68 CHAPTER 4 DESIGN OF CUK CONVERTER-BASED MPPT SYSTEM WITH VARIOUS CONTROL METHODS 4.1 INTRODUCTION The main objective of this research work is to implement and compare four control methods, i.e., PWM with periodic carrier, Zero voltage switching, Zero Current switching, chaotic PWM with chaotic carrier in terms of their performance in suppressing ripples, reducing peaky electromagnetic inference and increasing converter conversion efficiency in MPPT circuits of the solar PV powered Cuk converter system. This research work proposed to design chaotic PWM-based MPP tracking using Cuk converter in order to improve electromagnetic compatibility, converter conversion efficiency and this method is compared with soft switching Cuk converter based MPP tracking. Due to continuous power spectrum feature in chaotic PWM, the power density peak in output voltage and hence the electromagnetic inference is reduced to great extent. The proposed MPP tracking is achieved by connecting a chaotic PWM based Cuk converter between a solar panel and a load (Rheostat). 4.2 CUK CONVERTER WITH PERIODIC CARRIER Cuk converter shown in Figure 4.1 provides an output voltage which is less than or greater than the input voltage. It works based on the
69 capacitor energy transfer. It has low switching losses and highest efficiency among all non-isolated DC-DC converters. It exhibits non-pulsating input current characteristic due to the inductor in the input stage. Also Cuk converter is capable of sweeping the V-I curve of solar PV module in CCM from open circuit voltage to short circuit current condition and hence Cuk converter is a suitable converter to be employed in designing the MPPT circuits. Cuk converter is used as the power stage interface between PV module and the load. Cuk converter has two modes of operation. First mode of operation is when the switch (MOSFET) is closed (ON) and it is conducting as a short circuit. In this mode, the current through inductor L1 rises. At the same time, the voltage of capacitor C 1 reverse biases diode D and turns it off. The capacitor (C 1 ) releases its stored energy to the load. Figures 4.1 Cuk converter as the solar PV power stage interface Figure 4.2 shows the generation scheme of PWM pulse using periodic carrier to trigger main switch (MOSFET) of Cuk converter. The periodic carrier and PWM pulse are shown in Figures 4.3.and 4.4. Figure 4.2 MATLAB model to generate PWM pulse
70 Figure 4.3 PWM pulse using periodic carrier Figure 4.4 Periodic carrier Figures 3.6 and 3.7 show the relation between tracked power from the solar PV module and duty cycle for the change in irradiations using periodic carrier. The percentage duty cycle of the main switch of the Cuk converter is 35.3. The converter conversion efficiency of the Cuk converter is 86.26%. 4.3 ZVS-PWM CUK CONVERTER The converter that connects to the solar panel and load is a ZVS- Cuk shown in Figure 4.5. By connecting ZVS-PWM Cuk converter between
71 solar PV module and load, the APAO MPPT algorithm has been implemented. The ZVS is facilitated in order to reduce the EMI during switching transitions, high efficiency with high voltage inputs, no power loss due to the discharge of output capacitance of MOSFET, zero power lossless switching transition. To track maximum power from PV module, the duty cycle of the main switch is adjusted by using ATMEGA16 micro-controller such that the input resistance of the ZVS-PWM Cuk converter is equal to the output resistance of the solar PV module. The output power of solar module is equal to the input power of the converter which ensures maximum power transfer. To compare the adaptive MPPT algorithm with a traditional PAO method, the same converter is being used. The switching frequency of the converter is 25 khz. The Active-Clamp ZVS-PWM Cuk converter is shown in Figure 4.5 featuring PWM, and soft switching (ZVS) in all three active switches, resulting in high efficiency at high-frequency operation without significant increase in voltage and current stresses on switches. It consists of an input inductor L i, power switch S 1, S 2, S 3, energy transfer capacitor C a, output inductor L o, filter capacitor C o and resonant capacitors C r and resonant inductor L r. The ZVS-Cuk converter will not have any switching losses across all three power devices. The function of this converter is to transfer electrical energy from the input voltage supply V s to the output load R o at a voltage level that can be higher or lower than the input supply through the energy transfer capacitor. As in any power application, high efficiency is essential and, hence, the increasing of frequency can be problematic because of the direct dependence of switching losses on frequency.
72 Figure 4.5 Active Clamp ZVS-Cuk converter The input filter inductance is assumed to be a current source in the proposed circuit. The capacitors Cc and C a are chosen to have large capacitances so that the voltages V c and V ca are considered as constants. The six topological stages and key waveforms of the proposed modified Active Clamp ZVS-PWM Cuk converter, to one switching cycle, are shown in Figure 4.6 (a) to (f), respectively. The two switches S 1 and S 3 are triggered in a complementary way and soft switching is employed for all three switches. The main switch S 1 is switched off at t = t o, when the switching cycle starts. Prior to t= t o, the main switch S 1 is turned on, the auxiliary switch S 2 is turned off, and S 3 is also off. When S 1 is turned off, at t= t o, the first stage has started, as shown in Figure 4.6(a). The capacitor is charged under constant current. When V cr (t) reaches V 0 +V S, the diode of switch S 3 starts conducting, at this instant, S 3 also triggers. Thus switching losses across the diode is eliminated. At the second stage, current through L r and voltage across C r rings in a resonant way. Voltage V cr (t) increases until it reaches (V s +V 0 +V c ) when V cr (t) = V s +V o +V c, the antiparallel diode of S 2 starts conducting.
73 At the third stage, L r current ramps down, because it is considered as a constant voltage source, until it reaches zero, when it changes its direction and rises again and voltage across is clamped at V s +V o +V c. When the antiparallel diode of S 2 is conducting, the auxiliary switch S 2 is switched on to achieve a lossless turn-on. This stage ends when S 2 is turned off at t = t 4. At the fourth stage, the voltage across C r falls due to the resonance between L r and C r, until it reaches zero at t = t 4. This stage ends when V cr (t) becomes null and the antiparallel diode of S 1 and S 3 begins conducting. Hence lossless turn-on is achieved for the switch S 3. (a) (b) (C) (d) (e) (f ) Figure 4.6 Operational modes of the Active clamp ZVS-Cuk converter
74 At the fifth stage, S 1 is turned on without switching losses, in a ZVS way, because V cr (t) becomes zero. The current through L r changes its polarity and ramps up to reach Is at t = t 5. Then, the diode of switch S 3 becomes reversibly biased and turns off, and, at the sixth stage, S 1 conducts a current equal to I s +I 0 and the auxiliary switch S 2 is off. The S 3 is off and the stage ends when S 3 is turned off at the end of the period. The theoretical waveform during the one switching cycle is shown in Figure 4.7. Figure 4.7 Theoretical waveform of modified Active clamp ZVS- PWM Cuk converter Due to the capacitance C r, S 1, S 2 and S 3 are turned off with no losses, in an ZVS way. However, S 1, S 2 and S 3 will turn on with no losses, only if there is enough energy stored in L r to achieve soft commutation. At, t = t 1, it is necessary to charge C r from V 0 +V s to V 0 +V s +V c. At t = t 3, it is necessary to discharge C r from V 0 +V s +V c to zero. The latter is very tedious because it needs more energy. Thus, if enough energy is guaranteed to achieve soft commutation for S 1, then S 2 and S 3 will also achieve soft commutation. From the energy relationship in L r and C r, at t = t 3, we have
75 ½ L r.(i s +I o ) 2 ½ C r.(v c +V o +V s ) 2 (4.1) T= f s / f o and 0 = 1 / (L r C r ) ½ (4.2) f s -switching frequenc, f o -resonant frequency. The voltage gain is given by, = (4.3) where d is duty cycle. L - normalized load current which is given by L = (4.4) The normalized clamping voltage is given by = ( ) * ( ) (4.5) The output voltages for various values of duty cycle are shown in Figure 4.8. The input voltage is 16.4 V, swithing frequency (f s ) =25kHz, Figure 4.8 Duty cycle versus output voltage
76 The input resistance R i of the Active clamp ZVS-PWM Cuk converter is given by = ( ) (4.6) where L =L i // L o, f -switching frequency, D- is the duty cycle of the main switch S 1. Under discontinuous inductor current mode of operation, the voltage across C a is given by V ca = V s + V o. Where the V s is the converter input voltage and V o is the converter output voltage. The maximum voltage stress on the main switch S 1,V S1stress, occurs in the time interval from t o to t 4 when S 1 is off and S 3 is on. Maximum voltage stress on S 3, s 3stress occurs in the interval t 4 to t 6 when S 1 is on and S 3 is off. V S1stres = V s3stress =V ca = V s +V o (4.7) Table 4.1. The specifications of ZVS-PWM Cuk converter are given in Table 4.1 Specification of ZVS-PWM Cuk converter Maximum power 37watts Switching frequency 25kHz Converter Output voltage 8V Input voltage 16.4V Main inductor L1 500e-6H Resonant inductor Lr 5e -6 H Resonant capacitor Cr 50e -9 F Capacitor C a 220e -6 F Capacitor C c 20e -6 F Capacitor C o 220e -6 F ESR (Element Series 0.5 Resistance) Load resistance R 2 Step size 1e-7 MOSFET IRF510
77 4.4 ZCS-CUK CONVERTER To achieve ZCS, the resonant inductor L r, is in a series with a switch as shown in Figure 4.9. If zero current resonant switching operates in half-wave mode, the switch current can flow unidirectionally and is permitted to resonate only in the positive half cycle. To avoid overvoltage across the switch S, and parasitic oscillation, a unidirectional current switch is implemented. IGBT S 1 and series-wound diode D form the principal unidirectional current switch, L r, and C r, form the resonant circuit, C 2 is the output filter capacitor and R, represents the load. The equivalent circuit of ZCS-Cuk converter is shown in Figure 4.10 Five modes of operation in the ZCS-Cuk converter have been identified during one switching cycle. Figure 4.9 ZCS-Cuk converter Figure 4.10 Equivalent circuit of ZCS-Cuk converter
78 MODE 1: Linear stage (t o - t 1 ) Mode1 starts a switching cycle at time t 0, IGBT ( S 1 ), is turned on at zero current condition with a positive gating signal due to the resonant inductor L r, which limits the di/dt of the IGBT S 1 current. Output diode D remains on. The IGBT current, i s1, rises linearly and the current through L r, will decrease linearly at the same rate. The voltage across C r is equal to input voltage of the converter. This mode ends at time t 1, when the switch current, i s1,, is equal to i s. The equivalent circuit for mode 1 is shown in Figure 4.11. Figure 4.11 Mode 1: Linear stage MODE 2: Resonant stage (t 1 - t 2 ) At t= t 1, the output diode D is off and resonant mode starts. The equivalent circuit of mode 2 is shown in Figure 4.12. The IGBT current i s1 increases in a sinusoidal fashion and the voltage across C 1 decreases in a linear fashion at the same time. The resonant capacitor C r, is discharged and its voltage becomes negative. The resonant inductor current i r starts to decrease in a resonant mode, becomes negative, and increases again. At time t 2, the resonant capacitor voltage reaches the minimum negative peak voltage and the switch current is equal to (i s + i o ).
79 Figure 4.12 Mode 2: Resonant stage MODE 3: Resonant Stage (t 2 - t 3 ) After time t 2, the current i s1, is less than the input current i s, and it decreases until zero. The resonant capacitor voltage is negative, and increases again. This mode ends at time t 3, the switch current i s, is equal to zero. The resonant inductor current i Lr reaches the maximum positive peak current. At this moment, the voltage of capacitor C 1 will reach a minimum positive voltage, and increase linearly again. The equivalent circuit of mode 3 is shown in Figure 4.13. Figure 4.13 Mode 3: Resonant stage
80 MODE 4: Linear stage (t 3 - t 4 ) At time t 3, IGBT S is turned off and the resonant capacitor C r and the capacitor C 1, respectively are charged linearly by the input current i s. The resonant inductor current i Lr, is reduced to zero. This mode ends when the resonant capacitor voltage V cr is more than the capacitor voltage V c1. At time t 4, D, will turn on. The equivalent circuit for mode 4 is shown in Figure 4.14. Figure 4.14 Mode 4: Linear stage MODE 5: Power Transfer Stage (t 4 - t 5 ) At time t 4, the resonant capacitor voltage is clamped by D. This mode ends when the resonant capacitor voltage V cr,, is equal to Vs. The equivalent circuit for mode 5 is shown in Figure 4.15. Figure 4.15 Mode 5: Power transfer stage
81 The voltage and current across the IGBT is shown in Figure 4.16. Figure 4.16 Voltage and current waveforms across IGBT under ZCS Table 4.2 shows the designed values for ZCS-Cuk converter-based MPP tracking. The power rating is 37 W. The duty cycle is around 20 %. The operating frequency is 25 khz. Table 4.2 Specifications of ZCS-Cuk converter Maximum power 37watts Switching frequency 25kHz Input voltage 16.4V Main inductor L 1 500e -6 H Resonant inductor L r 5e -6 H Resonant capacitor C r 2.2e -6 F Capacitor C a 220e -6 F ESR(Element Series Resistance) 0.5 Load resistance 2 Step size 1e-7 IGBT G15N60 Diode BY129
82 4.5 CHAOTIC PWM-CUK CONVERTER The output waveform of the Cuk converter-based MPPT solar PV system, controlled by the traditional PWM, consists of many harmonic components. The distribution of harmonics is influenced by the periodic carrier. The carrier frequency and carrier amplitude are invariant under traditional PWM. Thus, the spectrum has biggish peaks close to the carrier frequency or its multiples. This makes the Cuk converter difficult to satisfy the international standards of Electro Magnetic compatibility (EMC). Conventionally, filters are used to reduce EMI of Cuk converter-based MPPT system. Moreover, each filter can only restrain EMI in a certain frequency band. The existence of a number of biggish peaks of the spectrum with traditional PWM makes it difficult to design filters for the Cuk converter. It is desirable for DC-DC Cuk converter used in MPPT system to eliminate EMI without using filters. The distribution of harmonics is influenced by the carrier and the chaotic behavior of DC-DC Cuk converter can be used to reduce EMI. So, chaotic frequency or chaotic amplitude can be used to distribute the harmonics continuously and evenly over a wide frequency range. Although the total energy is not changed, the peaks of the harmonics are reduced, thereby restraining the EMI. Chaotic phenomena are useful in suppressing electromagnetic interference by adjusting the parameters of the Cuk converter and making it operate in chaos, a chaos-based pulse width modulation is proposed to distribute the harmonics of the DC-DC converters continuously and evenly over a wide frequency range, thereby reducing the EMI. The output waves and spectral properties of the EMI are simulated and analyzed. In order to improve the electromagnetic compatibility of solar PV- powered system, direct control chaotic pulse width modulated Cuk converter
83 as shown in Figure 4.17 is proposed to track maximum power from the solar PV module. Therefore, in order to get chaotic frequency f or chaotic amplitude A, chaotic PWM, as shown in Figure 4.17, is proposed and analyzed. The analogue chaotic PWM has its advantages over the digital one in its low cost and easy design, making it suitable for high-frequency operation. Figure 4.17 Chaotic PWM-Cuk converter The analogue chaotic PWM generation circuit consists of 555 timer circuit (triangular or saw toothed waveform circuit) in combination with Chua s diode in order to generate chaotic pulse width modulation which is used to trigger the main switch of Cuk converter, and used for reducing EMI in tracked converter voltage. The CPWM adopts sawtooth to modulate, but its carrier period T changes according to the equation 4.8. T = ( ) * T (4.8) where T is invariant period, x, i= 1,2,.N, a chaotic sequence,
84 x=(x 1,x 2 x N ), and Mean(x),average of the sequence, defined as Mean(x) = Lim Xi N (4.9) Similarly, the CPWM also adopts sawtooth to modulate, but its carrier amplitude A changes according to A = {1+ k ( ) } A (4.10) where A is the invariant amplitude, X, i= 1,2,.N, a chaotic sequence, x = (x 1,x 2 x N ), and Mean(x), average of the sequence, and K is the modulation factor of the amplitude. The value of K is selected as low so that the ripple in the output voltage of the Cuk converter is low. Also the ripple in the output voltage controlled by chaotic PWM is low when compared with soft-switching Cuk converter-based MPPT system. 4.6 SIMULATION RESULTS The closed loop diagram was simulated in MATLAB /Simulink which is given in Figure 4.22 that includes a PV module electric circuit subsystem (MATLAB model), a DC-DC converter and an adaptive PAO algorithm. Four different control methods, i.e., traditional PWM with periodic carrier, ZVS-soft switching, ZCS-soft switching and PWM with chaotic carrier are simulated and compared in terms of their performance in suppressing ripples, reducing EMI and increasing converter conversion efficiency. The soft-switching ZVS-PWM active clamp Cuk converter has been simulated with the solar PV module rating of 37Wp in MATLAB/ Simulink as shown in Figure 4.22. The maximum power tracking efficiency at
85 the input of the converter from the solar PV module is 98.9%. The ZVS-PWM converter conversion (output power to the input power) efficiency is 91.3%. PV module is modeled based on the electrical Equations (2.1) and (2.2) to provide voltage and current to the Cuk converter and the microcontroller simultaneously. Using the adaptive PAO algorithm, the duty cycle is adjusted. High perturbation is selected when the operating point is far away from MPP and low perturbation is selected when the operating point is closer to MPP. When the obtained tracked power is equal or nearby actual maximum, the variation in the duty cycle is minimum in such a way that the memory increment value is selected. Using APAO algorithm, the output is obtained in terms of pulses as shown in Figure 4.18. The method of generation of 3 pulses which can be given to the switches of a ZVS-Cuk converter. Figure 4.18 PWM-Pulse generation scheme for three switches 4.6.1 Generation of Chaotic PWM in MATLAB The chaotic PWM is generated in MATLAB using the following circuit model shown in Figure 4.19. The generated chaotic carrier and chaotic PWM are shown in Figures 4.20 and 4.21.
86 Figure 4.19 MATLAB model to generate Chaotic PWM Figure 4.20 Generation of chaotic PWM in MATLAB Figure 4.21 Generation of chaotic carrier in MATLAB
87 The adaptive PAO MPPT algorithm is coded in embedded file and the maximum power is tracked using ZVS-Cuk converter which is shown in Figure 4.22. Figure 4.22 ZVS-PWM Cuk converter-based MPP tracking The changing irradiation is modeled to study the system operation. The temperature is constant at 25 C and the illumination level is varying between two levels. Initial irradiation is set as 1000 W/m 2. After 0.01sec, the irradiation (G) is suddenly changed to 500 W/m 2. The relationship between the duty cycle and solar PV power are shown in Figures 4.23 and 4.24. They show that the output power at G =1000 W/m 2 and 500 W/m 2 is 36.74 W and 17 W, respectively, for ZVS- Cuk converter-based MPP tracking. The percentage duty cycle of the main switch S 1 is 43%. The ZVS-Cuk converter conversion efficiency is 91.26%. The voltage across and current through the main switch of ZVS-Cuk converter is shown in Figure 4.25.
88 Figure 4.23 Change in duty cycle for various irradiation levels for ZVS Cuk converter-based tracking Figure 4.24 Change in power for various irradiation levels for ZVS-Cuk converter-based tracking
89 Figure 4.25 Voltage and current waveform across main switch S of ZVS- Cuk converter Similarly, the ZCS-Cuk converter is used to track maximum power from the solar PV module which is shown in Figure 4.26. Figure 4.26 ZCS-Cuk converter-based MPP tracking
90 Initial irradiation is set as 1000 W/m 2. After 0.02sec, the irradiation (G) is suddenly changed to 500 W/m 2. The relationship between the duty cycle and solar PV power is shown in Figures 4.27 and 4.28. They show that the output power at G=1000 W/m 2 and 500 W/m 2 is 36.74 W and 17 W, respectively, for ZCS-Cuk converter-based MPP tracking. The percentage duty cycle of the main switch is 18.5%. The ZCS-Cuk converter conversion efficiency is 91.12%. The voltage across and current through the main switch is ZCS-Cuk converter as shown in Figure 4.29. Figure 4.27 Change in duty cycle for various irradiation levels in ZCS- Cuk coneverter- based tracking Figure 4.28 Change in power for various irradiation levels in ZCS-Cuk converter-based tracking
91 Figure 4.29 Voltage and current waveform across main switch S of ZCS- Cuk converter Chaotic PWM-Cuk converter shown in Figure 4.30 is used to track maximum power from the solar PV module. The chaotic PWM shown in Figure 4.19 is used to trigger the main switch of the Cuk converter. The initial irradiation is set as 1000 W/m 2. After tracking of maximum power at 0.1sec, the irradiation (G) is suddenly changed to 500W/m 2. The percentage duty cycle of the main switch S is 26%. The tracked solar PV power using chaotic PWM-Cuk converter is shown in Figure 4.31. The voltage across and current through the main switch of CPWM-Cuk converter is shown in Figure 4.32.
92 Figure 4.30 Chaotic PWM Cuk converter based MPP tracking Figure 4.31 Change in power for various irradiation levels in CPWM- Cuk converter-based tracking
93 Figure 4.32 Voltage across and current through main switch of CPWM Cuk converter The maximum power tracking efficiency is 99.3% without considering the efficiency of solar PV module and converter. The converter conversion efficiency is improved to 93.1% when chaotic PWM Cuk converter is used for MPPT purposes. 4.7 CONCLUSION Cuk converter-based tracking with conventional PWM, and zero voltage switching for all the three switches, zero current switching and chaotic PWM were proposed to overcome the limitations of the conventional Cuk converter-based MPPT. The zero voltage switching reduces the EMI during switching transitions. The converter conversion efficiency is improved to 93.1% when chaotic PWM Cuk converter is used to track maximum power from solar PV module.