Design of MPPT Based Microcontroller Using Constant Voltage and Perturb & Observ Algorithm, With Buck Converter Trough Fuzzy Control

Transcription

1 Proceedings of the 3rd Applied Science for Technology Innovation, ASTECHNOVA 2014 International Energy Conference Yogyakarta, Indonesia, August 2014 Design of MPPT Based Microcontroller Using Constant Voltage and Perturb & Observ Algorithm, With Buck Converter Trough Fuzzy Control Prisma Megantoro, Rachmawan Budiarto Physics Engineering Department, Gadjah Mada University Jln. Grafika 2 Yogyakarta Indonesia prisma.megantoro@giz.de; rachmawan@yahoo.com Irawan Eko Prabowo Center For Energy Studies, Sekip Blok K1-A, Bulaksumur, Yorgyakarta INDONESIA prabowo_14@yahoo.com ABSTRACT Maximum Power Point Tracking (MPPT) is supporting device for power harvesting on Solar Power Generator. This system means to optimalize power transfer from solar panel array to battery. Power transfering process shown in solar panel characteristic as I-V curve. This curve depends on weather condition, such as temperature and irradiation level of the sunlight. The purpose of this research is to design the MPPT system device based on Atmega16 microcontroller. This system implements combination of Constant Voltage and Perturb & Observ algorithms. Then fuzzy logic system used to control the output voltage The conclusions of this research are; first, devising switching PWM using prescaler of 1024, with clock of microncontroller on its default 1 MHz, and 50% duty cycle generated the highest efficiency of the system, that is 58.03%, on the other hand, it has the lowest inductor impedance value, that is 5.03 ohm. Second, MPPT device which has been devised has the average efficiency of 88.89% and provides up 11,36 % to increase the power generated KEYWORDS: Maximum Power Point Tracking, fuzzy, microcontroller 1 INTRODUCTION Each Solar Panel has charateristic, that is I-V Curve which depends on temperature and level irradiation of the sunlight. Inthis curve, there is a valueoccured when the maximum power transfer happens, the coordinate is (Vmpp, Impp). This point called the Maximum Power Point. MPPT algorithm used to track the Maximum Power Point, then convert it into operational voltage which is Vmpp voltage with buck converter (DC/DC Converter) circuit use switching mode PWM. This research began with devising buck converter circuit, then followed by devising the reading voltage and current and its software, then implement the tracking algorithm and fuzzy logic control method on C language use AVRStudio4 IDE, and the last is total devising of MPPT hardware and its trial. The trial has been done at the frontyard of Center for Energy Studies Office of UGM. Boundary problem in this research is the MPPT system designed for solar system with operating voltage of 12/24 V and maximum current of 20 A using ATmega16 microcontroller as processor. 210

2 2 LITERATURE Research conducted by Beng Tito entitled New MPPT method for photovoltaic cells based on PI controller, discusses a new method for MPPT solar cells based on PI controller. Analysis of simulation work done by MPPT design based on boost converter. In the study that produced the tracking of bothalgorithm, from the simulation obtained that the ICM algorithm generated watts rated power. while the PI algorithm generated watts of power. This study concluded that the PI algorithm has advantages in terms of speed and tracking process than the ICM algorithm. Research conducted by Dianggoro entitled Design Maximum Power Point Tracker (MPPT) Solar Panels ForCuk Converter Using Hill Climbing Methods, discusses the design of the MPPT system using Continuous Conduction Mode (CCM). Research that conducted by Dianggoro (2011) use simulation to determine the reliability of the system. The algorithm used in this MPPT is Hill Climb Search (HCS). The result is the system which has implemented MPPT could increase the output power ranging from 4.70 to 36.49% compared to the system without MPPT. This test done with some variations of load 5 ohms, 10 ohms, and 15 ohms as well as varying levels of irradiation. 3 BASIC THEORY Maximum Power Point Tracking is a technique to track the voltage which could generate the maximum power from solar array. Energy from solar panels has the characteristics described in the V-I curve. Therefore, this technique is applied to find the maximum power point on the V-I curve. This MPPT device integrated in a electronic device for voltage conversion. MPPT here is a DC / DC converter with a controller. Existing work processes within the MPPT controller can be classified into several algorithms such as Constant Voltage and Perturb &Observ. In this algorithm, the controller regulates the voltage variation to be included to the data array. Along with the process of setting this voltage, voltage variations that occur every measured current. Current were measured then entered to the data array. Each variation of voltage multiplied by the current measured,then the power value is obtained. This voltage setting process leads to changes the value of power. The process will continue to run forward or increase the voltage settings if the power continue to rise. And the process will stop if the obtained value of power began to fall. Then Constant Voltage, In this algorithm refers to the open circuit voltage (Voc) were measured when the DC connection is opened. The voltage at the maximum power point can be calculated by the equation below. Vref = kxvoc (3.1) Then for voltage conversion, this system using fuzzy logic control to determine the value of the PWM. This PWM value used to obtain the duty cycle of the DC / DC converter. Fuzzy logic is a method that uses the membership function of a zero-one membership for each set (Ginanjar, 2011). Membership value in the fuzzy set is a closed interval [0,1]. Fuzzy set has two attributes, namely linguistically and numerical. Linguistics, namely the naming of a group represents a particular condition in a language that is often used by humans. While the numericalis a value that indicates the size of a variable. There are some parts of the fuzzy system, ie; variables, the set, the universe of discourse, and domain. In the set there are multiple representations of fuzzy membership functions, one of which is a linear representation. In the linear representation, mapping input to the degree of membership is described as a straight line. After that, is Inference, thatis the process of combining many rules based on available data. Fuzzy inference system receives input crisp. This input sent to a knowledge base that contains n fuzzy rules in the form of IF-THEN. antecedent membership value (y) will be sought on each rule. To obtain a crisp output value / value of Z, sought a way to change the input (in the form of fuzzy sets obtained from composition of fuzzy rules) into a fuzzy set of numbers in the domain. This method is called 211

3 defuzzification method. Defuzzificationin Tsukamoto method mean centered (Center Average Defuzzyfier) formulated in equation. z = n i=0 yixi n i=0 yi The voltage conversion systems using a buck converter to reduce the voltage of the solar panel voltage to Vmpp. Z is the value used to determine the value of the PWM duty cycle of the buck converter. (3.2) 4 IMPLEMENTATION OF RESEARCH The design is done in several stages. Diagram of the study design of the MPPT device shown in Figure 4.1. Figure 4.1: Block Diagram MPPT Controller Device The first phase is the design of current and voltage measurements. Used IC ACS A for current sensor and a voltage divider circuit for the voltage sensor. The second stage is the design of DC / DC Converter buck converter topology. Figure 4.2: Schematic Design of Buck Converter In figure above shown the 4N35 optocoupler is used as the first switch the voltage signal 0/5V from OCR1A pin. This IC is used as a switch activation voltage (Vgs) MOSFET. Then the second switching use MOSFET IRF 540 which is trigger the switch of the MPPT device. The third stage is the design and its interface microcontroller ATmega16 LCD, LED, SPI, and others. After that is the implementation of 212

4 MPPT tracking algorithm with a combination of Constant Voltage and Pertub&Observ. Tracking process is shown in Figure 4.3. Figure 4.3: Process Tracking MPPT Figure 4.14 above is the process of tracking algorithm. Vref1 point is a point for the calculation algorithm Vmppof the Constant Voltage. (r) is the tracking range for Pertub&Observ algorithm that starts from Vref1-1V. MPPT1 and MPPT2 result point that indicates MPPT tracking process is done. In addition to tracking the implementation of the second algorithm, fuzzy control is also used as a control output voltage. This fuzzy system has 2 variables, namely dv and pwm. Each variable has a set that combined the 3 inference rules; [R1], [R2] and [R3]. Each of them obtained for value of x and y, then the average defuzzyfication done to determine the value of pwm. Pwm value determines the value of the duty cycle of the switching process MPPT. 5 RESULT 5.1 Relationship between switching frequency, impedance of the inductor, and the efficiency of the system In this test, the duty cycle is fixed at 50% with input voltage of 12 volts from the adapter and load remain a motorcycle headlamp 10 ohms. Using a clock frequency of 1 MHz. Input voltage, input current, output voltage and output current are measured by varying the value of the prescaler register TCCRA. Prescaler value affects the switching frequency in Equation 5.1. Switching frequency = frequency clock prescaler 256 (5.1) 213

5 Table 5.1: Relationship between frequency switching, inductor impedance, and efficiency of the system. prescaler Fs (Hz) eff (%) Xl ,25 49,78 11, ,28 51,37 6, ,04 47,48 5, ,26 55,00 5, ,81 58,03 5,03 In Table 5.1 indicated that the 50% duty cycle at the lowest frequencies, ie 3.81 Hz in the prescaler is set at 1024 system produces the highest efficiency and lowest inductor impedance, which is 5.03 ohms. 5.2 Testing MPPT system The trial of the MPPT system done by measuring the voltage and current input and output voltage and current solar system. The trial used 195 WpSharp branded solar panel and 100Ah Sigellum battery.the trial was conducted over 2 days, on Saturday dated August 31, 2013 and Sunday, September 1, The test has been performed at the frontyard ofcentre for Energy Studies UGM Office which is located in Block K1 Sekip A, UGM, Yogyakarta Sleman. The office position is at latitude 7 46'17.80 "S and longitude '30.82" T. The trial was conducted over 2 days with each day divided into 3 periods of testing, ie morning, afternoon, and evening. The trial data is input voltage and current, output voltage and current, and PWM values.that five parametersare displayed in the LCD display. Data is collected every 1 minute. In this test,mppt device installed on the mini-off grid solar system. Table 5.2: Results of the trial on Saturday periods Pin ave. Pout ave. eff ave. (%) PWM pagi 36,32 32,73 90, siang 35,41 31,05 88, sore 7,35 6,59 89, In Table 5.1 is shown that the power input and power output on smaller than the period during the morning period. That's because the weather during the period of changed significantly at the beginning and end of the test. Figure 5.1: Graphic comparison of the power input and power output in the whole testing period on Saturday output in the whole testing period on Saturday 214

6 However, peak power (Ppeak) input and output, remains highest during the period compared to the morning and afternoon. It shows the characteristics of daily solar irradiation levels. Table 5.3: Test results for each period on Sunday. periods Pin ave. Pout ave. eff ave. (%) PWM pagi 33,77 30,00 88, siang 40,85 35,90 88, sore 30,57 26,84 88, In Table 5.3, it was shown that the power input and power output during the period of the highest among morning and afternoon periods. Figure 5.2: Graphic comparison of the power input and power output in the whole testing period on Sunday the whole testing period on Sunday Peak power (Ppeak) input and output, also the highest during the period compared to the morning and afternoon. It shows the characteristics of daily solar irradiation levels. Overall system efficiency MPPT system showed good performance with a value of not less than 88%. Then to determine the performance of solar power system, this MPPT device tested in it in two identical test conditions. This test done by compared the generated solar power system with and without MPPT. Table 5.4: Comparison of MPPT solar systems without and with MPPT Parameters System without System with MPPT MPPT Ave. Voltage 12,39 V 12,91 V Ave. Current 0,36 A 0,38 A Ave. Power 4,40 Watt 4,90 Watt 6 CONCLUSION This research result some conclusions, such as: 1. Frequency of 3.81 Hz prescaler is set on 1024, microcontroller clock default 1 MHz, and duty cycle of 50% resultedthe highest system efficiency, ie 58.03% and yield the smallest inductor impedance value, which is 5.03 ohms. 215

7 2. On duty cycle values varied from 0-250, set the prescaler is varied, 64 prescaler generated the best graph efficiency with the value of efficiency is 70.07% on average. 3. MPPT device tested in the field used a battery load, the tracking algorithm implements a combination of Constant Voltage and Perturb and Observ, and voltage control method using fuzzy logic Tsukamoto, has system efficiency in average of 88.89%. While the trial done in the laboratory, solar mini grid system could increase % reated power more than the solar system without MPPT. Recomendations for this research are: Using a microcontroller or varying the clock frequency over its default frequency for better system efficiency. And also use other tracking algorithms, such as, Hill Climb Search, Incremental conductance, PI, etc. REFERENCES Abdurrahman, G. (2011). Implementation Method of Tsukamoto (Fuzzy Logic) Indecision Support System for Determining the Total Production of Data Availability and Based on Total Request. Yogyakarta State of University, Yogyakarta, Indonesia. Darmawan, D. (2011). Design of Maximum Power Point Tracker (MPPT) for Solar Panels Using Cuk Converter Hill Climbing Method. Essay. Institute of Technology Sepuluh Nopember, Surabaya, Indonesia. Harmini, (2010). Implementation of MPPT (Maximum Power Point Tracker) DC-DC Converter on Photovoltaic Sistem using Constant Voltage, Perturb and Observe (P&O) and Incremental Conductance. Master thesis. Unversity of Gadjah Mada, Yogyakarta, Indonesia. Tito, B. (2012), The New MPPT Method for Photovoltaic Cells Based on PI Controller. Essay. University of Indonesia, Depok, Indonesia. 216