- Email: [email protected]
Integrated Solar Controller for Solar Powered off-Grid Lighting System
Available online at www.sciencedirect.com
Energy Procedia
Energy Procedia (2011) 000–000 Energy00Procedia 12 (2011) 570 – 577 www.elsevier.com/locate/procedia
ICSGCE 2011: 27–30 September 2011, Chengdu, China
Integrated Solar Controller for Solar Powered Off-grid Lighting System Zhong Yi Hea,b, Hong Chenb* a
School of Electronic, Information and Electrical Engineering, Shanghai Jiao Tong University, 800 Dong Chuan Road, Shanghai, China b Philips Research Asia – Shanghai, 10, Lane 888, Tian Lin Road, Shanghai, China
Abstract An integrated solar controller for off-grid lighting system is proposed to synthesize battery charger and light emitting diode (LED) driver with the benefit of simplified system architecture and reduced system cost. Based on bidirectional converter (BDC) with Sepic and Zeta topologies, high efficient solar energy collection is realized by maximum power point tracking (MPPT) method via battery parameter, and constant current driving for LED module is implemented by digital hysteresis control strategy. Theoretical analysis and experiment results are presented to verify the validity and feasibility of the proposed control.
©©2011 Ltd. Selection and/or peer-review under responsibility of University of Electronic 2011Published PublishedbybyElsevier Elsevier Ltd. Selection and/or peer-review under responsibility of ICSGCE 2011 Science and Technology of China (UESTC). Keywords: Photovoltaic (PV), Maximum power point tracking (MPPT), Controller, Light emitting diode (LED), Bi-directional converter (BDC)
Introduction Solar photovoltaic (PV) panel by converting solar radiation into DC electricity using semiconductors that exhibit the photovoltaic effect is suitable for small-scale solar application system because of its implementation flexibility. The output power of solar panel depends on solar insolation level and PV module temperature, as well as load property. The control method of maximum power point tracking (MPPT) enables the solar charge controller to track the MPP under any input and output conditions. Many MPPT methods have been developed and implemented. The methods vary in complexity, sensors required, convergence speed, cost, implementation hardware, range of effectiveness, popularity, and in other respects [1]-[3].
* Corresponding author. Tel.: +86-21-24127131. E-mail address: brian.chen@ philips.com.
1876-6102 © 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of University of Electronic Science and Technology of China (UESTC). doi:10.1016/j.egypro.2011.10.077
Zhong Yi HeYiand / Energy Procedia 12 (2011) 570 – 577 Zhong HeHong et al. Chen / Energy Procedia 00 (2011) 000–000
2
Light emitting diodes (LEDs) have been widely used in many products such as liquid crystal display (LCD) panel backlighting [4] and street lighting [5]. The significant improvements achieved for highpower or high-brightness LEDs are gradually realizing the possibility of replacing conventional light sources based on heated filaments and gas discharges with high-power LEDs [6]. The emission intensity of LEDs varies linearly with the forward current for small currents, but it shows a tendency to saturate at high currents. This phenomenon implies that the efficacy of an LED is lower if operated at high forward currents [7]. Two commonly used driving techniques for LEDs employ DC and PWM current driving with their inherent advantages and disadvantages [8]. As a typical solar powered off-grid application, lighting system converts electricity generated by solar panel to light load. In conventional solar lighting system, charger and driver are independent controllers and responsible for energy collection and energy utilization by interacting with battery. Many solar powered off-grid lighting systems have the feature of time-separate energy harvest stage and energy utilization stage e.g. solar street lamp and solar landscape lights. In these applications, single solar controller integrating battery charging and discharging functions is attractive because it makes full use of converter hardware and simplifies system configuration. An integrated solar controller for off-grid lighting system is proposed. Based on Sepic-type bidirectional converter (BDC), both battery charging and LED driving are realized with improved flexibility of voltage level matching among solar PV panel, battery, and LED module. A modified perturb & observe (P&O) MPPT control via battery parameters is proposed and implemented to improve control performance with reduced implementation cost [9]. Digital hysteresis constant current control for driving LED module is also analyzed and verified by experiments. System Architecture Optimization Typical solar off-grid lighting system consists of solar panel, solar charge controller, battery, and light source. Optimized system architecture brings benefit to customer with decreased cost, improved reliability, and enhanced flexibility. Based on BDC, the system architecture of solar powered off-grid lighting is shown in Fig. 1. During daytime when solar irradiance is available, PV panel charges battery through diode DPV and BDC with battery switch Sbat closed and LED switch SLED open. Power flows from PV panel to battery. During nighttime when sunlight is not available and light is needed, PV panel doesn’t work, and battery supplies energy to LED module through LED switch SLED with battery switch Sbat closed. Power flows from battery to LED module. Improved flexibility will be obtained by changing the stepping-up or stepping-down topology with BDC such as Sepic topology which has both stepping-up and stepping-down functions. Solar off-grid lighting system based on Sepic-type BDC is shown in Fig. 2, in which battery switch and LED switch are Dpv
Dpv PV
SLED PV LED
Integrated solar controller (BDC)
La
LED
Va
Ca
Qa Cc
Sbat
Bat
Fig. 1. System architecture of solar lighting system based on BDC
Qb Bat
Vb
Cb
Lb
Fig. 2. Solar lighting system based on Sepic-type BDC
571
572
ZhongYiYiHe Heetand Chen / Energy00Procedia 12 (2011) 570 – 577 Zhong al. Hong / Energy Procedia (2011) 000–000
omitted for simplicity. Battery Charging by Novel MPPT Method Perturb and observe (P&O) MPPT method is adopted for its important advantages as simplicity and applicability to almost any PV system configuration. Conventional P&O technique requires to sense PV panel output voltage and output current for MPPT implementation. In off-grid solar system with battery as energy storage, the charging current and battery need to be monitored to realize charging control. By taking the PV panel and the MPPT converter as an integrated energy source, the MPPT algorithm could be implemented from the battery side to obtain the maximum charging power without the information of PV panel. Less sensing circuitry contributes to decrease the complexity and cost of the MPPT controller in off-grid solar applications. The MPPT control method based on load parameters have proved the feasibility of the P&O MPPT implementation via load parameters. However, the optimization of transient MPPT behaviour has not been covered by these literatures [10]-[12]. In practical off-grid solar applications, the control objective is to maximize the power/energy flow delivered to the load, e.g. energy storage. From this point of view, it is reasonable to choose load power instead of PV power as control variable and to estimate output power of PV panel by battery charging power. Analysis shows that the mathematical expression of typical single switch DC/DC converter suitable for solar charge controller is monotonous regarding duty cycle. In this way, it is possible to judge the operating point location without the information of PV panel. Improved P&O MPPT algorithm based on load parameter is implemented with the control flow chart in [9]. Both the continuous current mode (CCM) and the discontinuous current mode (DCM) of Sepic converter are studied, in which the definition of DCM is the sum of both inductors current has the duration of being zero. Solar charge controller based on Sepic topology is shown in Fig. 3, where the port connected with inductor La and power switch Qb are defined as port a and port b respectively, Va and Vb represent output voltage of PV panel and battery voltage, and Cc is coupling capacitor. There are two operating modes in CCM and three modes in DCM, respectively. In CCM, by applying voltage-second balance principle for La and Lb, Dpv
La
PV
Va
Ca
Qa Cc
Qb Bat
Vb
Fig. 3. Sepic-based solar charger
(
)
Va DQaTs + (Va − VCc − Vb ) 1 − DQa Ts = 0,
(1)
(
)
VCc DQaTs + ( −Vb ) 1 − DQa Ts =0,
Cb
Lb
3
573
Zhong Yi HeYiand / Energy Procedia 12 (2011) 570 – 577 Zhong HeHong et al. Chen / Energy Procedia 00 (2011) 000–000
4
(2) where DQa is duty cycle of Qa. DQa Vb = Va 1 − DQa
(3) VCc = Va
(4) By replacing Va with VPV and Vb with Vbat in (3), DQa Vbat . = VPV 1 − DQa
(5) For mode 3 in DCM, Qa is disabled while Qb is enabled, but no current flows through Qb. The average power is determined by P= P= a b
2 DQa Va2 1 1 + . 2 f s La Lb
(6) By replacing Va with VPV, Vb with Vbat, and Pb with Pbat, = Pbat
2 DQa 1 1 2 + VPV . 2 f s La Lb
(7)
vPV (10V/div) iLb (1A/div)
vCc (10V/div) vbat (5V/div)
vPV (10V/div) iLa (1A/div)
vCc (10V/div) vbat (5V/div)
The intersection of the parabola expressed by (7) and the P-V characteristic curve of PV panel determine the operating point in DCM. Because Vbat rises very slowly, the parabola shows the monotonicity between VPV and DQa in DCM. As for CCM, VPV is inversely proportional to DQa. For the whole duty cycle range, the larger the DQa is, the smaller the VPV will be. This phenomenon proves that it is applicable to judge VPV variation direction by means of comparing the DQa’s of two adjacent control intervals. The key waveforms of Sepic-based charger are shown in Fig. 4. The profile of VCc is almost the same as that of VPV. As for their mean value, VCc is about 0.5 V lower than VPV due to the forward voltage drop of schottky diode DPV in Fig. 3. The level of Vbat during the time interval of Qa off or Qb on is higher than that of Qa on or Qb off.
Time (5µs/div)
Time (5µs/div)
(a) Current of inductor a
(b) Current of inductor b
574
ZhongYiYiHe Heetand Chen / Energy00Procedia 12 (2011) 570 – 577 Zhong al. Hong / Energy Procedia (2011) 000–000
Fig. 4. Output waveforms of Sepic-based solar charger
The MPPT performance of Sepic-based solar charge controller is measured in the experimental setup, in which the energy source is a solar array simulator (SAS), and the data acquisition system (DAS) records the input and output data of the solar charge controller. The target solar panel is the STP070D12/SEA from Suntech. The MPPT efficiency of the Sepic-based charger is calculated in terms of [13] with the data sensed by the DAS. The static MPPT testing results are shown in Table 1, where the testing conditions include solar cell temperature and solar irradiance level. The dynamic testing data is at the same level as that of the static one, verifying the validity and feasibility of the MPPT control. Table 1: Static MPPT performance of Sepic-based charger Conditions
10 ºC & 200 W/m2
10 ºC & 600 W/m2
45 ºC & 200 W/m2
45 ºC & 600 W/m2
MPPT efficiency
94.9%
97.4%
94.7%
97.0%
LED Driving Based on Digital Hysteresis Control When power flows from battery to LED module, the circuit in Fig. 2 is actually a Zeta topology, with the schematic diagram re-drawn in Fig.5. The operating principle analysis of Zeta converter is similar to that of Sepic converter in Section III, with two operating modes in CCM and three modes in DCM. The definition of DCM and CCM in Zeta converter is whether the sum of inductor a current iLa and inductor b current iLb has the possibility of being zero. Qb Bat
Vb
Cb
Lb Cc La
Va
Ca
Qa
Fig. 5. Zeta-based solar charger
By applying voltage-second balance principle for La and Lb in CCM, DQb Va , = Vb 1 − DQb (8) VCc = Va ,
(9) where DQb is duty cycle of Qb. For mode 3 in DCM, Qb is disabled while Qa is enabled, but no current flows through Qa. The average power is given by 2 DQb Vb2 1 1 P= P = + . b a 2 f s Lb La (10)
5
Zhong Yi HeYiand / Energy Procedia 12 (2011) 570 – 577 Zhong HeHong et al. Chen / Energy Procedia 00 (2011) 000–000
6
The system setup based on Zeta topology is shown in Fig. 6 (a), where lead acid battery specification is 12 V 100 Ah, and LED module consists of 15 pieces of Philips Lumileds Luxeon LXK2-PB12-L00. The prototype of LED driver shares the same controller as that of solar charger shown in Fig. 6 (b), where the SAS charges the same battery through Sepic-based charger circuit. The LED driver in Fig. 6 (a) works at constant current control mode. In conventional implementation of constant current driving, the practical LED driving current is sensed and compared with the current reference, and the error signal is processed by a proportional and integral (PI) type current regulator. In Zeta-based driver, another type of current regulator is studied and implemented based on digital hysteresis control, as is shown in Fig. 7, where ILED is practical driving current of LED module, Iset is current reference, ΔIth1 is threshold of current adjustment, and ΔIth2 is larger threshold that ΔIth1 for determining the perturbation steps (ΔD2 > ΔD1) of DQb.
(a) LED driver
(b) Solar charger
Fig. 6. Sepic-Zeta based BDC Start N
Timing reached for hysteresis current control? Y Reset timing count and timing again Sample ILED Y
Y
ILED - Iset < ΔIth1?
N
ILED - Iset ≥ -ΔIth1? N ILED - Iset < ΔIth2?
ILED - Iset ≥ -ΔIth2? Y D + ΔD1 → D
N D + ΔD2 → D
Y D - ΔD1 → D
N D - ΔD2 → D
Return
Fig. 7. Control flow chart of proposed digital hysteresis constant current control
The analysis of the two LED driving techniques shows that the proposed one has the merits of lower control processor requirements because no multiplication operation is needed to implement the constant driving control. The conventional PI method has the advantage of errorless control in theory. The experimental results of Zeta-based LED driver with constant current control implementation shown by Fig. 7 are presented in Fig. 8, where VLED is terminal voltage of LED module. The setting point of current reference is 350 mA.
575
576
vLED (10V/div)
iLED (0.1A/div)
iLED (0.1A/div) vbat (5V/div)
vCc (20V/div) vLED (10V/div)
ZhongYiYiHe Heetand Chen / Energy00Procedia 12 (2011) 570 – 577 Zhong al.Hong / Energy Procedia (2011) 000–000
Time (5µs/div)
(a) Instantaneous output
Time (5µs/div)
(b) Regulation process
Fig. 8. Output waveforms of Zeta-based LED driver
In Fig. 8 (a), the mean value of ILED is very approximate to the current reference, showing the validity of proposed control. When the power switch Qb is turned off, lead acid battery is in the empty load state, and its terminal voltage will rise comparing to battery voltage with load according to the properties of lead acid battery. In Fig. 8 (b), ILED fluctuates around the setting point of 350 mA. The ripple component of ILED could be further suppressed by fine-tuning control parameters in Fig. 7. Conclusion An integrated solar controller synthesizing both battery charging and LED driving functions for offgrid lighting system is proposed. The BDC for the integrated solar controller with Sepic and Zeta modes improves the control flexibility of voltage level matching among solar PV panel, battery, and LED module. Analysis shows that there is always a monotonic relationship between the operating voltage of PV panel and duty cycle of power switch regardless of inductor current continuous or discontinuous modes. Either static or dynamic MPPT efficiency of the integrated solar controller in charging mode is quite good. The digital hysteresis constant control of the integrated solar controller in driving mode is proposed and verified by experimental results. Acknowledgment The authors would like to thank Dr. Yongqin Zeng of Philips Research Asia Shanghai and Prof. Junmin Pan of Shanghai Jiaotong University for their help in this study. References [1] T. Esram and P. L. Chapman, “Comparison of photovoltaic array maximum power point tracking techniques,” IEEE Trans. Energy Conversion, Vol. 22, No. 2, 2007, pp. 439-449. [2] R. Faranda, S. Leva, and V. Maugeri, “MPPT techniques for PV systems: energetic and cost comparison,” in Proc. IEEE Power and Energy Society General Meeting – Conversion and Delivery of Electrical Energy in the 21st Century, 2008, pp. 1-6. [3] R. Faranda and S. Leva, “Energy comparison of MPPT techniques for PV systems,” WSEAS Trans. Power Systems, 2008, Vol. 3, No. 6, pp. 446-455. [4] C. Y. Wu, T. F. Wu, J. R. Tsai, Y. M. Chen, and C. C. Chen, “Multistring LED backlight driving system for LCD panels with color sequential display and area control,” IEEE Trans. Industrial Electronics, 2008, Vol. 55, No. 10, pp. 3791-3800. [5] X. Long, R. Liao, and J. Zhou, “Development of street lighting system-based novel high-brightness LED modules,” IET Optoelectron., 2009, Vol. 3, Iss. 1, pp. 40-46. [6] M. R. Krames, H. Amano, J. J. Brown, and P. L. Heremans, “Introduction to the issue on high-efficiency light-emitting diodes,” IEEE J. Selected Topics in Quantum Electronics, 2002, Vol. 8, Iss. 2, pp. 185-188.
7
Zhong Yi HeYiand / Energy Procedia 12 (2011) 570 – 577 Zhong HeHong et al. Chen / Energy Procedia 00 (2011) 000–000
8
[7] K. H. Loo, W. K. Lun, S. C. Tan, Y. M. Lai, and C. K. Tse, “On driving techniques for LEDs: toward a generalized methodology,” IEEE Trans. Power Electronics, Vol. 24, No. 12, pp. 2967-2976. [8] W. K. Lun, K. H. Loo, S. C. Tan, Y. M. Lai, and C. K. Tse, “Bilevel current driving technique for LEDs,” IEEE Trans. Power Electronics, 2009, Vol. 24, No. 12, pp. 2920-2932. [9] Z. Y. He, C. Y. Liu, Y. D. Liu, Y. Q. Zeng, and J. M. Pan, “Low cost MPPT controller for off grid solar applications,” in Proc. IEEE ICEMS, 2010, pp. 447-451. [10] D. Shmilovitz, “On the control of photovoltaic maximum power point tracker via output parameters,” IEE Proc. Electric Power Applications, 2005, Vol. 152, No. 2, pp. 239-248. [11] D. Shmilovitz, “Photovoltaic maximum power point tracking employing load parameters,” in Proc. IEEE ISIE, 2005, pp. 1037-1042. [12] C. Fu, M. Chen, Y. L. Shen, and S. J. Yu, “A control method of maximum power point based on output parameters,” Transactions of China Electrotechnical Society, 2007, Vol. 22, No. 2, pp. 148-152 (in Chinese). [13] European Standard EN 50530, “Overall efficiency of grid connected photovoltaic inverters,” CENELEC 2010.
577
Energy Procedia
Energy Procedia (2011) 000–000 Energy00Procedia 12 (2011) 570 – 577 www.elsevier.com/locate/procedia
ICSGCE 2011: 27–30 September 2011, Chengdu, China
Integrated Solar Controller for Solar Powered Off-grid Lighting System Zhong Yi Hea,b, Hong Chenb* a
School of Electronic, Information and Electrical Engineering, Shanghai Jiao Tong University, 800 Dong Chuan Road, Shanghai, China b Philips Research Asia – Shanghai, 10, Lane 888, Tian Lin Road, Shanghai, China
Abstract An integrated solar controller for off-grid lighting system is proposed to synthesize battery charger and light emitting diode (LED) driver with the benefit of simplified system architecture and reduced system cost. Based on bidirectional converter (BDC) with Sepic and Zeta topologies, high efficient solar energy collection is realized by maximum power point tracking (MPPT) method via battery parameter, and constant current driving for LED module is implemented by digital hysteresis control strategy. Theoretical analysis and experiment results are presented to verify the validity and feasibility of the proposed control.
©©2011 Ltd. Selection and/or peer-review under responsibility of University of Electronic 2011Published PublishedbybyElsevier Elsevier Ltd. Selection and/or peer-review under responsibility of ICSGCE 2011 Science and Technology of China (UESTC). Keywords: Photovoltaic (PV), Maximum power point tracking (MPPT), Controller, Light emitting diode (LED), Bi-directional converter (BDC)
Introduction Solar photovoltaic (PV) panel by converting solar radiation into DC electricity using semiconductors that exhibit the photovoltaic effect is suitable for small-scale solar application system because of its implementation flexibility. The output power of solar panel depends on solar insolation level and PV module temperature, as well as load property. The control method of maximum power point tracking (MPPT) enables the solar charge controller to track the MPP under any input and output conditions. Many MPPT methods have been developed and implemented. The methods vary in complexity, sensors required, convergence speed, cost, implementation hardware, range of effectiveness, popularity, and in other respects [1]-[3].
* Corresponding author. Tel.: +86-21-24127131. E-mail address: brian.chen@ philips.com.
1876-6102 © 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of University of Electronic Science and Technology of China (UESTC). doi:10.1016/j.egypro.2011.10.077
Zhong Yi HeYiand / Energy Procedia 12 (2011) 570 – 577 Zhong HeHong et al. Chen / Energy Procedia 00 (2011) 000–000
2
Light emitting diodes (LEDs) have been widely used in many products such as liquid crystal display (LCD) panel backlighting [4] and street lighting [5]. The significant improvements achieved for highpower or high-brightness LEDs are gradually realizing the possibility of replacing conventional light sources based on heated filaments and gas discharges with high-power LEDs [6]. The emission intensity of LEDs varies linearly with the forward current for small currents, but it shows a tendency to saturate at high currents. This phenomenon implies that the efficacy of an LED is lower if operated at high forward currents [7]. Two commonly used driving techniques for LEDs employ DC and PWM current driving with their inherent advantages and disadvantages [8]. As a typical solar powered off-grid application, lighting system converts electricity generated by solar panel to light load. In conventional solar lighting system, charger and driver are independent controllers and responsible for energy collection and energy utilization by interacting with battery. Many solar powered off-grid lighting systems have the feature of time-separate energy harvest stage and energy utilization stage e.g. solar street lamp and solar landscape lights. In these applications, single solar controller integrating battery charging and discharging functions is attractive because it makes full use of converter hardware and simplifies system configuration. An integrated solar controller for off-grid lighting system is proposed. Based on Sepic-type bidirectional converter (BDC), both battery charging and LED driving are realized with improved flexibility of voltage level matching among solar PV panel, battery, and LED module. A modified perturb & observe (P&O) MPPT control via battery parameters is proposed and implemented to improve control performance with reduced implementation cost [9]. Digital hysteresis constant current control for driving LED module is also analyzed and verified by experiments. System Architecture Optimization Typical solar off-grid lighting system consists of solar panel, solar charge controller, battery, and light source. Optimized system architecture brings benefit to customer with decreased cost, improved reliability, and enhanced flexibility. Based on BDC, the system architecture of solar powered off-grid lighting is shown in Fig. 1. During daytime when solar irradiance is available, PV panel charges battery through diode DPV and BDC with battery switch Sbat closed and LED switch SLED open. Power flows from PV panel to battery. During nighttime when sunlight is not available and light is needed, PV panel doesn’t work, and battery supplies energy to LED module through LED switch SLED with battery switch Sbat closed. Power flows from battery to LED module. Improved flexibility will be obtained by changing the stepping-up or stepping-down topology with BDC such as Sepic topology which has both stepping-up and stepping-down functions. Solar off-grid lighting system based on Sepic-type BDC is shown in Fig. 2, in which battery switch and LED switch are Dpv
Dpv PV
SLED PV LED
Integrated solar controller (BDC)
La
LED
Va
Ca
Qa Cc
Sbat
Bat
Fig. 1. System architecture of solar lighting system based on BDC
Qb Bat
Vb
Cb
Lb
Fig. 2. Solar lighting system based on Sepic-type BDC
571
572
ZhongYiYiHe Heetand Chen / Energy00Procedia 12 (2011) 570 – 577 Zhong al. Hong / Energy Procedia (2011) 000–000
omitted for simplicity. Battery Charging by Novel MPPT Method Perturb and observe (P&O) MPPT method is adopted for its important advantages as simplicity and applicability to almost any PV system configuration. Conventional P&O technique requires to sense PV panel output voltage and output current for MPPT implementation. In off-grid solar system with battery as energy storage, the charging current and battery need to be monitored to realize charging control. By taking the PV panel and the MPPT converter as an integrated energy source, the MPPT algorithm could be implemented from the battery side to obtain the maximum charging power without the information of PV panel. Less sensing circuitry contributes to decrease the complexity and cost of the MPPT controller in off-grid solar applications. The MPPT control method based on load parameters have proved the feasibility of the P&O MPPT implementation via load parameters. However, the optimization of transient MPPT behaviour has not been covered by these literatures [10]-[12]. In practical off-grid solar applications, the control objective is to maximize the power/energy flow delivered to the load, e.g. energy storage. From this point of view, it is reasonable to choose load power instead of PV power as control variable and to estimate output power of PV panel by battery charging power. Analysis shows that the mathematical expression of typical single switch DC/DC converter suitable for solar charge controller is monotonous regarding duty cycle. In this way, it is possible to judge the operating point location without the information of PV panel. Improved P&O MPPT algorithm based on load parameter is implemented with the control flow chart in [9]. Both the continuous current mode (CCM) and the discontinuous current mode (DCM) of Sepic converter are studied, in which the definition of DCM is the sum of both inductors current has the duration of being zero. Solar charge controller based on Sepic topology is shown in Fig. 3, where the port connected with inductor La and power switch Qb are defined as port a and port b respectively, Va and Vb represent output voltage of PV panel and battery voltage, and Cc is coupling capacitor. There are two operating modes in CCM and three modes in DCM, respectively. In CCM, by applying voltage-second balance principle for La and Lb, Dpv
La
PV
Va
Ca
Qa Cc
Qb Bat
Vb
Fig. 3. Sepic-based solar charger
(
)
Va DQaTs + (Va − VCc − Vb ) 1 − DQa Ts = 0,
(1)
(
)
VCc DQaTs + ( −Vb ) 1 − DQa Ts =0,
Cb
Lb
3
573
Zhong Yi HeYiand / Energy Procedia 12 (2011) 570 – 577 Zhong HeHong et al. Chen / Energy Procedia 00 (2011) 000–000
4
(2) where DQa is duty cycle of Qa. DQa Vb = Va 1 − DQa
(3) VCc = Va
(4) By replacing Va with VPV and Vb with Vbat in (3), DQa Vbat . = VPV 1 − DQa
(5) For mode 3 in DCM, Qa is disabled while Qb is enabled, but no current flows through Qb. The average power is determined by P= P= a b
2 DQa Va2 1 1 + . 2 f s La Lb
(6) By replacing Va with VPV, Vb with Vbat, and Pb with Pbat, = Pbat
2 DQa 1 1 2 + VPV . 2 f s La Lb
(7)
vPV (10V/div) iLb (1A/div)
vCc (10V/div) vbat (5V/div)
vPV (10V/div) iLa (1A/div)
vCc (10V/div) vbat (5V/div)
The intersection of the parabola expressed by (7) and the P-V characteristic curve of PV panel determine the operating point in DCM. Because Vbat rises very slowly, the parabola shows the monotonicity between VPV and DQa in DCM. As for CCM, VPV is inversely proportional to DQa. For the whole duty cycle range, the larger the DQa is, the smaller the VPV will be. This phenomenon proves that it is applicable to judge VPV variation direction by means of comparing the DQa’s of two adjacent control intervals. The key waveforms of Sepic-based charger are shown in Fig. 4. The profile of VCc is almost the same as that of VPV. As for their mean value, VCc is about 0.5 V lower than VPV due to the forward voltage drop of schottky diode DPV in Fig. 3. The level of Vbat during the time interval of Qa off or Qb on is higher than that of Qa on or Qb off.
Time (5µs/div)
Time (5µs/div)
(a) Current of inductor a
(b) Current of inductor b
574
ZhongYiYiHe Heetand Chen / Energy00Procedia 12 (2011) 570 – 577 Zhong al. Hong / Energy Procedia (2011) 000–000
Fig. 4. Output waveforms of Sepic-based solar charger
The MPPT performance of Sepic-based solar charge controller is measured in the experimental setup, in which the energy source is a solar array simulator (SAS), and the data acquisition system (DAS) records the input and output data of the solar charge controller. The target solar panel is the STP070D12/SEA from Suntech. The MPPT efficiency of the Sepic-based charger is calculated in terms of [13] with the data sensed by the DAS. The static MPPT testing results are shown in Table 1, where the testing conditions include solar cell temperature and solar irradiance level. The dynamic testing data is at the same level as that of the static one, verifying the validity and feasibility of the MPPT control. Table 1: Static MPPT performance of Sepic-based charger Conditions
10 ºC & 200 W/m2
10 ºC & 600 W/m2
45 ºC & 200 W/m2
45 ºC & 600 W/m2
MPPT efficiency
94.9%
97.4%
94.7%
97.0%
LED Driving Based on Digital Hysteresis Control When power flows from battery to LED module, the circuit in Fig. 2 is actually a Zeta topology, with the schematic diagram re-drawn in Fig.5. The operating principle analysis of Zeta converter is similar to that of Sepic converter in Section III, with two operating modes in CCM and three modes in DCM. The definition of DCM and CCM in Zeta converter is whether the sum of inductor a current iLa and inductor b current iLb has the possibility of being zero. Qb Bat
Vb
Cb
Lb Cc La
Va
Ca
Qa
Fig. 5. Zeta-based solar charger
By applying voltage-second balance principle for La and Lb in CCM, DQb Va , = Vb 1 − DQb (8) VCc = Va ,
(9) where DQb is duty cycle of Qb. For mode 3 in DCM, Qb is disabled while Qa is enabled, but no current flows through Qa. The average power is given by 2 DQb Vb2 1 1 P= P = + . b a 2 f s Lb La (10)
5
Zhong Yi HeYiand / Energy Procedia 12 (2011) 570 – 577 Zhong HeHong et al. Chen / Energy Procedia 00 (2011) 000–000
6
The system setup based on Zeta topology is shown in Fig. 6 (a), where lead acid battery specification is 12 V 100 Ah, and LED module consists of 15 pieces of Philips Lumileds Luxeon LXK2-PB12-L00. The prototype of LED driver shares the same controller as that of solar charger shown in Fig. 6 (b), where the SAS charges the same battery through Sepic-based charger circuit. The LED driver in Fig. 6 (a) works at constant current control mode. In conventional implementation of constant current driving, the practical LED driving current is sensed and compared with the current reference, and the error signal is processed by a proportional and integral (PI) type current regulator. In Zeta-based driver, another type of current regulator is studied and implemented based on digital hysteresis control, as is shown in Fig. 7, where ILED is practical driving current of LED module, Iset is current reference, ΔIth1 is threshold of current adjustment, and ΔIth2 is larger threshold that ΔIth1 for determining the perturbation steps (ΔD2 > ΔD1) of DQb.
(a) LED driver
(b) Solar charger
Fig. 6. Sepic-Zeta based BDC Start N
Timing reached for hysteresis current control? Y Reset timing count and timing again Sample ILED Y
Y
ILED - Iset < ΔIth1?
N
ILED - Iset ≥ -ΔIth1? N ILED - Iset < ΔIth2?
ILED - Iset ≥ -ΔIth2? Y D + ΔD1 → D
N D + ΔD2 → D
Y D - ΔD1 → D
N D - ΔD2 → D
Return
Fig. 7. Control flow chart of proposed digital hysteresis constant current control
The analysis of the two LED driving techniques shows that the proposed one has the merits of lower control processor requirements because no multiplication operation is needed to implement the constant driving control. The conventional PI method has the advantage of errorless control in theory. The experimental results of Zeta-based LED driver with constant current control implementation shown by Fig. 7 are presented in Fig. 8, where VLED is terminal voltage of LED module. The setting point of current reference is 350 mA.
575
576
vLED (10V/div)
iLED (0.1A/div)
iLED (0.1A/div) vbat (5V/div)
vCc (20V/div) vLED (10V/div)
ZhongYiYiHe Heetand Chen / Energy00Procedia 12 (2011) 570 – 577 Zhong al.Hong / Energy Procedia (2011) 000–000
Time (5µs/div)
(a) Instantaneous output
Time (5µs/div)
(b) Regulation process
Fig. 8. Output waveforms of Zeta-based LED driver
In Fig. 8 (a), the mean value of ILED is very approximate to the current reference, showing the validity of proposed control. When the power switch Qb is turned off, lead acid battery is in the empty load state, and its terminal voltage will rise comparing to battery voltage with load according to the properties of lead acid battery. In Fig. 8 (b), ILED fluctuates around the setting point of 350 mA. The ripple component of ILED could be further suppressed by fine-tuning control parameters in Fig. 7. Conclusion An integrated solar controller synthesizing both battery charging and LED driving functions for offgrid lighting system is proposed. The BDC for the integrated solar controller with Sepic and Zeta modes improves the control flexibility of voltage level matching among solar PV panel, battery, and LED module. Analysis shows that there is always a monotonic relationship between the operating voltage of PV panel and duty cycle of power switch regardless of inductor current continuous or discontinuous modes. Either static or dynamic MPPT efficiency of the integrated solar controller in charging mode is quite good. The digital hysteresis constant control of the integrated solar controller in driving mode is proposed and verified by experimental results. Acknowledgment The authors would like to thank Dr. Yongqin Zeng of Philips Research Asia Shanghai and Prof. Junmin Pan of Shanghai Jiaotong University for their help in this study. References [1] T. Esram and P. L. Chapman, “Comparison of photovoltaic array maximum power point tracking techniques,” IEEE Trans. Energy Conversion, Vol. 22, No. 2, 2007, pp. 439-449. [2] R. Faranda, S. Leva, and V. Maugeri, “MPPT techniques for PV systems: energetic and cost comparison,” in Proc. IEEE Power and Energy Society General Meeting – Conversion and Delivery of Electrical Energy in the 21st Century, 2008, pp. 1-6. [3] R. Faranda and S. Leva, “Energy comparison of MPPT techniques for PV systems,” WSEAS Trans. Power Systems, 2008, Vol. 3, No. 6, pp. 446-455. [4] C. Y. Wu, T. F. Wu, J. R. Tsai, Y. M. Chen, and C. C. Chen, “Multistring LED backlight driving system for LCD panels with color sequential display and area control,” IEEE Trans. Industrial Electronics, 2008, Vol. 55, No. 10, pp. 3791-3800. [5] X. Long, R. Liao, and J. Zhou, “Development of street lighting system-based novel high-brightness LED modules,” IET Optoelectron., 2009, Vol. 3, Iss. 1, pp. 40-46. [6] M. R. Krames, H. Amano, J. J. Brown, and P. L. Heremans, “Introduction to the issue on high-efficiency light-emitting diodes,” IEEE J. Selected Topics in Quantum Electronics, 2002, Vol. 8, Iss. 2, pp. 185-188.
7
Zhong Yi HeYiand / Energy Procedia 12 (2011) 570 – 577 Zhong HeHong et al. Chen / Energy Procedia 00 (2011) 000–000
8
[7] K. H. Loo, W. K. Lun, S. C. Tan, Y. M. Lai, and C. K. Tse, “On driving techniques for LEDs: toward a generalized methodology,” IEEE Trans. Power Electronics, Vol. 24, No. 12, pp. 2967-2976. [8] W. K. Lun, K. H. Loo, S. C. Tan, Y. M. Lai, and C. K. Tse, “Bilevel current driving technique for LEDs,” IEEE Trans. Power Electronics, 2009, Vol. 24, No. 12, pp. 2920-2932. [9] Z. Y. He, C. Y. Liu, Y. D. Liu, Y. Q. Zeng, and J. M. Pan, “Low cost MPPT controller for off grid solar applications,” in Proc. IEEE ICEMS, 2010, pp. 447-451. [10] D. Shmilovitz, “On the control of photovoltaic maximum power point tracker via output parameters,” IEE Proc. Electric Power Applications, 2005, Vol. 152, No. 2, pp. 239-248. [11] D. Shmilovitz, “Photovoltaic maximum power point tracking employing load parameters,” in Proc. IEEE ISIE, 2005, pp. 1037-1042. [12] C. Fu, M. Chen, Y. L. Shen, and S. J. Yu, “A control method of maximum power point based on output parameters,” Transactions of China Electrotechnical Society, 2007, Vol. 22, No. 2, pp. 148-152 (in Chinese). [13] European Standard EN 50530, “Overall efficiency of grid connected photovoltaic inverters,” CENELEC 2010.
577