MICROCONTROLLER BASED INTELLIGENT DC/DC CONVERTER TO TRACK MAXIMUM POWER POINT FOR SOLAR PHOTOVOLTAIC MODULE Siwakoti, Yam Prasad1, Bhupendra Bimal Chhetri2*, Brijesh Adhikary2†, Diwakar Bista2‡ 1. Department of Electric Power Engineering, NTNU, Norway/ Department of Electrical and Electronics Engineering, Kathmandu University, Nepal E-mail: [email protected] 2. Department of Electrical and Electronics Engineering, Kathmandu University, Nepal E-mail: *[email protected], †[email protected], ‡[email protected]

Abstract— Maximum Power Point Tracking (MPPT) is widely used control technique to extract maximum power available from the solar cell of photovoltaic (PV) module. Since the solar cells have non-linear i-v characteristics. The efficiency of PV module is very low and power output depends on solar insolation level and ambient temperature, so maximization of power output with greater efficiency is of special interest. Moreover there is great loss of power due to mismatch of source and load. So, to extract maximum power from solar panel a MPPT needs to be designed. The objective of the paper is to present a novel cost effective and efficient microcontroller based MPPT system for solar photovoltaic system to ensure fast maximum power point operation at all fast changing environmental conditions. The proposed controller scheme utilizes PWM techniques to regulate the output power of boost DC/DC converter at its maximum possible value and simultaneously controls the charging process of battery. Incremental Conductance algorithm is implemented to track maximum power point. For the feasibility study, parameter extraction, model evaluation and analysis of converter system design a MATLAB/Simulink model is demonstrated and simulated for a typical 40W solar panel from Kyocera KC-40 for hardware implementation and verification. Finally, a hardware model is designed and tested in lab at different operating conditions. Further, MPPT system has been tested with Solar Panel at different solar insolation level and temperature. The resulting system has high-efficiency, lower-cost, very fast tracking speed and can be easily modified for additional control function for future use.

development done in the area of solar cells and power electronic converter technology to reduce the cost and increase the system efficiency. Solar photovoltaic generation systems have two inherent major problems. The first is low conversion efficiency (10 to 16% efficiency for commercially available amorphous silicon solar cells). Second is presence of highly nonlinear i-v characteristics. The problem gets worse due to dependence of characteristics of solar cell on temperature and insolation level [1]. Further, due to mismatch between the operating point and Maximum Power Point (MPP) of the solar cells, the power available from the solar cells is not always fully extracted [2]. In order to extract the maximum amount of energy, the PV system must be capable of tracking the solar panel unique maximum power point that varies with irradiance and temperature. A MPPT is a power electronic DC/DC converter inserted between the PV module and load to achieve optimum matching. By using an intelligent algorithm, it ensures that the PV module always operates at its maximum power point. Several MPPT algorithms have been proposed in different literatures like Perturb & Observe (P&O), Incremental Conductance, Constant Voltage, Constant Current and fuzzy based algorithms [3]. These techniques differ in many aspects like complexity, convergence speed, hardware implementation, sensors required, cost, range of effectiveness and need for parameterization. The P&O and Incremental Conductance algorithms are more common due to effectiveness of extracting maximum power from the panel, ease of hardware implementation, and less sensor requirement and consequently relative low cost [1], [2] & [4]. But is has problem of oscillation around the MPP, due to this there is considerable loss of power. Also the response of P&O algorithms is slow under fast changing environmental condition [2] & [3]. So in this paper Incremental Conductance based algorithm is implemented in MATLAB/Simulink to track maximum power point.

Key words-MPPT Techniques, Photovoltaic Module, Microcontroller, PWM Techniques, Boost DC/DC Converter.

1. INTRODUCTION Global climatic change has become serious concern these days triggered by greenhouse gas emission due to production of energy from conventional sources of energy. Renewable energy especially solar photovoltaic is seen as important alternative source of energy for the future, and has recently attained importance due to lot of research and

978-1-4244-6078-6/10/$26.00 ©2010 IEEE

Different converter topologies like buck, boost, buck-boost and cuk converters are implemented for MPPT design. Boost converter can track MPP with maximum efficiency and can work for wide range of input voltage [2]. However, other advantages like less component needed for hardware

94

implementation and cost, makes this topology a better choice than the buck or buck-boost for MPPT system design in this paper.

dV

R s = − − dI

In this paper duty cycle of boost DC/DC converter is controlled by PWM signal from microcontroller implementing Incremental Conductance algorithm. Firstly, a model of MATLAB PV model under different temperature and insolation is simulated, tested and verified. Further, a boost converter and battery is modelled and tested under different conditions. Whole MPPT system is simulated using real environmental data and battery conditions. Result validates the good performance of the control technique and improved efficiency for hardware implementation and realization. Finally, a hardware model is designed and tested in lab at different operating conditions. Further, MPPT system has been tested on Solar Panel at different solar insolation level and temperature. The resulting system has high-efficiency, lower-cost, very fast tracking speed and can be easily modified for additional control function for future use.

2. MODELLING AND VALIDATION CELL IN MATLAB

OF

(6)

q (V + IRs ) I 0 e nkT − 1

Figure 1: Equivalent circuit used for MATLAB Simulation Table 1: Electrical characteristics data of Kyocera KC-40 S.No. 1 2 3 4

SOLAR

A model of moderate complexity is chosen for the MATLAB modelling and simulation. A simple PV cell is represented by a light dependent current source (Iph) in antiparallel with a diode driven by current Id (Shockley diode equation) [2] and a series resistance in the current path through the semiconductor material, the metal grid, contacts, and current collecting bus [5] & [7]. The presence of diode determine the output V-I characteristics of solar cell.

Electrical Specification Maximum Power Voltage Maximum Power Current Open Circuit Voltage Short-circuit Current

16.9V 2.34A 21.5V 2.48A

*Note: The electrical specifications are under test conditions of irradiance of 1kW/m2, spectrum of 1.5 air mass and cell temperature of 250C

Kyocera KC-40 PV module is chosen for a MATLAB simulation model. The module is made of 36 multicrystalline silicon solar cells in series and provides 40W of nominal maximum power. All the constants in above equations are determined by examining the datasheet provided by manufacturer. Table 1 shows its electrical specification that is necessary for cell/module modelling.

The parallel resistance associated with a small leakage of current through a resistive path in parallel with the intrinsic device [5], [6] & [7] is very large, so its effect is very less and is neglected. The equations which govern the characteristics of solar cell are: + IRs ) q (VnkT I = I sc − I 0 e − 1

nkT q

Above characteristics equations were solved with MATLAB using Newton Raphson method as it converge more rapidly to find the output current [2]. A pre-simulation result is shown in figure 2 for model validation.

(1) V-I Curve at STC (KYOCERA KC40) Pmpp = 39.35 W ( Error = -1.6 %)

2.48

I sc

G

=

G I sc G0 3

I0

T

(2)

G0

qE g 1

1

T n − nk T − Tref e = I 0 Tref Tref I sc T = I sc Tref 1 + a (T − Tref

[

Module Current (A)

2.3239

2

Impp = 2.32 A (Error = - 0.7%) 1.5

1

Vmpp = 16.93 V (Error = + 0.2%) 0.5

(3)

0

0

5

10

15

16.9339

21.5

Module Voltage (V)

)]

(4)

q (V + IRs ) V + IRs I = I sc − I 0 e nkT − 1 − Rp

(5)

Figure 2: I-V Characteristics of Solar PV from MATLAB Model Comparing the manufacturers’ datasheet and simulated result, it is found that the MATLAB model thus designed shows excellent correspondence. So, same model is used in this paper for further analysis.

95

The I-V curves and PV curves at different temperature and insolation are shown in figure 3 and 4.

is implemented because of its ease of digital implementation, fast response and efficiency. Incremental Conductance algorithm is based on differentiation of PV power to its voltage and on condition of zero slope of P – V curve at maximum power point (MPP). Power output from solar PV panel is: P = VI (7) Where P = module power, V = module voltage and I = module current; Differentiation with respect to V gives dP/dV = I+V(dI/dV) (8) The algorithm works on this basis. At peak power point, dP/dV = 0 (9) dI/dV = -I/V (10)

50 45

Module Current (A)

40

50 C

35

0C

30

1000 W/m2

25

25 C

75 C

20 15 10

500 W/m2 5 0

0

5

10

15

20

25

Module Voltage (V)

Figure 3: i-v curve at different environmental conditions 3

1000 W/sq.m

50

dP/dV = 0

2.5 45

Module Current (A)

0C

40

Module Power (W)

2

50 C

1.5

500 W/sq.m

25 C

1

75 C

35 30

dP/dV > 0

25

dP/dV < 0

20 15

0.5 10 5

0

0

5

10

15

20

25 0

Module Voltage (V)

10

15

20

If the operating point is on the right of the power curve we have dP/dVI/V

Locus of MPP at different Temp. (KYOCERA KC40) 45

40

G=1 T = 0 - 75 C

30

Power (W)

5

25

Figure 7: Operating point of Incremental Conductance algorithm on P-V curve.

The locus of Maximum Power Point (MPP) at different temperature and insolation is shown in figure 5 and 6. The effect of irradiance on power output is much higher than the effect of temperature. The maximum power point fall steeply as the irradiance level fall in case of irradiance change.

35

0

Module Voltage (V)

Figure 4: p-v curve at different environmental condition

25

20

then (11) (12) then (13) (14)

15

10

3. MPPT ALGORITHM AND BOOST CONVERTER

The present value and the previous value of the solar array voltage and current are used to calculate the values of dI and dV. If dV=0 and dI=0, then the atmospheric conditions have not changed and the MPPT is still operating at the MPP. If dV=0 and dI>0, then the amount of sunlight has increased, raising the MPP voltage. This requires the MPPT to increase the PV array operating voltage to track the MPP. Conversely, if dI - I/V, then dP/dV > 0, and the PV array operating point is to the left of the MPP on the P–V curve. Thus, the PV array voltage must be increased to reach the MPP. Similarly, if dI/dV < - I/V, then dP/dV < 0 and the PV array operating point lies to the right of the MPP on the P–V curve, meaning that the voltage must be reduced to reach the MPP.

MPPT optimize the power output of the solar cell by constantly tracking the varying maximum power operating point and adjusts the solar panel operating voltage. Different algorithms are used to track the maximum power point [3] & [4]. In this paper Incremental Conductance based algorithm

A small marginal error could be added to the maximum power condition such that the MPP is assumed to be found if [dI/dV + I/V] < . The value of was determined with consideration of the trade-off between the problem of not

5

0

0

5

10

15

20

25

Voltage (V)

Figure 5: Effect of temperature on locus of MPP Locus of MPP at different Irradiance (KYOCERA KC40) 45 40

T = 25 C G = 0.2 - 1.2

Power (W)

35 30 25 20 15 10 5 0

0

5

10

15

20

25

Voltage(V)

Figure 6: Effect of irradiance on locus of MPP

96

operating exactly at the MPP and the possibility of oscillating around it. It will also depend on the chosen perturbation step size C.

instantaneous maximum power from solar panel, power that can be transferred from the panel under direct coupling and maximum power that can be extracted from the panel using MPPT can be seen in Scope-3. Scope-2 plot battery current, solar current and load current. Negative battery current indicates battery is under discharge mode and positive indicates charging mode. If the load is high then battery and solar panel supply to load and if the load is light then solar panel supply current to load as well as it charge the battery simultaneously.

These changes in PV’s voltage may be done by coupling a DC/DC converter to PV and controlling properly its duty cycle (D) to track the maximum power point [8]. Most MPPT charge controllers are based on either the buck converter (step-down), boost convert (step-up) or buck-boost converter setup. In this paper boost converter is used because of its higher efficiency, reduced cost for realization and have large working voltage range [2].

4. SIMULATION MODEL

OF

State of Charge, battery voltage and battery current can be seen in Scope1. Different instantaneous values can be seen in display. For analysis of power under different condition is transferred to workspace. Current can be measured with current meter ‘CM2’. Similarly voltage with voltage meter ‘VM1’. Switching for various loads is performed by load switch controller.

COMPLETE MPPT MATLAB

The complete MPPT model is shown in figure 8. Three PV array with same characteristics is taken as current source for different measurement. First block ‘KYOCERA KC40-1’ is used to measure theoretical maximum power available from the panel under specified temperature and insolation level. Second block for panel ie. ‘KYOCERA KC40-2’ is used for analysis of power transfer for direct coupling to the load. Program for the third panel is written in the IC_algo’ block. This block is use for analysis of maximum power that can be extracted from the panel employing different algorithms. The battery is represented by a voltage source and the load as resistors.

5. SIMULATION RESULT AND DISCUSSION Simulation was run for 12 hour data (6am to 6pm) to see the output under real insolation and temperature. The graph shows the simulated result of PV panel under real insolation and temperature available from the solar lab of Kathmandu University. The increase in power due to MPPT is clearly seen from the above simulated result. The maximum power mismatch is about 7W. Theoretical maximum energy available = 357.45 Wh Energy extracted by MPPT = 346.66 Wh Direct battery coupled energy transferred = 296 Wh So, there is 50Wh loss of energy in a day without MPPT. These losses increase if the battery is in fully discharged condition and in cloudy day. The numerical result is shown in table 2.

For the present work Incremental Conductance algorithm is used to extract maximum power from the array. IMPP and VMPP are input to the boost converter. The change in duty cycle is necessary to control the output voltage. Manual switch 1 and 2 are used to run simulation under real data or any specific data. PV SW and Load SW are manual switch for solar panel and load side of battery. Plot of theoretical

Figure 8: Complete MATLAB/Simulink Block diagram

97

Theoretical MPP

Direct Battery Coupled Power MPPT

Figure 9: Simulated result Figure 10 shows the block diagram of the proposed MPPT system.

Table 2: Simulated result under real data and under STC. Energy

Real data (Wh)

At STC (Wh)

Theoretical Max. Energy (ETh)

357.45

800

MPPT Energy (EMPPT)

346.66

776.73

Energy transferred to Battery (EBatt)

296.00

615.14

Loss (EMPPT-EBatt)

50.66

161.59

6. HARDWARE DESIGN Figure 10: Maximum Power Point Tracking System block diagram

Simulation result shows the good performance of the control technique and improved efficiency for hardware implementation and realization. The important things to be considered for MPPT System design are the switch-mode topology and the control mechanism. The switch-mode topology is usually determined by the input and the output voltages desired. The tracker can be designed to either increase (known as a boost topology) or decrease voltage (known as a buck topology) from the array going into the battery or load. Because a tracker is essentially a specialized switching power supply, the input and output currents are such that the power into and out of the tracker are equal. In this research boost converter topology is chosen that will maximize the efficiency of the PV array. Boost converter are technically easier to build and also guarantee continuous current through the array.

6.1

Functional Specifications

High Frequency Boost Converter—A boost converter operating at high frequency (31 KHz) is designed which is controlled by the PWM signal from the microcontroller. Duty cycle is changed by microcontroller according to the temperature and insolation level. The microcontroller tries to maximize the watts output from the solar panel by controlling the step up ratio to keep the solar panel operating always at its Maximum Power Point.

The control section design involves the design of both analog and digital system. It will take in analog voltages and currents proportional to measured quantities, digitize them, process them in a microcontroller, and then convert a number back to a voltage proportional to what the system believes is the maximum power point voltage of the array and the state of charge of the battery.

Figure 11: Boost converter

98

Microcontroller—The DC/DC converter is controlled by the microcontroller. It calculates the solar watts generated by reading the voltage and current of the solar panels through the A/D port. It also calculate the battery side voltage and current in same way and send corresponding control signal to the converter and control the duty cycle of the converter to increase, decrease or turn off the converter accordingly. The ATMEGA16 is a perfect combination of features, performance, and low power consumption for this application.

modes and is given as: L = [(Ts Vo)/(2 Io,max)] D(1-D)2

The value of inductor should be greater than the calculated value for continuous conduction mode operation of converter [8]. MOSFET—The main criteria for selection of MOSFET is low switching power loss and able to handle the worst case current and voltage stresses i.e. it able to handle current and voltage of solar panel working in most favorable environmental condition (T = 0oC or low and G = 1kW/m2 or high). N-channel enhancement mode MOSFET IRFZ44N from Philips Semiconductor is chosen for best design. [8].

Voltage and Current Sensor—The current and voltage signals from the PV array are monitor by using the current and voltage sensing circuit. The sensing circuit that acts as transducer-converts the current reading to the voltage signals so that the microcontroller can understand and use the information to process and hence performs the PWMs. A simple voltage divider network to sense the voltage and a Hall Effect-Based Linear current sensor ACS714 from Allegro Microsystem Inc. is used.

MOSFET Driver—The PWM generated by microcontroller is not sufficient to drive the large capacitive loads such as MOSFET with high slew rate. The signals generated by the microcontroller can only deliver maximum current of 25mA. In order to achieve high speed switching in power MOSFET, a MOSFET driver chip TPS2812D was chosen.

Power Supply—Power necessary for microcontroller and other peripheral devices to work is taken from the solar panel itself using a suitable regulator. 6.2

Component Selection

7. EXPERIMENTAL RESULT

Since the objective of the tracker is to deliver maximum power from the solar module to the load, power losses associated with the tracker itself should be minimal. So, one of the very important factor to consider while selecting the component is the low power loss [8]. The other selection criteria associated with each major component are:

Experiment is carried out in a sunny day (29 May 2010) from 1100Hrs to 1300Hrs and data were taken for analysis. Data trackers log the necessary data for further analysis. Panel is fixed in adjustable frame to change the direction of panel for insolation control. The panel is kept initially at maximum insolation ie. horizontal position. Data was logged for 20 minutes in the same position. After every 20 minutes in first one hour of observation period, the insolation is decreased by tilting the panel away from the sun. Similarly, for next one hour, insolation is increased from minimum testing level to maximum level at 20 minutes step time and data is logged for further analysis.

Diode—Fast switching and low forward voltage drop Schottky diode (MBR745) with ability to block the required off–state voltage stress and have sufficient peak and average current handling capability. Capacitor—The primary criterion for selecting the output filter capacitor is its capacitance and equivalent series resistance (ESR). Since the capacitor’s ESR affects efficiency, low-ESR capacitors will be used for best performance. The output filter capacitors are chosen to meet an output voltage ripple specifications, as well as its ability to handle the required ripple current stress. An approximate expression [8] for the required capacitance as a function of ripple voltage requirement, Vo, D, switching frequency, Fs and output voltage, Vo is given as: C = VoD/(Fs Vo R)

(16)

The variation of efficiency of the MPPT system for the observation period is shown in figure 15. It is observed that the MPPT system is working at high efficiency during high irradiance level than those at low irradiance level. The decrease in system efficiency at low irradiance level is due to the constant losses in the converter section. The variation in efficiency for the total observation period is due to variation in duty cycle for converter MPPT operation. The performance of the algorithm and the system were found good for changing irradiance and hence temperature of the solar panel.

(15)

Inductor—Inductor with a ferrite core and low DC resistance is suitable for boost converter. Inductance is selected based on the maximum allowed ripple current at minimum duty cycle, D, at maximum output current Io,max. The critical inductance is defined as the inductance at the boundary edge between continuous and discontinuous

Solar Panel is tested with and without MPPT system at different environmental condition with a resistive load (bulb) of 2.2. It is observed from the figure 14 that the power extracted by peak power tracker is more than (nearly about two times) that without the MPPT system.

99

Pyranometer

Temperature Sensor

Kyocera KC-40 Figure 13: Experimental hardware setup

Figure 12: Solar Panel from Kyocera KC-40 used for experimental analysis

Figure 14: Comparison of module power with and without MPPT system

Figure 15: Efficiency of MPPT system

100

8. CONCLUSION

Sons, Ltd., Research and Applications, Prog. Photovolt: Res. Appl. 2003.

In this paper, MPPT is implemented by using a boost DC/DC converter, which is designed to operate under continuous conduction mode and a microcontroller using Incremental Conductance algorithm to control the PWM signals to the boost converter. An experimental result shows the system has good conversion efficiency of 92%. Even small improvement of efficiency could bring large saving, if the system is large e.g. Grid Connected System.

[4] D. P. Hohm, M. E. Ropp, “Comparative Study of Maximum Power Point Tracking Algorithms Using an Experimental, Programmable, Maximum Power Point Tracking Test Bed”, Photovoltaic Specialists Conference, Conference Record of the Twenty-Eighth IEEE Volume , Issue , 2000 Page(s):1699 – 1702. [5] Martin A. Green, Solar Cells: Operating Principles, Technology, and System Applications, Publisher: Prentice Hall (October, 1981) ISBN: 0138222703.

The designed intelligent MPPT system thus increase the power output from the solar PV drastically. The effect of variation of temperature and insolation in power output is observed and analyzed for MPPT design. Power transferred from the solar panel is affected by temperature and insolation variation which can be minimized by using MPPT system.

[6] Messenger, Roger & Jerry Ventre, Photovoltaic Systems Engineering 2nd Edition, CRC Press, 2005. [7] Ralph Gottschalg, Photovoltaic Technology for Bangladesh, Publisher: Department of Mechanical Engineering, BUET and CREST, Loughborough University, UK. March 2001.

This research presents a simple but cost effective control technique to harness solar energy optimally. It models each component of PV system and simulates in MATLAB to study the feasibility of the system under real weather conditions. Solar cell model using the equivalent circuit of moderate complexity was chosen for simulation and found to be good matching with the PV model. Simulation is carried out for different insolation and temperature for different interval of time. From the simulation result it is found that the use of MPPT system is more effective in the changing weather conditions and in cold places where we can extract extra power from the solar panel.

[8] N.Mohan, T.M Undeland, and W.P. Robbins, Power Electronics: Converters, Application and Design, John Wiley & Sons, Inc, New York, 2003. [9] Luis Castañer, Santiago Silvestre, Modelling photovoltaic systems using PSpice, John Wiley and Sons, January 2003. [10] Yen-Jung Mark Tung, Dr. Aiguo Patrick Hu, Dr. Nirmal-Kumar Nair, “Evaluation of Micro Controller Based Maximum Power Point Tracking Methods Using dSPACE Platform”, Australian University Power Engineering Conference 2006.

The hardware system is designed, tested and found to be satisfactorily working with reasonably higher efficiency at different operating conditions. The performance of system is tested in real sunny day and emulated to cloudy day for analysis. Incremental Conductance algorithms show better performance in terms of energy extraction under rapidly changing weather conditions. Efficiency of the system even under low irradiance, high temperature and fast changing operating conditions was found reasonably good.

[11] Eftichios Koutroulis, Kostas Kalaitzakis, “Development of a Microcontroller-Based, Photovoltaic Maximum Power Point Tracking Control System”, IEEE transactions on Power Electronics, Vol. 16, No .1, January 2001. [12] Francisco M. González-Longatt, “Model of Photovoltaic Module in Matlab™”, 2DO CONGRESO IBEROAMERICANO DE ESTUDIANTES DE INGENIERÍA ELÉCTRICA, ELECTRÓNICA YCOMPUTACIÓN (II CIBELEC 2005).

REFERENCES [1] Roberto faranda, Sonia Leva “Energy comparison of MPPT techniques for PV System”, WEAS Transaction on Power Systems, ISSN: 17905060, Issue 6, Volume 3, June 2008. [2] Geoff Walker “Evaluting MPPT converter topologies using a MATLAB PV model”, Journal of Electrical & Electronics Engineering, Australia; Volume 21, Issue 1; 2001; 49-55. [3] D. P. Hohm and M. E. Ropp, “Comparative Study of Maximum Power Point Tracking Algorithms”, Progress in Photovoltaics: Copyright-2002 John Wiley &

101

Abstract— Maximum Power Point Tracking (MPPT) is widely used control technique to extract maximum power available from the solar cell of photovoltaic (PV) module. Since the solar cells have non-linear i-v characteristics. The efficiency of PV module is very low and power output depends on solar insolation level and ambient temperature, so maximization of power output with greater efficiency is of special interest. Moreover there is great loss of power due to mismatch of source and load. So, to extract maximum power from solar panel a MPPT needs to be designed. The objective of the paper is to present a novel cost effective and efficient microcontroller based MPPT system for solar photovoltaic system to ensure fast maximum power point operation at all fast changing environmental conditions. The proposed controller scheme utilizes PWM techniques to regulate the output power of boost DC/DC converter at its maximum possible value and simultaneously controls the charging process of battery. Incremental Conductance algorithm is implemented to track maximum power point. For the feasibility study, parameter extraction, model evaluation and analysis of converter system design a MATLAB/Simulink model is demonstrated and simulated for a typical 40W solar panel from Kyocera KC-40 for hardware implementation and verification. Finally, a hardware model is designed and tested in lab at different operating conditions. Further, MPPT system has been tested with Solar Panel at different solar insolation level and temperature. The resulting system has high-efficiency, lower-cost, very fast tracking speed and can be easily modified for additional control function for future use.

development done in the area of solar cells and power electronic converter technology to reduce the cost and increase the system efficiency. Solar photovoltaic generation systems have two inherent major problems. The first is low conversion efficiency (10 to 16% efficiency for commercially available amorphous silicon solar cells). Second is presence of highly nonlinear i-v characteristics. The problem gets worse due to dependence of characteristics of solar cell on temperature and insolation level [1]. Further, due to mismatch between the operating point and Maximum Power Point (MPP) of the solar cells, the power available from the solar cells is not always fully extracted [2]. In order to extract the maximum amount of energy, the PV system must be capable of tracking the solar panel unique maximum power point that varies with irradiance and temperature. A MPPT is a power electronic DC/DC converter inserted between the PV module and load to achieve optimum matching. By using an intelligent algorithm, it ensures that the PV module always operates at its maximum power point. Several MPPT algorithms have been proposed in different literatures like Perturb & Observe (P&O), Incremental Conductance, Constant Voltage, Constant Current and fuzzy based algorithms [3]. These techniques differ in many aspects like complexity, convergence speed, hardware implementation, sensors required, cost, range of effectiveness and need for parameterization. The P&O and Incremental Conductance algorithms are more common due to effectiveness of extracting maximum power from the panel, ease of hardware implementation, and less sensor requirement and consequently relative low cost [1], [2] & [4]. But is has problem of oscillation around the MPP, due to this there is considerable loss of power. Also the response of P&O algorithms is slow under fast changing environmental condition [2] & [3]. So in this paper Incremental Conductance based algorithm is implemented in MATLAB/Simulink to track maximum power point.

Key words-MPPT Techniques, Photovoltaic Module, Microcontroller, PWM Techniques, Boost DC/DC Converter.

1. INTRODUCTION Global climatic change has become serious concern these days triggered by greenhouse gas emission due to production of energy from conventional sources of energy. Renewable energy especially solar photovoltaic is seen as important alternative source of energy for the future, and has recently attained importance due to lot of research and

978-1-4244-6078-6/10/$26.00 ©2010 IEEE

Different converter topologies like buck, boost, buck-boost and cuk converters are implemented for MPPT design. Boost converter can track MPP with maximum efficiency and can work for wide range of input voltage [2]. However, other advantages like less component needed for hardware

94

implementation and cost, makes this topology a better choice than the buck or buck-boost for MPPT system design in this paper.

dV

R s = − − dI

In this paper duty cycle of boost DC/DC converter is controlled by PWM signal from microcontroller implementing Incremental Conductance algorithm. Firstly, a model of MATLAB PV model under different temperature and insolation is simulated, tested and verified. Further, a boost converter and battery is modelled and tested under different conditions. Whole MPPT system is simulated using real environmental data and battery conditions. Result validates the good performance of the control technique and improved efficiency for hardware implementation and realization. Finally, a hardware model is designed and tested in lab at different operating conditions. Further, MPPT system has been tested on Solar Panel at different solar insolation level and temperature. The resulting system has high-efficiency, lower-cost, very fast tracking speed and can be easily modified for additional control function for future use.

2. MODELLING AND VALIDATION CELL IN MATLAB

OF

(6)

q (V + IRs ) I 0 e nkT − 1

Figure 1: Equivalent circuit used for MATLAB Simulation Table 1: Electrical characteristics data of Kyocera KC-40 S.No. 1 2 3 4

SOLAR

A model of moderate complexity is chosen for the MATLAB modelling and simulation. A simple PV cell is represented by a light dependent current source (Iph) in antiparallel with a diode driven by current Id (Shockley diode equation) [2] and a series resistance in the current path through the semiconductor material, the metal grid, contacts, and current collecting bus [5] & [7]. The presence of diode determine the output V-I characteristics of solar cell.

Electrical Specification Maximum Power Voltage Maximum Power Current Open Circuit Voltage Short-circuit Current

16.9V 2.34A 21.5V 2.48A

*Note: The electrical specifications are under test conditions of irradiance of 1kW/m2, spectrum of 1.5 air mass and cell temperature of 250C

Kyocera KC-40 PV module is chosen for a MATLAB simulation model. The module is made of 36 multicrystalline silicon solar cells in series and provides 40W of nominal maximum power. All the constants in above equations are determined by examining the datasheet provided by manufacturer. Table 1 shows its electrical specification that is necessary for cell/module modelling.

The parallel resistance associated with a small leakage of current through a resistive path in parallel with the intrinsic device [5], [6] & [7] is very large, so its effect is very less and is neglected. The equations which govern the characteristics of solar cell are: + IRs ) q (VnkT I = I sc − I 0 e − 1

nkT q

Above characteristics equations were solved with MATLAB using Newton Raphson method as it converge more rapidly to find the output current [2]. A pre-simulation result is shown in figure 2 for model validation.

(1) V-I Curve at STC (KYOCERA KC40) Pmpp = 39.35 W ( Error = -1.6 %)

2.48

I sc

G

=

G I sc G0 3

I0

T

(2)

G0

qE g 1

1

T n − nk T − Tref e = I 0 Tref Tref I sc T = I sc Tref 1 + a (T − Tref

[

Module Current (A)

2.3239

2

Impp = 2.32 A (Error = - 0.7%) 1.5

1

Vmpp = 16.93 V (Error = + 0.2%) 0.5

(3)

0

0

5

10

15

16.9339

21.5

Module Voltage (V)

)]

(4)

q (V + IRs ) V + IRs I = I sc − I 0 e nkT − 1 − Rp

(5)

Figure 2: I-V Characteristics of Solar PV from MATLAB Model Comparing the manufacturers’ datasheet and simulated result, it is found that the MATLAB model thus designed shows excellent correspondence. So, same model is used in this paper for further analysis.

95

The I-V curves and PV curves at different temperature and insolation are shown in figure 3 and 4.

is implemented because of its ease of digital implementation, fast response and efficiency. Incremental Conductance algorithm is based on differentiation of PV power to its voltage and on condition of zero slope of P – V curve at maximum power point (MPP). Power output from solar PV panel is: P = VI (7) Where P = module power, V = module voltage and I = module current; Differentiation with respect to V gives dP/dV = I+V(dI/dV) (8) The algorithm works on this basis. At peak power point, dP/dV = 0 (9) dI/dV = -I/V (10)

50 45

Module Current (A)

40

50 C

35

0C

30

1000 W/m2

25

25 C

75 C

20 15 10

500 W/m2 5 0

0

5

10

15

20

25

Module Voltage (V)

Figure 3: i-v curve at different environmental conditions 3

1000 W/sq.m

50

dP/dV = 0

2.5 45

Module Current (A)

0C

40

Module Power (W)

2

50 C

1.5

500 W/sq.m

25 C

1

75 C

35 30

dP/dV > 0

25

dP/dV < 0

20 15

0.5 10 5

0

0

5

10

15

20

25 0

Module Voltage (V)

10

15

20

If the operating point is on the right of the power curve we have dP/dVI/V

Locus of MPP at different Temp. (KYOCERA KC40) 45

40

G=1 T = 0 - 75 C

30

Power (W)

5

25

Figure 7: Operating point of Incremental Conductance algorithm on P-V curve.

The locus of Maximum Power Point (MPP) at different temperature and insolation is shown in figure 5 and 6. The effect of irradiance on power output is much higher than the effect of temperature. The maximum power point fall steeply as the irradiance level fall in case of irradiance change.

35

0

Module Voltage (V)

Figure 4: p-v curve at different environmental condition

25

20

then (11) (12) then (13) (14)

15

10

3. MPPT ALGORITHM AND BOOST CONVERTER

The present value and the previous value of the solar array voltage and current are used to calculate the values of dI and dV. If dV=0 and dI=0, then the atmospheric conditions have not changed and the MPPT is still operating at the MPP. If dV=0 and dI>0, then the amount of sunlight has increased, raising the MPP voltage. This requires the MPPT to increase the PV array operating voltage to track the MPP. Conversely, if dI - I/V, then dP/dV > 0, and the PV array operating point is to the left of the MPP on the P–V curve. Thus, the PV array voltage must be increased to reach the MPP. Similarly, if dI/dV < - I/V, then dP/dV < 0 and the PV array operating point lies to the right of the MPP on the P–V curve, meaning that the voltage must be reduced to reach the MPP.

MPPT optimize the power output of the solar cell by constantly tracking the varying maximum power operating point and adjusts the solar panel operating voltage. Different algorithms are used to track the maximum power point [3] & [4]. In this paper Incremental Conductance based algorithm

A small marginal error could be added to the maximum power condition such that the MPP is assumed to be found if [dI/dV + I/V] < . The value of was determined with consideration of the trade-off between the problem of not

5

0

0

5

10

15

20

25

Voltage (V)

Figure 5: Effect of temperature on locus of MPP Locus of MPP at different Irradiance (KYOCERA KC40) 45 40

T = 25 C G = 0.2 - 1.2

Power (W)

35 30 25 20 15 10 5 0

0

5

10

15

20

25

Voltage(V)

Figure 6: Effect of irradiance on locus of MPP

96

operating exactly at the MPP and the possibility of oscillating around it. It will also depend on the chosen perturbation step size C.

instantaneous maximum power from solar panel, power that can be transferred from the panel under direct coupling and maximum power that can be extracted from the panel using MPPT can be seen in Scope-3. Scope-2 plot battery current, solar current and load current. Negative battery current indicates battery is under discharge mode and positive indicates charging mode. If the load is high then battery and solar panel supply to load and if the load is light then solar panel supply current to load as well as it charge the battery simultaneously.

These changes in PV’s voltage may be done by coupling a DC/DC converter to PV and controlling properly its duty cycle (D) to track the maximum power point [8]. Most MPPT charge controllers are based on either the buck converter (step-down), boost convert (step-up) or buck-boost converter setup. In this paper boost converter is used because of its higher efficiency, reduced cost for realization and have large working voltage range [2].

4. SIMULATION MODEL

OF

State of Charge, battery voltage and battery current can be seen in Scope1. Different instantaneous values can be seen in display. For analysis of power under different condition is transferred to workspace. Current can be measured with current meter ‘CM2’. Similarly voltage with voltage meter ‘VM1’. Switching for various loads is performed by load switch controller.

COMPLETE MPPT MATLAB

The complete MPPT model is shown in figure 8. Three PV array with same characteristics is taken as current source for different measurement. First block ‘KYOCERA KC40-1’ is used to measure theoretical maximum power available from the panel under specified temperature and insolation level. Second block for panel ie. ‘KYOCERA KC40-2’ is used for analysis of power transfer for direct coupling to the load. Program for the third panel is written in the IC_algo’ block. This block is use for analysis of maximum power that can be extracted from the panel employing different algorithms. The battery is represented by a voltage source and the load as resistors.

5. SIMULATION RESULT AND DISCUSSION Simulation was run for 12 hour data (6am to 6pm) to see the output under real insolation and temperature. The graph shows the simulated result of PV panel under real insolation and temperature available from the solar lab of Kathmandu University. The increase in power due to MPPT is clearly seen from the above simulated result. The maximum power mismatch is about 7W. Theoretical maximum energy available = 357.45 Wh Energy extracted by MPPT = 346.66 Wh Direct battery coupled energy transferred = 296 Wh So, there is 50Wh loss of energy in a day without MPPT. These losses increase if the battery is in fully discharged condition and in cloudy day. The numerical result is shown in table 2.

For the present work Incremental Conductance algorithm is used to extract maximum power from the array. IMPP and VMPP are input to the boost converter. The change in duty cycle is necessary to control the output voltage. Manual switch 1 and 2 are used to run simulation under real data or any specific data. PV SW and Load SW are manual switch for solar panel and load side of battery. Plot of theoretical

Figure 8: Complete MATLAB/Simulink Block diagram

97

Theoretical MPP

Direct Battery Coupled Power MPPT

Figure 9: Simulated result Figure 10 shows the block diagram of the proposed MPPT system.

Table 2: Simulated result under real data and under STC. Energy

Real data (Wh)

At STC (Wh)

Theoretical Max. Energy (ETh)

357.45

800

MPPT Energy (EMPPT)

346.66

776.73

Energy transferred to Battery (EBatt)

296.00

615.14

Loss (EMPPT-EBatt)

50.66

161.59

6. HARDWARE DESIGN Figure 10: Maximum Power Point Tracking System block diagram

Simulation result shows the good performance of the control technique and improved efficiency for hardware implementation and realization. The important things to be considered for MPPT System design are the switch-mode topology and the control mechanism. The switch-mode topology is usually determined by the input and the output voltages desired. The tracker can be designed to either increase (known as a boost topology) or decrease voltage (known as a buck topology) from the array going into the battery or load. Because a tracker is essentially a specialized switching power supply, the input and output currents are such that the power into and out of the tracker are equal. In this research boost converter topology is chosen that will maximize the efficiency of the PV array. Boost converter are technically easier to build and also guarantee continuous current through the array.

6.1

Functional Specifications

High Frequency Boost Converter—A boost converter operating at high frequency (31 KHz) is designed which is controlled by the PWM signal from the microcontroller. Duty cycle is changed by microcontroller according to the temperature and insolation level. The microcontroller tries to maximize the watts output from the solar panel by controlling the step up ratio to keep the solar panel operating always at its Maximum Power Point.

The control section design involves the design of both analog and digital system. It will take in analog voltages and currents proportional to measured quantities, digitize them, process them in a microcontroller, and then convert a number back to a voltage proportional to what the system believes is the maximum power point voltage of the array and the state of charge of the battery.

Figure 11: Boost converter

98

Microcontroller—The DC/DC converter is controlled by the microcontroller. It calculates the solar watts generated by reading the voltage and current of the solar panels through the A/D port. It also calculate the battery side voltage and current in same way and send corresponding control signal to the converter and control the duty cycle of the converter to increase, decrease or turn off the converter accordingly. The ATMEGA16 is a perfect combination of features, performance, and low power consumption for this application.

modes and is given as: L = [(Ts Vo)/(2 Io,max)] D(1-D)2

The value of inductor should be greater than the calculated value for continuous conduction mode operation of converter [8]. MOSFET—The main criteria for selection of MOSFET is low switching power loss and able to handle the worst case current and voltage stresses i.e. it able to handle current and voltage of solar panel working in most favorable environmental condition (T = 0oC or low and G = 1kW/m2 or high). N-channel enhancement mode MOSFET IRFZ44N from Philips Semiconductor is chosen for best design. [8].

Voltage and Current Sensor—The current and voltage signals from the PV array are monitor by using the current and voltage sensing circuit. The sensing circuit that acts as transducer-converts the current reading to the voltage signals so that the microcontroller can understand and use the information to process and hence performs the PWMs. A simple voltage divider network to sense the voltage and a Hall Effect-Based Linear current sensor ACS714 from Allegro Microsystem Inc. is used.

MOSFET Driver—The PWM generated by microcontroller is not sufficient to drive the large capacitive loads such as MOSFET with high slew rate. The signals generated by the microcontroller can only deliver maximum current of 25mA. In order to achieve high speed switching in power MOSFET, a MOSFET driver chip TPS2812D was chosen.

Power Supply—Power necessary for microcontroller and other peripheral devices to work is taken from the solar panel itself using a suitable regulator. 6.2

Component Selection

7. EXPERIMENTAL RESULT

Since the objective of the tracker is to deliver maximum power from the solar module to the load, power losses associated with the tracker itself should be minimal. So, one of the very important factor to consider while selecting the component is the low power loss [8]. The other selection criteria associated with each major component are:

Experiment is carried out in a sunny day (29 May 2010) from 1100Hrs to 1300Hrs and data were taken for analysis. Data trackers log the necessary data for further analysis. Panel is fixed in adjustable frame to change the direction of panel for insolation control. The panel is kept initially at maximum insolation ie. horizontal position. Data was logged for 20 minutes in the same position. After every 20 minutes in first one hour of observation period, the insolation is decreased by tilting the panel away from the sun. Similarly, for next one hour, insolation is increased from minimum testing level to maximum level at 20 minutes step time and data is logged for further analysis.

Diode—Fast switching and low forward voltage drop Schottky diode (MBR745) with ability to block the required off–state voltage stress and have sufficient peak and average current handling capability. Capacitor—The primary criterion for selecting the output filter capacitor is its capacitance and equivalent series resistance (ESR). Since the capacitor’s ESR affects efficiency, low-ESR capacitors will be used for best performance. The output filter capacitors are chosen to meet an output voltage ripple specifications, as well as its ability to handle the required ripple current stress. An approximate expression [8] for the required capacitance as a function of ripple voltage requirement, Vo, D, switching frequency, Fs and output voltage, Vo is given as: C = VoD/(Fs Vo R)

(16)

The variation of efficiency of the MPPT system for the observation period is shown in figure 15. It is observed that the MPPT system is working at high efficiency during high irradiance level than those at low irradiance level. The decrease in system efficiency at low irradiance level is due to the constant losses in the converter section. The variation in efficiency for the total observation period is due to variation in duty cycle for converter MPPT operation. The performance of the algorithm and the system were found good for changing irradiance and hence temperature of the solar panel.

(15)

Inductor—Inductor with a ferrite core and low DC resistance is suitable for boost converter. Inductance is selected based on the maximum allowed ripple current at minimum duty cycle, D, at maximum output current Io,max. The critical inductance is defined as the inductance at the boundary edge between continuous and discontinuous

Solar Panel is tested with and without MPPT system at different environmental condition with a resistive load (bulb) of 2.2. It is observed from the figure 14 that the power extracted by peak power tracker is more than (nearly about two times) that without the MPPT system.

99

Pyranometer

Temperature Sensor

Kyocera KC-40 Figure 13: Experimental hardware setup

Figure 12: Solar Panel from Kyocera KC-40 used for experimental analysis

Figure 14: Comparison of module power with and without MPPT system

Figure 15: Efficiency of MPPT system

100

8. CONCLUSION

Sons, Ltd., Research and Applications, Prog. Photovolt: Res. Appl. 2003.

In this paper, MPPT is implemented by using a boost DC/DC converter, which is designed to operate under continuous conduction mode and a microcontroller using Incremental Conductance algorithm to control the PWM signals to the boost converter. An experimental result shows the system has good conversion efficiency of 92%. Even small improvement of efficiency could bring large saving, if the system is large e.g. Grid Connected System.

[4] D. P. Hohm, M. E. Ropp, “Comparative Study of Maximum Power Point Tracking Algorithms Using an Experimental, Programmable, Maximum Power Point Tracking Test Bed”, Photovoltaic Specialists Conference, Conference Record of the Twenty-Eighth IEEE Volume , Issue , 2000 Page(s):1699 – 1702. [5] Martin A. Green, Solar Cells: Operating Principles, Technology, and System Applications, Publisher: Prentice Hall (October, 1981) ISBN: 0138222703.

The designed intelligent MPPT system thus increase the power output from the solar PV drastically. The effect of variation of temperature and insolation in power output is observed and analyzed for MPPT design. Power transferred from the solar panel is affected by temperature and insolation variation which can be minimized by using MPPT system.

[6] Messenger, Roger & Jerry Ventre, Photovoltaic Systems Engineering 2nd Edition, CRC Press, 2005. [7] Ralph Gottschalg, Photovoltaic Technology for Bangladesh, Publisher: Department of Mechanical Engineering, BUET and CREST, Loughborough University, UK. March 2001.

This research presents a simple but cost effective control technique to harness solar energy optimally. It models each component of PV system and simulates in MATLAB to study the feasibility of the system under real weather conditions. Solar cell model using the equivalent circuit of moderate complexity was chosen for simulation and found to be good matching with the PV model. Simulation is carried out for different insolation and temperature for different interval of time. From the simulation result it is found that the use of MPPT system is more effective in the changing weather conditions and in cold places where we can extract extra power from the solar panel.

[8] N.Mohan, T.M Undeland, and W.P. Robbins, Power Electronics: Converters, Application and Design, John Wiley & Sons, Inc, New York, 2003. [9] Luis Castañer, Santiago Silvestre, Modelling photovoltaic systems using PSpice, John Wiley and Sons, January 2003. [10] Yen-Jung Mark Tung, Dr. Aiguo Patrick Hu, Dr. Nirmal-Kumar Nair, “Evaluation of Micro Controller Based Maximum Power Point Tracking Methods Using dSPACE Platform”, Australian University Power Engineering Conference 2006.

The hardware system is designed, tested and found to be satisfactorily working with reasonably higher efficiency at different operating conditions. The performance of system is tested in real sunny day and emulated to cloudy day for analysis. Incremental Conductance algorithms show better performance in terms of energy extraction under rapidly changing weather conditions. Efficiency of the system even under low irradiance, high temperature and fast changing operating conditions was found reasonably good.

[11] Eftichios Koutroulis, Kostas Kalaitzakis, “Development of a Microcontroller-Based, Photovoltaic Maximum Power Point Tracking Control System”, IEEE transactions on Power Electronics, Vol. 16, No .1, January 2001. [12] Francisco M. González-Longatt, “Model of Photovoltaic Module in Matlab™”, 2DO CONGRESO IBEROAMERICANO DE ESTUDIANTES DE INGENIERÍA ELÉCTRICA, ELECTRÓNICA YCOMPUTACIÓN (II CIBELEC 2005).

REFERENCES [1] Roberto faranda, Sonia Leva “Energy comparison of MPPT techniques for PV System”, WEAS Transaction on Power Systems, ISSN: 17905060, Issue 6, Volume 3, June 2008. [2] Geoff Walker “Evaluting MPPT converter topologies using a MATLAB PV model”, Journal of Electrical & Electronics Engineering, Australia; Volume 21, Issue 1; 2001; 49-55. [3] D. P. Hohm and M. E. Ropp, “Comparative Study of Maximum Power Point Tracking Algorithms”, Progress in Photovoltaics: Copyright-2002 John Wiley &

101