
D(k+1) = D(k) + dD(k+1) (7) and 𝑁𝑆 series is shown in Fig. 2(a). As shunt resistance
approaches infinity the terminal equation for the current and model takes voltage, irradiance and cell temperature as input
voltage of the array becomes as follows. and gives current as output.
𝑞 𝑉/𝑁𝑆 +𝐼𝑅𝑆 /𝑁𝑃
𝐼 = 𝑁𝑃 𝐼𝑃𝐻 − 𝑁𝑃 𝐼𝑆 exp 𝐾𝑇𝐶 𝐴
−1 (12)

(a)

Fig. 9 Subsystem implementation of generalized PV array model.

B. DC-DC converter modeling

The buck converter is a step down converter where boost
is step up converter. The buck-boost converter can act both as
(b)
a step up and step down converter according to the duty cycle
Fig. 6 (a) General model and (b) appropriate model of a solar cell. [6]. This topology was used for DC-DC converter for MPPT
system. In Simulink, SimPowerSystems library was used to
create a buck-boost converter. The input was connected to the
solar panel block and output with a battery. MOSFET was
used as the switch.

Fig. 7 generalized model of a PV array.

Since normally 𝐼𝑃𝐻 >>𝐼𝑆 and ignoring the small diode and
ground-leakage currents under zero-terminal voltage, the
short-circuit current 𝐼𝑆𝐶 is approximately equal to the
(a)
photocurrent 𝐼𝑃𝐻 . On the other hand, the V_OC parameter is
obtained by assuming the output current is zero. Given the PV
open-circuit voltage 𝑉𝑂𝐶 at reference temperature and ignoring
the shunt-leakage current, the reverse saturation current at
reference temperature can be approximately obtained as,
𝐼𝑆𝐶
𝐼𝑅𝑆 = 𝑞 𝑉 𝑂𝐶 (13)
exp −1
𝑁 𝑆 𝑘𝐴 𝑇 𝐶

(b)
Fig. 10 (a) MATLAB model and (b) subsystem implementation of buck-boost
converter with battery as load.

C. MPPT modeling
MPPT block can be described as the brain of the system. It
controls the duty cycle of the converter by using some
algorithm to track the MPP of the solar panel. After the duty
cycle, D is calculated it is converted to pulses using repeating
sequence, sum and compare to zero block. The PWM
frequency was selected to be 20 KHz. Fig. 11 shows the
Fig. 8 Solar panel model.
Simulink model.

Using these equations a generalized solar panel mask The MPPT block contains necessary parameters and codes
model was made in MATLAB/Simulink [3]. The solar panel for implementing MPPT. Fig. 12 shows subsystem
implementation of MPPT block. The Embedded MATLAB
Function block contains codes. For P&O, IC and fractional
open circuit voltage algorithms MATLAB codes were written
using Table I, Fig. 3 and Fig. 4. Fuzzy Logic based algorithm
subsystem implementation is different from Fig. 12 and
presented in Fig. 13.

Fig. 14 MATLAB/Simulink based PV array MPPT system.

IV. SIMULATION
The MPPT algorithms were simulated for two profiles of
solar irradiance and cell temperature. Case 1 depicts when the
Fig. 11 MATLAB/Simulink model for MPPT block. irradiance and temperature undergoes relatively low changes.
The irradiance level changes from 800 𝑊/𝑚2 to 1000 𝑊/𝑚2
and then to 900𝑊/𝑚2 . The temperature changes from 30°C to
32°C. Case 2 depicts the condition when the irradiance and
temperature is highly varying. The irradiance level changes
from 600 𝑊/𝑚2 to 400 𝑊/𝑚2 to 500 𝑊/𝑚2 to 700 𝑊/𝑚2
and finally to 900𝑊/𝑚2 . The temperature changes from 26°C
to 25°C to 27°C. The time duration for both cases of
simulation duration was 1.02 sec and MPPT was enabled after
.02 second.
1.0 0.9 0.9
0.8
(KW/𝑚 2 )
Irradiance

0.6 0.7
Fig. 12 Subsystem implementation of P&O, IC and fractional open circuit 0.5
voltage MPPT block. 0.4
Temperature

32
30
(°C)

26 27
25

(a) (b)
Fig. 15 Irradiance and temperature profile for (a) Case 1 and (b) Case 2

Simulation was done for 5 cases. First case was no MPPT

or the battery was directly connected to the panel. The voltage
and current were multiplied and then taken to workspace to
Fig. 13 MATLAB/Simulink based PV array MPPT system. calculate the harvested energy. Similarly, simulation were
done for perturb and observation, incremental conductance,
fractional open circuit voltage and fuzzy logic based.
D. Total PV system modeling To determine the true maximum power the array could
The PV panel block, Converter with battery block and produce under a given temperature and irradiance the P-V
MPPT block previously created were connected as the Fig. 14. curves were plotted under that condition. From that curve the
The output of the solar panel block is the panel current (𝐼𝑖𝑛𝑝𝑢𝑡 ) maximum available power were determined.
and is connected to the input of the converter. The converter
with battery block gives panel voltage (𝑉𝑖𝑛𝑝𝑢𝑡 ), Load voltage V. RESULTS
(𝑉𝑜𝑢𝑡 ) and current (𝐼𝑜𝑢𝑡 ) as the output. Panel voltage (𝑉𝑖𝑛𝑝𝑢𝑡 ) The available true maximum power is determined by
is input of the solar panel block. The MPPT block takes constructing P-V curve for different irradiance and
𝑉𝑖𝑛𝑝𝑢𝑡 and 𝐼𝑖𝑛𝑝𝑢𝑡 as the input and gives duty cycle, D as the temperature of the two profiles. The simulation gave the actual
output which controls the converter block. The solar irradiance powers obtained by using four different. These are plotted in
and cell temperature profile were given as input to the solar Fig. 16 for both cases of irradiation and temperature variation
panel. profile.
 Case 2 Case 1
70 70

60 60
100
50 50

40 40
90 98.43 98.57 98.67

P=VI
P=VI

30 30 93.89
20 20
81.14
10 10
80
No Mppt P&O Inc. Con OV Fuzzy
0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time Time
(a)
Case 1 Case 2
70 70

60 60
Fig. 17 Average efficiency of Solar PV system with and without different
MPPT algorithms
50 50

40 40
VI. CONCLUSION
P=VI

P=VI

30 30
Fig. 17 presents a comparison of the efficiencies of four
20 20
MPPT control algorithms with no MPPT situation. From the
10 10
simulation results we found that all MPPT algorithms
0
0 0.1 0.2 0.3 0.4 0.5
Time
0.6 0.7 0.8 0.9 1
0
0 0.1 0.2 0.3 0.4 0.5
Time
0.6 0.7 0.8 0.9 1 performed much better than the case where no MPPT was
Case 1
(b) Case 2 used. So, we can say that using MPPT can increase the
70 70
efficiency of the solar PV system considerably.
60 60

50 50
The fractional open circuit voltage method was the
simplest considering ease of implementation and number of
40 40
sensors but its performance was least impressive with 93.89%
P=VI
P=VI

30 30 efficiency. Other three MPPT algorithms have higher

20 20 efficiency. Among them Perturb and observation is the most
10 10
popular algorithm and by optimizing the step size we found
out that its performance is satisfactory with 98.43% efficiency.
0 0
0 0.1 0.2 0.3 0.4 0.5
Time
0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5
Time
0.6 0.7 0.8 0.9 1 Incremental Conductance has higher efficiency (98.57%) than
(c) P&O and Fuzzy logic based algorithm has the highest
efficiency (98.67%). We also found out that among the four
Case 1 Case 2
70 70 algorithms fuzzy logic based algorithm performed best in case
60 60 1 and incremental conductance on case 2. On an average basis
50 50
the fuzzy logic based algorithm outperformed other
algorithms. But the complexity of implementing higher
40 40
implementation cost can only be justified in much larger PV
P=VI

P=VI

30 30
systems.
20 20

10 10
REFERENCES
[1] Md. Rashed Hassan Bipu, Syed Mohammad Sifat Morshed Chowdhury,
0 0
0 0.1 0.2 0.3 0.4 0.5
Time
0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5
Time
0.6 0.7 0.8 0.9 1 Manik Dautta and Md. Zulkar Nain. 2014. Study and analysis of solar
charge controller and different mppt algorithms, B.Sc. Thesis,
Case 1
(d) Case 2
Bangladesh University of Engineering and Technology.
70 70
[2] D. P. Hohm and M. E. Ropp, “Comparative Study of Maximum Power
60 60 Point Tracking Algorithms”, PROGRESS IN PHOTOVOLTAICS:
50 50
RESEARCH AND APPLICATIONS, Published online 22 November
2002.
40 40
[3] Trishan Esram and Patrick L. Chapman, “Comparison of Photovoltaic
P=VI

P=VI

30 30 Array Maximum Power Point Tracking Techniques”, IEEE

TRANSACTIONS ON ENERGY CONVERSION, VOL. 22, NO. 2,
20 20
JUNE 2007.
10 10 [4] Huan-Liang Tsai, Ci-Siang Tu, and Yi-Jie Su, “Development of
0 0
Generalized Photovoltaic Model Using MATLAB/SIMULINK”,
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time Time Proceedings of the World Congress on Engineering and Computer
(e) Science 2008, San Francisco, USA.
Fig. 16 Actual maximum power available (blue line) and power produced [5] Jiyong Li and Honghua Wang, “Maximum Power Point Tracking of
by PV array with (a) no MPPT, (b) P&O, (c) IC, (d) Fractional open circuit Photovoltaic Generation Based on the Fuzzy Control Method”.M.
voltage and (e) fuzzy logic based MPPT algorithm (red line) Young, The Technical Writer’s Handbook. Mill Valley, CA: University
Science, 1989.
The areas under the curves were determined. Using these [6] Wikipedia, (2014). Buck–boost converter. [online] Available at:
values the efficiency of the algorithms were calculated in both http://en.wikipedia.org/wiki/Buck%E2%80%93boost_converter
cases and then averaged and presented in Fig. 17. [Accessed 12 Jan. 2014].