Zakariah 2015

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

2015 IEEE Student Conference on Research and Development (SCOReD)

Dual-Axis Solar Tracking System Based on Fuzzy


Logic Control and Light Dependent Resistors as
Feedback Path Elements
Azwaan Zakariah, Jasrul Jamani Jamian, Mohd Amri Md Yunus
Innovative Engineering Research Alliance
Control and Mechatronics Engineering Department, FKE
Universiti Teknologi Malaysia
81310 Johor Bahru, Johor, Malaysia
amri@fke.utm.my

Abstract—The power generated by solar panels is mainly growing type of renewable energy technology, because the
influenced by the intensity of solar irradiance. This project primary source is always costless, accessible, plentiful, non-
utilized dual-axis solar tracking based on light intensity polluting, safe, and clean. Solar radiation is the quantity of solar
comparison with a fuzzy logic intelligent method. The energy that falls on a solar PV panel over a period of time, and
hardware system consists of an Arduino UNO is called solar irradiation. Solar radiation is made up of a
microcontroller as the main controller and four Light combination of three types of radiation: direct, diffuse, and
Dependent Resistor (LDR) sensors for sensing the reflected. The limitation of solar energy is that when the
maximum incident intensity of solar irradiance. Initially, weather is cloudy and/or wet, solar photovoltaic (PV) panels
two power window motors (one for the horizontal axis and cannot capture enough solar radiation to produce electricity. In
the other for the vertical axis) coordinate and alternately most cases, the solar-tracking systems that have been
rotate to scan the position of the sun. The mechanism of developed by researchers are closed-loop systems and employ
each axis movement is based on a power window motor many kinds of tracking techniques to track the sun’s position.
which is controlled by a fuzzy logic controller system that Most previous researches combined the system with a
moves according to the inputs received from the LDR microcontroller or programmable logic controller with sensors,
sensors. The fuzzy logic controller helps the microcontroller a real-time clock (RTC), or a combination of an RTC with
to give the best inference concerning the direction to which sensors. Nevertheless, only a simple system will be applied to
the solar PV panel should rotate and the position in which this project in order to make sure the system is ready to harvest
it should stay. In conclusion, the solar tracking system the maximum output power and reduce the operating energy
delivers high efficiency of output power with a low power consumption. The angles and fuzzy logic control will be
intake while it operates. implemented to reduce the time taken by the system to make
Keywords—Dual-axis tracker; solar renewable energy; the decision to rotate the solar PV panel.
sunlight intensity method; styling; light dependent resistor
In these circuit and model, two power window motors are
(LDR); fuzzy logic.
operated separately at any time. Since the sun changes its
I. INTRODUCTION position, after every five minutes the sensor detects the position
of the sun. If the sun moves from east to west, the second power
Nowadays, fossil fuel energy, such as coal energy, natural
window motor, which is responsible for vertical movement of
gas energy, and so on, contributes significantly to the
the solar tracker, will stop working. The second power window
utilization and development of industrial society. Electric
motor will start running only after the sun moves to a northward
power generation using renewable energy sources has begun to
or southward position. Thus, this system will result in low
take on a dominant role in fulfilling the demand for electric
power consumption. The vertical and horizontal movements of
energy. The diminishing fossil fuel resources are expected to
the solar PV panel on the azimuth and altitude angles are taken
be exhausted within the next hundred years [1]. Renewable
as a reference. The exact position of the sun is estimated from
energy is energy that is generated from natural surroundings
these two angles.
and is continuously replenished. Wind, sunlight, water,
geothermal heat, and biomass are sources of energy that are In addition, one problem of solar PV is that it has low
constantly renewed. efficiency in producing the maximum output power from
sunlight. To solve this problem, many research works have
Among the different types of renewable energy, solar
been carried out to increase the efficiency of output power by
energy is the most effective and efficient for producing
using solar-tracking systems. The solar-tracking system is used
electricity. In countries where a warm climate is dominant,
to keep the solar PV panels perpendicular to the direction of
solar energy is examined as one of the predominant energy
solar radiation. Solar-tracking systems are also considered as
resources [2]. Moreover, solar technology is also the fastest
mechatronics systems, which merge elements of mechanics,
The authors wish to acknowledge the finance assistance from Universiti
Teknologi Malaysia under GUP Research Vote entitled “An Efficient Solar
Tracking System with Fuzzy Logic Method for Tropical Environment”, in part
by the Fundamental Research Scheme Grant (FRGS) from the Ministry of
Higher education Research (MOHE) Malaysia (Vote No. 4F594), and in part
by the ScienceFund through the Ministry of Science, Technology and
Innovation Malaysia (Vote No. 4S098).

978-1-4673-9572-4/15/$31.00 ©2015 IEEE 139


2015 IEEE Student Conference on Research and Development (SCOReD)

electronics, and information technology [3]. Increases in the


efficiency of energy production of 35% per year have been
Fuzzy Main Output
achieved by solar tracking systems [4].
Logic Controller (Power
Thus, this research aims to design and develop a dual axis Controller (Arduino) Window)
solar tracking system based on light intensity comparison by
using four LDR sensors. Fuzzy logic algorithm for gives the
best accuracy in positioning is also intensively highlighted.
Moreover, the system should be robustly developed with high
Input
quality materials, but in low cost [5]. Lastly, the power
efficiency of the developed solar tracking system will be (LDR)
compare with a non-tracing system under the same location and
environment.
Fig. 1. Block diagram of overall system
II. SOLAR TRACKER DESIGN implemented to reduce the time taken by the system to make the
The solar-tracking system with an intelligent method decision to rotate the solar PV panel.
consists of an Arduino UNO microcontroller board, four LDR
sensors, two power window motors, a motor driver board, and In this circuit and model, two power window motors are
an MPPT charger module. An FLC is applied in the system operated separately at any time. Since the sun changes its
through Arduino programming. The overall system is position, after every five minutes the sensor detects the position
illustrated in Fig. 1. The aim of this circuit design is to of the sun. If the sun moves from east to west, the second power
minimize the power consumption and maximize the solar window motor, which is responsible for vertical movement of
energy generation. the solar tracker, will stop working. The second power window
motor will start running only after the sun moves to a northward
The Arduino UNO microcontroller board which controls or southward position. Thus, this system will result in low power
the overall performance of the systems also controls all the consumption. The vertical and horizontal movements of the
subsystems and feedback operation. The MPPT charger adjusts solar PV panel on the azimuth and altitude angles are taken as a
the current and voltage outputs of the solar PV panel, and then reference. The exact position of the sun is estimated from these
the FLC supplies the leading inference and decision to an two angles.
Arduino UNO board concerning the position to which the solar
PV panel should rotate and the direction of rotation.
B. Controller Ciruit Design
Four LDRs sensors are placed on the solar PV panel, one
Basically, the circuit design comprises the solar controller
on each side. These sensors are used for sensing the positional
(DC to DC converter), MPPT charger module, tracker
change of the sun, thus tracking the movement of the sun. The
controller, and inverter (DC to AC inverter), but this project
outputs of the LDR sensors are inputted into the Arduino and
focuses only on the tracker controller circuit design. The solar
subsequently used as the input variable for the FLC. Then the
controller, MPPT charge controller, and inverter were used for
FLC makes decisions based on its approximate reasoning and
the current generator. In this project, an Arduino UNO
sends the outputs to the motor drivers. Two power window
microcontroller is integrated with LDR sensors, motor driver
motors pivot the solar PV panel horizontally and vertically to
board, power window motors, and MPPT charger.
the position in which the panel is perpendicular to the sun and
the intensity of solar irradiance is maximized. The sensors
1. Solar PV Panel
continuously monitor the solar radiation and these data are
transferred to the power window motor via the Arduino UNO This project utilizes a poly-crystalline (m-Si) solar PV panel,
microcontroller. which was chosen because of its good longevity, efficiency,
lower installation cost, embodied energy, the fact that
monocrystalline solar PV panels are not hazardous to the
A. System Block Diagram environment, greater heat resistance, and higher electricity
In most cases, the solar-tracking systems that have been generation. This 10-W solar panel is a combination of a solar
developed by researchers are closed-loop systems and employ cell which generates a maximum of 10 W of electricity when
many kinds of tracking techniques to track the sun’s position. exposed to sunlight. It outputs from 2 to 20 V depending on the
Most previous researches combined the system with a sunlight intensity and comes ready with two cables terminated
microcontroller or programmable logic controller with sensors, with crocodile clips (black and red). With a medium size (78 mm
a real-time clock (RTC), or a combination of an RTC with × 26 mm) and light weight (1.2 kg), it is suitable for a variety of
sensors. applications.
Nevertheless, only a simple system will be applied to this 2. Arduino Uno R3 Microcontroller
project in order to make sure the system is ready to harvest the
The Arduino UNO R3 microcontroller board was based on
maximum output power and reduce the operating energy
the ATmega328 microcontroller chip. The board consists of 14
consumption. The angles and fuzzy logic control will be

140
2015 IEEE Student Conference on Research and Development (SCOReD)

bi-directional digital input and output pins which six pins sensors are placed on the north, south, east and west sides of the
support PWM output, six analog input pins, ICSP header and a solar PV panel. Fig. 2 shows the type of LDR sensor. This
reset button. Basically, this microcontroller board contains sensor detects the intensity of the sunlight and compares the
everything needed to support its operations such as power jack intensity analogue value with a 10 kΩ variable resistor which
supply and USB connection cable. acts as a proportional controller. The configuration of the LDR
sensors is located on the voltage divider circuit, where the
3. Motor Driver Module sensor is fixed in series with a resistor.
A motor driver was used to drive the motor in this project. By The difference in voltage appears at their junction for
controlling the pulse width modulation (PWM) duty cycle, the different levels of light, where the LDR changes its resistive
motor speed can be controlled. The motor driver model is value depends on the light intensity, and the output voltage, Vout,
MD30C. It is designed to drive a medium- to high-power is determined by the voltage divider formula. The resistance of
brushed DC motor. Its current capacity is up to 80 A peak for 1 the LDR can vary from about 100 Ω in the sunlight to over
second and 30 A continuously. It has reverse polarity protection. 100 MΩ in absolute darkness. When the incident sunlight
Its PWM generator enables it to operate without a host intensity increases, Vout also rises to a bigger value. The output
controller. This motor driver has bidirectional control of one voltage of the LDR is calculated according to the voltage
brushed DC motor and the voltage of the motor is 5 to 25 V. Its divider equation:
logic level input is 3.3 and 5 V. It has better efficiency due to its
full NMOS H-bridge, and no heat sink is needed. The PWM R2
Vout  Vin  (1)
frequency of speed control can go up to 20 kHz. It can support RLDR  R2
both locked-antiphase and sign-magnitude of external PWM
operation.
The feedback input from the LDR sensor is calculated
4. Power Window Motor according to the equation:
The power window motor was chosen in this project  2500 
because of its high torque, availability, and cheap price. The   500 
(2)
power window motor functions or is used like a brushless DC Illumination ( Lux) 
 Vout 
motor. It can rotate clockwise (CW) or counter clockwise 3.3
(CCW) by using the motor driver. The specifications of the
power window motor are as follows: the rated voltage is 12
VDC, the no-load speed is 85 rpm, the no-load current is less III. HARDWARE DEVELOPMENT
than 3 A, the load current is not more than 7 A, the lock current The development of the hardware and mechanism is quite
is less than 20 A, and the torque is 30 kg.cm. important in order to make the solar-tracking system robust
In general, a DC motor has two wires which are known as the against the weather’s elements. Firstly, for the mechanism of
positive and negative terminals to distinguish between them. this project, the size of the solar-tracker mechanism was
The motor rotates clockwise when the positive wire of the motor considered; it should be portable and easy to relocate to other
is connected to the positive terminal of the power supply while locations. The tower consists of three main subparts: the base,
the negative wire is connected to the negative terminal of the the vertical axis motor connecter and the shaft holder, and the
power supply. The motor rotates anticlockwise if the wire rotating shaft mounted with the horizontal motor and the solar
polarity is reversed. PV panel on the top, as shown in Fig. 3.
The DC motor consists of two main parts: a stator and a The placement of the LDRs sensors on the solar PV panel
rotor (armature). The armature rotates due to the is shown in Fig. 4. Each LDR sensor was confined inside a
electromagnetic interaction between the stator and the rotor. rectangular pipe to help the collimation of direct solar
The rotation speed is proportional to the voltage applied to the irradiance and to reduce the effect of reflected and diffuse solar
armature. The speed of the motor can be controlled by irradiances. The S1 (west side) and S2 (east side) LDR sensors
regulating the amount of voltage across the motor terminals. were placed on the sides of the solar PV panel to control the
One popular way to achieve voltage regulation is by PWM. The
idea behind PWM speed control is to drive the DC motor using
a series of ON and OFF pulses of voltage level and varying the
duty cycle while keeping the frequency constant. When the duty
cycle is small the average voltage is low, and therefore the
motor speed is small. When the duty cycle is big the average
voltage is high, and therefore the motor speed is high.

5. Light Dependent Resistor (LDR)

Next, four LDR sensors are fixed on top of the solar PV


panel. It is connected to the Arduino UNO analogue pins. The Fig. 2. LDR Sensor

141
2015 IEEE Student Conference on Research and Development (SCOReD)

vertical rotation, and the S3 (north side) and S4 (south side) operator experience, it deals with nonlinearities of the system
LDR sensors were placed on the solar PV panel to control the and Linguistic system definitions can be converted into control
horizontal rotation. Each LDR sensor was placed exactly at the rule base or control algorithm.
midpoint of each side. Furthermore, three limit switches were Every FLC has three main parts that are fuzzification, rule
placed below the solar PV panel. The function of the limit base and defuzzification [6]. Error and change in the error are
switches was to limit the horizontal rotation of the solar PV the inputs to the fuzzy logic controller. Output of the fuzzy logic
panel. Fig. 5 shows the overall electronic circuit design. controller is fed to the power window motor. Fig. 6 shows the
flowchart of the overall system.
A. Fuzzification
IV. IMPLEMENTATION OF FUZZY LOGIC CONTROLLER
Basically, it involves the conversion of the input or output
Intelligent controller is needed in order to rotate the PV panel signals into a number of fuzzy values or fuzzy sets. There is a
according to the sensors output in solar tracking system. Hence degree of membership function for each linguistic term that
in sun tracking intelligent controller like PID controller or fuzzy
logic controller can be used. Fuzzy logic controller is having
advantages over PID controller and these are mathematical
model of the control system is not required, totally depend upon

Fig. 3. Mechanism of Solar Tracker

Fig. 4. Position of LDR Sensors on Solar PV Panel

Fig. 6. Overall Flowchart of the system


Fig. 5. Overall Electronic Circuit Design

142
2015 IEEE Student Conference on Research and Development (SCOReD)

applies to that input variables. Membership functions is of TABLE I. CONTROL RULE BASE FOR THE FUZZY LOGIC CONTROLLER
different type like triangular membership function, trapezoidal E/DE NB NM NS ZE PS PM PB
membership function. For example, triangular membership NB NB NB NB NB NM NS ZE
function for error, change in the error and output is shown NM NB NB NM NM NS ZE PS
in Fig. 7. NS NB NM NS NS ZE PS PM
B. Control Rule Base ZE NB NM NS ZE PS PM PB
PS NM NS ZE PS PS PB PB
The knowledge base of a fuzzy logic controller consists of a
data base and a rule base. The fuzzy control rules in this project PM NS ZE PS PM PM PB PB
are based on fuzzy model of the process, means that the fuzzy PB ZE PS PM PB PB PB PB
control rules IF-THEN was generated in order to track the
maximum intensity of sunlight. This method is somewhat more TABLE II. LABEL DEFINITIONS OF THE LINGUISTIC TERMS
complicated than other approaches, but it yields better Linguistic Term Label
performance and reliability for dealing with nonlinearities of the Negative Big NB
sun movement. Table I shows the control rule base for the fuzzy Negative Medium NM
logic controller. The complete label definitions in Table I is Negative Small NS
listed in Table II. Zero ZE
Positive Small PS
For example:
Positive Medium PM
If ERROR is NB and CHANGE IN ERROR is NM then Positive Big PB
OUTPUT will be NB. Error E
C. Defuzzification Change in Error DE
Generally, this technique converts the fuzzy value into a tracking systems for seven days between 8.00 am and 6.00 pm.
crisp value that is required to actuate the motor to rotate the solar all the data were recorded and graphically represented as
PV panel. There are many methods of defuzzification and in this in Fig. 8. The main parameters for this experiment to calculate
project, the center of gravity method or known as centroid the input power were the voltage across the battery terminal and
method is used because this method is the most prevalent and the current flow through the battery terminal. From the data
physically appealing of all the defuzzification methods. The collected, the input power from the solar PV panel with the
equation of centroid method can be calculated according to the tracking mode is almost always more than the input power with
equation, the non-tracking mode. For one week, the input power from the
x* 
 u ( x).x
x
(3)
tracking mode gradually increased from 7.5 to 9 W between 8.00
 u ( x) x
am and 5.00 pm. The input power decreased dramatically after
5 pm. The input power harvested from the non-tracking mode
Where x* is the defuzzified output, 𝑢𝑥(x) is the aggregated showed a less significant peak at around 12.00 pm.
membership function and x is the output variables.
Fig. 9 shows the graph of input power in watt-hours collected
V. EXPERIMENTAL RESULT by the non-tracking mode and by the solar tracking system with
the intelligent method (tracking mode) during a one-week
The experiments were carried out by comparing the input
operation period. The expected maximum input power in watt-
power collected from solar energy using the tracking and non- hours is 100 Wh. The graph shows that the efficiency of the solar
tracking system with the intelligent method increased compared
to the stationary panel (non-tracking) between day one and day
seven. The input power collected in one week is not constant,
but always varies depending on the weather. On 4 May 2015, the
input power collected showed almost 90% efficiency because
the weather on that day was quite sunny and hot, and as a result
the highest input was collected compared to the other days.
Unlikely, on 6 May 2015, the graph demonstrates that the total
input power collected showed around 75% efficiency, which
was somewhat lower than on other days. The reason is that on
that day the weather was cloudy in the morning and rainy in the
afternoon, and only in the evening was the system able to track
the maximum intensity of sunlight.

Fig. 7. Fuzzy Logic Controller

143
2015 IEEE Student Conference on Research and Development (SCOReD)

Tracking_2 May 2015


9.5
Tracking_3 May 2015
9 Tracking_4 May 2015
Tracking_5 May 2015
Input Power (Watt)

8.5
Tracking_ 6 May 2015
8 Tracking_7 May 2015
7.5 Tracking_8 May 2015
Non-Tracking_2 May 2015
7
Non-Tracking_3 May 2015
6.5 Non-Tracking_4 May 2015
Non-Tracking_5 May 2015
6
8.00 10.00 12.00 14.00 16.00 18.00 Non-Tracking_6 May 2015
Non-Tracking_7 May 2015
Time (Hour)
Non-Tracking_8 May 2015

Fig. 8. Input Power Collected in Seven Days between 8.00 am and 6.00 pm.

100 VI. CONCLUSION


A solar-tracking system with an intelligent method aims to
Total Power (Wh)

90 track the maximum intensity of solar irradiance. Solar energy


has become one of the most effective types of alternative energy
80 because of its availability and cost effectiveness. One of the
challenges of solar energy technology is to increase the
70 maximum output power by collecting as much direct solar
irradiance as possible. Therefore, this system demonstrates a
60 good platform for contributing to the solution to the problem. It
2-May-15 4-May-15 6-May-15 8-May-15 is clearly seen that this system is able to provide a small
improvement of output power efficiency of 18.13% as compared
Date
Tracking Non-Tracking to the stationary panel system. Lastly, it is clearly seen that the
characteristics of the LDR sensors had been described to be
Fig. 9. Total Input Power in Watt-hour Collected for Different Days. sensitive to the sun intensity and the performance of the
application is experimentally testified in a dual-axis solar
Equation (4) is for the improvement percentage of tracking system.
efficiency for harvesting of the solar energy.
REFERENCES
Pt (total) [1] Stamatescu, I., G. Stamatescu, N. Arghira, I. Fagarasan, et al. Fuzzy
  100% (4) Decision Support System for Solar Tracking Optimization. 2014 IEEE
Pt (expected) International Conference on Development and Application Systems
where Pt (total) is the sum of the total input power collected and (DAS). May 2014, Suceava: IEEE. 2014. 16-20.
[2] Ahmad, S., Shafie, S., Kadir, M.Z.A.A. A High Power Generation, Low
Pt (expected) is the expected sum of the total input power, Power Consumption Solar Tracker. 2012 IEEE International Conference
which is equal to 100 Wh. on Power and Energy (PECon). Dec 2012, Kota Kinabalu: IEEE. 2012
366 - 371.
Hence, the percentage efficiency improvement as compared [3] Alexandru, C. The Design and Optimization of a Photovoltaic Tracking
Mechanism. 2009 International Conference on Power Engineering,
to the non-tracking mode is: Energy and Electrical Drives. March 2009. Lisbon: IEEE. 2009. 436–
441.
 ' ' ' [4] Usta, M.A.; Akyazi, O.; Altas, I.H. Design and Performance of Solar
Improvement,%   100% Tracking System with Fuzzy Logic Controller Used Different
' Membership Functions. 2011 7th International Conference on Electrical
and Electronics Engineering (ELECO). Dec 2011. Bursa: IEEE. 2011.
85.54  70.03 381–385.
  100  18.13% [5] Yamin A. H., Ibrahim M. N., Idroas M. and Zin A. R. Embedded Solar
85.54 Tracking Instrumentation System. 2013 IEEE 7th International Power
where η’ is the efficiency of the tracking mode and η’’ is the Engineering and Optimization Conference (PEOCO2013). June 2013.
efficiency of the non-tracking mode. Therefore, it is seen that Langkawi: IEEE. 2013. 223-227.
[6] Stamatescu, I., G. Stamatescu, N. Arghira, I. Fagarasan, et al. Fuzzy
the solar tracking system with the intelligent method is not only Decision Support System for Solar Tracking Optimization. 2014 IEEE
able to collect more sunlight but can also increase the efficiency International Conference on Development and Application Systems
of harvesting of the solar energy by 18.13%. (DAS). May 2014, Suceava: IEEE. 2014. 16-20.

144

You might also like