Experimental performance of pv fed led lighting system employing particle swarm optimization (pso)

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
_______________________________________________________________________________________________
Volume: 03 Issue: 10 | Oct-2014, Available @ http://www.ijret.org 19
EXPERIMENTAL PERFORMANCE OF PV FED LED LIGHTING
SYSTEM EMPLOYING PARTICLE SWARM OPTIMIZATION (PSO)
B. Venkatesh Reddy1
, D. Giri Babu2
1
M. Tech, Department of EEE, ASRA, Hyderabad, Telangana, India
2
M. Tech from National Institute of Technology Tiruchirapalli, Tamil Nadu, India
Abstract
This paper proposes Particle Swarm Optimization (PSO) based approach for light intensity control in a Photo Voltaic (PV)
supply fed LED lighting system. The output power regulation of LED lighting system is formulated as an optimization problem
and the output power regulation is carried out using a real time PSO technique. The concept is first simulated and verified and
subsequently implemented on a 16-bit PIC microcontroller. The proposed method possesses the major advantage of system
independence while regulating the output power. The computed and measured results clearly illustrate the effectiveness of the new
method.
Keywords: Particle Swarm Optimization, Light Emitting Diode and Photo Voltaic System etc…
--------------------------------------------------------------------***---------------------------------------------------------------------
1. INTRODUCTION
With the rapidly increasing energy demand all over the
world, saving lighting energy in building environments has
acquired a new research area since the lighting energy
constitutes about 20% of energy consumption in commercial
establishments [1]. In this context solid state LEDs are
expected to play a significant role in lighting systems due to
the advantages of a Light Emitting Diode (LED) include
long lifetime, low maintenance requirements, improved
luminance, fast response, and smaller size [2]-[5].
Therefore, LEDs have been widely used for various lighting
applications, such as traffic lights, flash lights, automotive
lightings, signage lights, and liquid-crystal display (LCD)
television backlights [6]. The LED driver circuit is an
important component in LED lighting system. The methods
for LED driving are generally classified into three basic
categories [7]-[9]: constant voltage control, constant current
controller, and constant power control. It is well established
in the literature that constant power control of LED is far
superior to constant voltage control and constant current
control as given in [9].
The architecture of existing LED lighting systems consists
of a Photovoltaic (PV) source charging a battery and the
battery is feeding the LED. This type of schematic is best
suitable for applications such as night street lighting.
However, there are several applications where the battery
can be dispensed with and the sun light is directly converted
into artificial light. This scheme is called light-to-light (LtL)
conversion [10]. One typical application is high-energy-
efficient photo catalytic reactors [11]. One can also think of
the use of LtL conversion schemes in indoor applications
during day time such as air-conditioned cabins and
conference halls, decorative lighting schemes in hotels and
marriage halls. In all the above applications the PV can
directly feed the LED system during the day time and
intensity of LED illumination can be adjusted as desired.
While this scheme is simple in its perspective, the design of
the feedback controller for output power regulation is a
challenging task due to the following reasons: (i). The
voltage-current characteristic is nonlinear. (ii). The dc-dc
converter acting as interface between the PV array and the
LED is also a non-linear system. With rapid change in
climatic conditions such as solar insolation and ambient
temperature, the PV array also possess a non-lineart
characteristic.
Particle Swarm Optimization (PSO) is a stochastic
optimization algorithm [12]-[16] and has simple structure,
fast computation capability and can be incorporated in low
cost micro controllers. Another advantage of PSO algorithm
is that it is system independent when employed in real time
applications.
This paper proposes the output power regulation for a light
to light conversion scheme based on Particle Swarm
Optimization (PSO). A boost type dc-dc converter is
interfaced between the PV module and the LED and the
duty ratio of the converter is identified through PSO for a
typical operating point. The PSO code is downloaded to a
16-bit PIC microcontroller and is used for tracking the
reference power. The measured results are given to
substantiate the new method.
2. ECONOMIC ANALYSIS
The present trend is to replace the existing lighting system
with LED lighting system. This section therefore describes
the total power saved by upgrading conventional lighting
technology to LED technology. For a sample test data, LED
lighting is proposed in a conference hall of 9 meter by 4
meter where the conventional fluorescent lamp fixtures were
used to illuminate the hall during daytime. Figure 1 shows
the hall layout where 12 fluorescent lamps, each 75W were
used. With the existing luminaire, the lux distribution is

_______________________________________________________________________________________________
shown in Figure 2. In order to obtain the same lux
distribution and visual comfort, each fluorescent lamp is
now replaced with LED system of 32W. Thus the total
power saved for the conference hall is (75-32)*12=516W.
Accordingly the savings obtained are computed in Tables 1,
2 and 3.
Fig.1. Hall Layout
Fig.2. Lux distribution with existing luminaire
Table 1 Total energy (kilowatt hours, kWh) saved annually
Total kilowatts saved 0.516 kW
Hours of use per day 6 Hrs
Days of use per Week 5
Weeks of use per Year 52
Total kWh saved per Year 804.96 kWh
Table 2 Total energy cost savings per year
Energy cost per kWh Rs. 7
Energy Cost savings per year Rs. 5634.72
Table 3 Energy cost savings over the life of the LED
LED life (hours) 50000 Hrs
Hours per year usage 1560 Hrs
LED est. life 32.05 years
Energy cost savings per year Rs. 5634.72
Lifetime energy cost savings Rs. 1,80,600
Table 4 Parameters of Buck Converter
S.No. Parameter Value
1 Inductance L=1.812mH, rL= 0.394
2 Capacitance C=100F, rC= 0.3
3 Potential divider R1=2.2M; R2 =1.1 M
The details of PV array are given in Table 5.
Table 5: Details of PV array
S.No. Parameter Value
1 Open circuit voltage, (V) 21
2 Short circuit current, (A) 1.3
3 Voltage at maximum power, (V) 17.4
4 Current at maximum power, (A) 2.4
5 Maximum power, (W) 40
6 Maximum system voltage (V) 600
7 Tolerance (%) (+/-) 3
2.1 Payback from Energy Cost Saving
Payback from energy savings is the ratio of LED cost or
difference between LED & original lamp and energy Cost
savings per year = (Rs. 1500 x 12) /Rs. 5634.72 = 3.2 years
Return on Investment (ROI) = 100/ payback = 31.35%
3. EXPERIMENTAL SYSTEM DESCRIPTION
The proposed LED power control scheme is given in Figure
3. It consists of a PV array, a boost converter, LED lighting
system and a feedback system for output power regulation.
The LED current and voltage are sampled at the rate of 100
milliseconds. A multiplier is now used to compute the LED
power, p0 as in Figure 3. If there is an error between the
measured power output and the reference value, the particle
swarm optimization algorithm is made to run on line. The
program is terminated when the displacement among the
particles reaches a set minimum value. Then the dc-dc
converter is activated with this optimum duty ratio.
Fig.3. Constant Power Control scheme
4. PARTICLE SWARM OPTIMIZATION
The particle swarm optimization (PSO) algorithm belongs to
the category of Swarm Intelligence methods [12-16]. PSO is
a population-based algorithm that exploits a population of
individuals to probe promising regions of the search space.

_______________________________________________________________________________________________
In this context, the population is called a swarm and the
individuals are called particles. PSO algorithm iteratively
explores a multidimensional search space with a swarm of
particles, looking for the global minimum (or maximum).
Each particle “flies” through the search space according to
its velocity vector. In every iteration, the velocity vector is
adjusted so that prior best positions (cognitive aspect) and
the best position found by particles within a specific
neighborhood (social aspect) act as attractors. In the original
PSO algorithm, the neighborhood of a particle consists of all
particles, so that the global best position, i.e., the best
solution found so far, directly influences its behaviour. Each
particle moves with an adaptable velocity within the search
space, and retains in its memory the best position it has
encountered so far. In PSO, the agents are considered to
have the ability to memorize the best position they have
attained.
The steps of PSO technique for LED power control are
given below.
1. Initialize the parameters maximum number of
iterations N, population size n, acceleration parameters
1c and 2c and inertia weight w .
2. Randomly generate n particles.
3. Initialize for each particle, the pbest as current
location.
4. While ( Ni  )
4.1. While  nj 
4.1.1. Calculate the fitness value of the particle.
4.1.2. If current fitness of the particle is better
than previous best position, update
the pbest.
4.1.3. 1 jj .
4.2. If the fitness of the best particle is better than the
global best position, gbest reached so far, update
the global best position, gbest.
4.3. 1 ii .
5. End.
The output power control of the LED lighting system is
operated as given in the flow chart in Figure 4.
Fig.4. Flow chart
5. PSO ALGORITHM FOR CONSTANT POWER
CONTROLLER
This section explains the implementation of Particle Swarm
Optimization (PSO) based algorithm for regulating PV fed
LED output power. A boost type dc-dc converter is used as
the power converter and PIC microcontroller is used to
generate pulses for the converter. The circuit diagram of
constant power control scheme for Photo Voltaic (PV) array
fed LED is shown in Fig. 3. It employs a PV array as source,
a boost converter comprising a power electronic switch such
as IGBT, a power diode, an inductor, and a capacitor, a
current sensor to sense the current flowing through the LED
array, a potential divider to sense the voltage across the LED
array, a multiplier to get the product of voltage across and
current through the LED array, and a microcontroller to
generate pulses to the boost converter.

_______________________________________________________________________________________________
Fig.5. Constant power controller using PSO
The output voltage across the LED array is measured and is
given as one of the inputs to multiplier IC through the
potential divider forming with R1 and R2. The other input to
the multiplier IC is the current flowing though the LED
array. Now, the multiplier IC gives the output power (p0) of
the LED array which is given to the microcontroller. To set
the required power, i.e., set power (Pset) is also given in
terms of voltage to the microcontroller which uses particle
swarm optimization (PSO) algorithm to process and gives
pulses to the boost converter to maintain constant power to
the LED array. The PIC16F876A microcontroller is used for
producing firing pulses to the boost converter. This IC is
available in 28-pin package. This device has three I/O ports
and five A/D input channels and A/D conversion module is
in-built.
A dedicated program for particle swarm optimisation based
power regulation is developed in MPLAB and then
downloaded to the PIC microcontroller for the closed loop
operation. The experimental results are explained in the
following section.
6. EXPERIMENTAL RESULTS
The PV array fed LED lighting system was fabricated in the
laboratory. The specifications of the experimental system
are given in Table 4. Experiments were conducted at
different operating points of the LED lighting system.
Figure 6(a) shows the output power response of the system
for a step input of amplitude 12W. It is seen that the power
output oscillates evidently due to random initial positions of
the particles. As the iteration proceeds, the particles do come
closer to each other and the PSO algorithm is terminated
when the position of the best particle corresponds to the
reference value of 12W. The response time is approximately
5 seconds.
After 20 seconds, the reference power is now increased to
24 W. The microcontroller senses the reference value and
the PSO algorithm is re-started again. It is seen that the
power output oscillates but settles to the reference value less
than 5 seconds. The experimental results given in Figure
6(a) clearly illustrate the tracking capability of the proposed
algorithm.
During day time the solar insolation as well as ambient
temperature will change due to environmental changes:
sometimes these changes may be drastic such as a cloudy
day or a very hot day in summer. The proposed controller
must maintain the required power level in the conference
hall for the efficiency of the working staff inside the
conference hall. In order to test the controller with solar
insolation change, artificial shading of the solar panel is now
performed and this is shown in Figure 6(b) with reduced
solar insolation the LED output power comes down forcing
the microcontroller to initialise the PSO algorithm. As can
be seen from Figure 6(b), the algorithm effectively identifies
the new duty ratio so as to meet the reference power with
decreased insolation. The search time is observed to be 5
seconds within which the dc-dc converter regulates the
output power to 12W. The artificial shading is now removed
increasing the solar insolation. This increases the output
power from the PV array and hence the PSO algorithm is
started once again to regulate the output power. The
required regulation time is around 5 seconds. The
performance curves in Figure 6(b) illustrate the robust
nature of the PSO algorithm in rejecting external as well as
internal disturbances.
(a)
(b)
Fig.6. Experimental Results
AN0
AN1

_______________________________________________________________________________________________
7. CONCLUSION
This paper has illustrated a new method of output power
regulation in LED lighting systems employing Particle
Swarm Optimization approach. The proposed method is
simple to implement and is also system independent. The
measured results for different types of internal as well as
external disturbances clearly illustrate that the proposed
algorithm is robust and immune to various disturbances.
REFERENCES
[1]. M. S. Shur and A. Zukauskas, “Sokid-state lighting:
Toward superior illumination,” Proc. IEEE, vol. 93, no. 10,
pp. 1691-1703, Oct. 2005.
[2]. H.J. Chiu and S.J. Cheng, “LED backlight driving
system for large-scale LCD panels,” IEEE Trans. Ind.
Electron., vol. 54, no. 5, pp. 2751-2760, Oct. 2007.
[3]. N. Narendran and Y. Gu, “Life of LED-based white
light sources,” J. Display Technol., vol. 1, no. 1, pp. 167-
171, Sep. 2005.
[4]. I. Hino and K. Iwamoto, “LED pulse response analysis
considering the distributed CR constant in the peripheral
junction,” IEEE Trans. Electron. Devices, vol. ED-26, no. 8,
pp. 1238-1242, Aug. 1979.
[5]. S. D. Lester, M. J. Ludowise, K. P. Killeenm, B. H.
Perez, J. N. Miller, and S. J. Rosner, “High-efficiency
InGaN MQW blue and green LEDs,” J. Cryst. Growth,
vol.189/190, pp. 786-789, Jun. 1998.
[6]. H. J. Chiu, Y. K. Lo, T. P. Lee, S. C. Mou, and H. M.
Huang, “Design of an RGB backlight circuit for liquid
crystal display panels,” IEEE Trans. Ind. Electron., vol. 56,
no. 7, pp. 2793-2795, Jul. 2009.
[7]. G. Carrato, “Solving high-voltage off-line HB-LED
constant-current control circuit issues,” in Proc. IEEE Appl.
Power Electron. Conf. (APEC), 2007, pp. 1316-1318.
[8]. Y. Hu, and M. M. Jovanovic, “LED driver with self-
adaptive drive voltage,” IEEE Trans. Power Electron., vol.
23, no. 6, Nov. 2008.
[9]. B. J. Huang, C. W. Chen, P. C. Hsu, W. M. Tseng, and
M. S. Wu, “Direct battery-driven solar LED lighting using
constant-power control,” Solar Energy, 86 (2012), pp. 3250-
3259.
[10]. Femina, N., Fortunato N., Vitelli, M., “Light to Light,
PV fed lighting systems,” IEEE trans. Power Electronics,
Vol.28, Iss.8 (2013), pp. 4063-4073.
[11]. D. H. Chen, X. Ye and K. Li "Oxidation of PCE with
a UV LED photocatalyst reactor", Chem. Eng. Tech., vol.
28, no. 1, pp.95 -97 2005
[12]. J. Kennedy, and R. Eberhart, “Particle Swarm
Optimization,” in Proc. of the IEEE Int. Conf. on Neural
Networks, Pisctaway, NJ, pp. 90-97.
[13]. H. Xiaohui, S. Yuhui, and R. Eberhart, “Recent
advances in particle swarm,” in Proc. Congr. Evolut.
Comput., 2004, vol. 1, pp. 90-97.
[14]. Y. Shi, and R. Eberhart, “A modified particle swarm
optimizer,” in Proc. IEEE World Congr. Comput. Intell.,
Piscataway, NJ, pp. 69-73, 1998.
[15]. A. Ratanaweera, Saman K. Halgamuge, Hary C.
Watson, “Self organizing hierarchical particle swarm
optimizer with time-varying acceleration coefficients,”
IEEE Trans. Evol. Comput., vol. 8, no. 3, Jun. 2004.
[16]. P. N. Suganthan, “Particle swarm optimizer with
neighbourhood operator,” Proc. Evol. Comp., 1999.
[17]. J. Robinson, Y. Rahamat Samii, “Particle Swarm
Optimization in electromagnetics,” IEEE Trans. Antennas
and Propagation, vol. 52, no. 2, pp. 397-407, Feb. 2004.

Experimental performance of pv fed led lighting system employing particle swarm optimization (pso)

More Related Content

What's hot (18)

Viewers also liked (17)

Similar to Experimental performance of pv fed led lighting system employing particle swarm optimization (pso) (20)

More from eSAT Journals (20)

Recently uploaded (20)

Experimental performance of pv fed led lighting system employing particle swarm optimization (pso)