Enhancing the design of battery charging controllers for photovoltaic ...

37 downloads 10392 Views 3MB Size Report
Enhancing the design of battery charging controllers for photovoltaic systems. Article · May 2016 with 879 Reads. DOI: 10.1016/j.rser.2015.12.061. Yasser E ...
Renewable and Sustainable Energy Reviews 58 (2016) 646–655

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews journal homepage: www.elsevier.com/locate/rser

Enhancing the design of battery charging controllers for photovoltaic systems Yasser E. Abu Eldahab a, Naggar H. Saad b, Abdalhalim Zekry c a

Electrical Engineer at Arab Contractor Co., Cairo, Egypt Electrical Power and Machines Department, Faculty of Engineering, Ain Shams University, Cairo, Egypt c Electronics and Communications Department, Faculty of Engineering, Ain Shams University, Cairo, Egypt b

art ic l e i nf o

a b s t r a c t

Article history: Received 1 August 2015 Received in revised form 14 December 2015 Accepted 15 December 2015

Batteries are the power tank of solar power systems. They play the role of power supply when the sun does not shine. This paper provides a review of battery charging control techniques for photovoltaic systems. In addition, it presents a new battery charge controller that keeps on the good features and resolves the drawbacks and limitations of the traditional controllers. The new controller is based on a newly developed maximum power point tracking (MPPT) technique enabling very fast maximum power point (MPP) capture. Moreover, it utilizes the constant current, constant voltage (CCCV) charging scheme to reduce the battery charging time. In addition, it enables accessing all system parameters remotely for monitoring and administration purposes. In order to determine the performance parameters of the proposed controller, a prototype was implemented together with microcontroller based DC–DC converter. The experimental results show that, the new controller tracks the MPP faster than the conventional controllers do. Moreover, the charging period is significantly reduced. Moreover, the proposed controller has high accuracy and minimizes the steady state oscillation errors around the target MPP. & 2015 Elsevier Ltd. All rights reserved.

Keywords: Renewable energy Photovoltaic system Charger controller Solar regulator Battery charger Maximum power point

Contents 1. 2.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646 Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647 2.1. System description and operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648 2.1.1. Improving the maximum power point tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648 2.1.2. Reducing the charging time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649 2.2. System hardware structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649 2.3. I–V and P–V characteristics of the PV module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649 2.4. System software. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650 3. Experiment results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651 3.1. Test case I, fixed temperature, and irradiance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652 3.2. Test case II, varying temperature, and/or irradiance with time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652 3.3. Test case III, the new MPPT test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654 3.4. Charging process test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654 3.5. Validation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654 4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655

1. Introduction

E-mail address: [email protected] (Y.E. Abu Eldahab). http://dx.doi.org/10.1016/j.rser.2015.12.061 1364-0321/& 2015 Elsevier Ltd. All rights reserved.

Photovoltaic (PV) systems have high fabrication cost and low energy conversion efficiency due to their nonlinear and atmosphere dependent current to voltage (I–V) and power to voltage

Y.E. Abu Eldahab et al. / Renewable and Sustainable Energy Reviews 58 (2016) 646–655

(P–V) characteristic [1]. Therefore, the maximum output power changes with the incident solar radiation and weather conditions especially the temperature. Moreover, the location of the MPP on the I–V curve could not be easily located. Therefore, it must be determined either through calculation models or by search algorithms. In addition, the process of tracking the maximum power point should be very fast to deal with the fast changes in temperature and or irradiance. The partial shading is not our addressed points. But, it is one of the cases that need fast tracking response. In partial shading, objects like planes, trees, clouds, or buildings cover the sun partially or wholly. Consequently, the MPP may change suddenly and rapidly. In order to maximize the power transfer from the photovoltaic array to the battery bank, a battery charger with charge controller should be utilized. It performs two main functions. The first one is tracking accurately the maximum power point (MPP) so fast in order to keep the operating point of the PV panels at the MPP for the most of the time. The other function is minimizing the battery charging time to back up the PV arrays as fast as possible. In addition, it should protect the battery from overcharging and under discharging. The algorithm of a battery charge controller determines the effectiveness of battery charging as well as the PV array utilization, and ultimately the ability of the system to meet the electrical load demands. The most common approaches for charge controllers are the shunt, series, pulse width modulation (PWM) and MPPT charge controllers. The shunt regulator controls the charging of a battery from the PV array by short-circuiting the array internal to the controller. The series controller utilizes some type of control element connected in series between the array and the battery. While this type of controller is commonly used in small PV systems, it is also the practical choice for larger systems due to the current limitations of shunt controllers. The MPPT battery charge controller incorporates a DC-to-DC converter such that the PV array can operate at the maximum power point at the prevailing solar irradiance [2]. The battery charging control methods are classified into two classes: single stage, and multi-stage method. The constant current charging is a good example for single stage method, while the constant current, constant voltage technique is a good example for multistage charging method. Studies show that, the multi-stage charging is the most efficient for battery charging regardless of the battery type [3]. The structure of battery charge controllers depends on the type of the controller. In the series and shunt controllers, it simply consists of a switching element, such as a relay that is switched on/off based on the value of a predefined set point. In a PWM and MPPT controllers, the circuits are more sophisticated. In PWM generator circuits or microcontrollers are needed in order to drive the switches of a DC–DC converter. However, the MPPT controller consists of a controller that manages the maximum power point tracking process and DC–DC converter [2]. This paper provides a review of the conventional battery charging control techniques pointing out their good features, drawbacks and limitations, as well. Then it is augmented by the design and practical implementation of a new charge controller that keeps on the good features and resolves the drawbacks and limitations of the traditional controllers. The new controller utilizes a new MPPT algorithm based on genetic neural algorithm (GA). The simulation results of this controller show that, it tracks the maximum power point much faster than the conventional ones [4]. On the other hand, the proposed battery charge controller uses the constant current–constant voltage as a charging scheme in order to reduce the charging time. Moreover, the proposed controller overcomes the other limitations such as depending on battery types, charging voltage levels, and size of PV modules. In addition, the proposed controller has high accuracy

647

and minimizes the steady state error with zero oscillation around the target MPP.

2. Materials and methods There are intensive and continuous research efforts on the design and implementation of the solar charger regulators to improve their performance parameters. The targets are: improving their efficiency, increasing their speed of maximum power point tracking and reducing the period of charging. In this section, the recent publications on the design and implementation of charger controllers will be reviewed to show their main features, drawbacks, and limitations. This would assist us to achieve advancements in the solar charger controller. A new MPPT technique was introduced in [4]. It has significant improvements in tracking the maximum power point thanks to employing the binary search instead of using linear search technique in the traditional MPPT methods. In addition, a review of different strategies, algorithms, and methods to implement a smart charging control system was presented in [5]. Furthermore, Ref. [6] analyzed the life cycle of three types of batteries, flowassisted nickel zinc-, manganese dioxide-, and valve-regulated lead-acid batteries. It introduced a comprehensive review and a full process-based life cycle analysis of batteries. This study provided good information for sizing and selecting the appropriate battery type. An optimized model of hybrid battery and energy storage system based on cooperative game model was proposed in [7]. This study is helpful in planning and designing of battery and energy storage station with the most economical types of batteries and optimal capacity configuration of energy storage station. In [8] the design considerations and evaluation of the performance of PV chargers used to charge major batteries including nickel–cadmium (Ni–Cd), nickel–metal-hydride (Ni–MH), lithium-ion (Li-ion) and sealed lead-acid batteries at real operating conditions are presented. Furthermore, in [9], a socio-technical approach was taken to understand the reasons for failure. A strategy was subsequently developed to influence user behavior and increase the PV array size to reduce capacity shortage through the year and improve the lifetime of the lead acid batteries found on these systems. Ref. [10] presented the potential of lithium ion (Li-ion) batteries to be the major energy storage in off-grid renewable energy. It introduced the electric vehicle sector as the driving force of Li-ion batteries in renewable energies. In addition, it presented the incomparable advantages of Li-ion batteries over other technologies even if some challenges are still to overcome for a wider usage in stationary energy storage. In addition, the impact of the charging methodology on the battery lifetime was investigated in [11]. Three charging techniques was used in this work: constant current (CC), constant current–constant voltage (CC–CV) and Constant Current–Constant Voltage with Negative Pulse (CC–CVNP). A comparative study between these techniques was presented in this research. In addition, the State of Health (SOH) determination for lithium ion batteries was discussed in [12]. In this study, the equivalent DC resistances of Lithium ion battery cells of various health conditions during charging under different temperatures were collected and the relationships between equivalent DC resistance, health condition, and working temperature were identified. Furthermore, Ref. [13] proposed a control strategy based on DC link voltage sensing for PV powered smart charging station. This system designed to charge the plug in hybrids electric vehicles (PHEVs) from grid-connected photovoltaic generation or the utility or both. Furthermore, Ref. [14] introduced an approach for estimating the open circuit voltage as a function of parameters of the electrical

648

Y.E. Abu Eldahab et al. / Renewable and Sustainable Energy Reviews 58 (2016) 646–655

equivalent model. This approach was based on the recursive algorithm of Kalman filter to estimate the state of charge of a battery dynamically. In addition, Ref. [15] presented an intelligent method of MPPT using fuzzy logic control for stand alone photovoltaic. Moreover, an Adaptive fuzzy controller based MPPT for photovoltaic systems was introduced in [16]. Ref. [17] demonstrated an evaluation of the performance of various types of MPPT algorithms under partial shading. Moreover, Ref. [18] presented a modified artificial bee colony based approach for mitigating the power loss in the PV module under partial shading effect. In addition, Ref. [19] proposed an MPPT method based on an Artificial Neural Network (ANN). It faced the tradeoff between the number of preselected power to voltage characteristic scansions, the size of the ANN and its prediction accuracy. Ref. [20] discussed a genetic algorithm based method to reconfigure the electrical connections between panels in order to fetch the optimum MPP. Moreover, an intelligent charge controller for prolonging battery life was introduced in [21]. It discussed the need for advantages of such charger controller and applied tests on a prototype. Ref. [22] proposed a low cost and fast solar charger by employing a buck converter with a digital signal processor. The work was tested by using the Elgar Terra as solar array simulator to verify the feasibility and validity of the system and its control algorithm. Moreover, a design and development of a microcontroller based solar charge controller was introduced in [23,24,25]. While Ref. [26] presented a design of MPPT based PV charger. In addition, an MPPT controller with control algorithms for the rotatable solar panel, which increases the efficiency of the plug-in hybrid electric vehicle (PHEV) by using rotatable solar panels, was given in [27]. This dual mechatronic MPPT controller is suitable for the PHEV system. Furthermore, a designed solar battery charge controller that combines both MPPT and over-voltage controls as a single control function was introduced in [28]. The designed controller was demonstrated to have good transient response with only small voltage overshoot. While Ref. [29] presented an analog MPPT controller for a solar system that utilizes the load current to achieve maximum output power from the solar panel. In addition, the modeling and control design of the PV charger system using a Buck-Boost converter was discussed in [30]. This controller designed to balance the power flow from PV module to the battery and the load such that PV power was utilized effectively. The battery was charged in three charging steps. A technique for extracting maximum power from a photovoltaic panel to charge the battery was introduced in [31]. This MPPT charge controllers can be used to utilize the maximum power output of solar panels instead of investing in number of

solar panels. In addition, an RS485 interface was included for monitoring purpose. While Ref. [32] introduced a practical bucktype power converter for PV system for energy storage application based on constant voltage MPPT algorithm. Simulation and experimental results demonstrate the effectiveness and validity of the proposed system. Furthermore, an efficient MPPT solar charge controller was presented to increase the efficiency of power transfer in comparison to systems with direct connection [33], thus reducing the size and the cost of the PV panel. In addition, a Peripheral Interface Controller (PIC) based solar charge controller for battery was discussed in [34]. From the outgoing literature survey on the photovoltaic battery charge controller, one can categorize the contributions as given in Table 1. As listed in Table 1, the limitations of the published controllers are slow tracking response, low accuracy and large steady state error, oscillation around the target MPP and long charging time, battery type dependency, and charging voltage level dependency. The next sections present the design, implementation, and validation of the proposed charge controller showing how it successfully keeps on the good features of the traditional controllers and resolving the drawbacks and limitations. Besides, showing the new features added such as remote monitoring and administration capabilities. 2.1. System description and operation The proposed system is shown in Fig. 1. It is composed of the PV array, the microcontroller based DC–DC boost converter, the battery bank, and the serial communication interface RS232. All of these components will be described in detail in the next section. In order to improve the tracking process, the new controller utilizes a new MPPT algorithm [4]. It is based on a genetic algorithm. Furthermore, it uses a multistage charging algorithm for reducing the charging time. 2.1.1. Improving the maximum power point tracking This method employs a genetic algorithm as optimization algorithm to search the training database for the nearest point to the MPP for specific operating condition. The training database initially contains some records coming out of real measurements and it grows up by recording every new case. Every record contains the optimum values of the maximum power P m , the maximum power point current I m , and the maximum power point voltage V m for a given insolation G, and an environment temperature T.

Table 1 List of contributions including the good features, drawbacks, and limitations. Ref.

Contribution

Limitations

– Review on the control methods and – They do not present any treatment and or solution for the addressed point. strategies of battery charging techniques – Simulation model for MPPT control techniques [20–34] Implementation and Experimental work. Every contribution has one or more of the following limitations – Working with specific type of battery – Depending on the PV module size – Output specific voltage level for charging – No MPPT control – Long charging time – High implementation cost – Low tracking speed and accuracy – Large steady state error and oscillation around the target [4–19]

Proposed controller It Introduced a practical and efficient solution avoiding the drawbacks and resolving the limitation of the traditional techniques while preserving the good features

– Designed based on a new , smart and fast MPPT control algorithm – Charging period is reduced thanks to utilizing CCCV charging algorithm – Capable of charging the major type of battery chemistries (nickel–cadmium (Ni–Cd), nickel–metal-hydride (Ni–MH), lithium-ion (Li-ion) and sealed lead-acid)

Y.E. Abu Eldahab et al. / Renewable and Sustainable Energy Reviews 58 (2016) 646–655

649

Fig. 3. Battery charging characteristic. Fig. 1. The battery charger controller block diagram.

Fig. 4. Circuit block diagram of the charger controller.

Fig. 2. The proposed search process for the GMPP second stage.

The principle of the new tracking method is using the binary search technique instead of searching every point in the PV curve from [0  V OC ] linearly. Every time the searching path is divided into two halves and only the point in the middle is checked to select a half as the new searching path and discards the other half. Fig. 2 demonstrates the search path staring from point 1 and ending at point 4 where the maximum power point resides .It shows four checkpoints. It starts from the middle point ( V2OC ) at the intersection of line number one and the PV curve. Then, it . If it determines the right direction to move based on the sign of dp dv is positive, search goes to the right to select the right half [V2OC  V OC ] and discards the other half [0  V2OC ]. The searching path is updated to be [V2OC  V OC ]. Again, it is divided into two halves. So, the next checkpoint will be (3V4OC ) at the intersection of line number two and the PV curve. The process is repeated until reaching the target MPP. This example shows that, there are only four iterations to reach the target point when it needs to check hundreds of points in the traditional linear search methods. The duty cycle of the boost converter is controlled by the pulse width modulation signal generated by the microcontroller. The width of these pulses are determined based on the values of the system parameters, which are the panel ambient temperature (T pv ), irradiance (G), panel current (I pv ) and voltage (V pv ), battery charging current (I bat ), battery voltage (V bat ) and battery temperature (T bat ). The system software program contains the logic of the MPPT algorithm, the GA algorithm, and the constant current constant voltage battery-charging algorithm. 2.1.2. Reducing the charging time On the other hand, the controller uses a multi-stage charging algorithm, which is the most safe and effective method of charging. The principle of operation is shown in Fig. 3. In the first stage, the battery voltage is increased gradually to the preset voltage level, which is called the bulk level using constant charge current. When the bulk level voltage is reached, the absorption stage starts. During this phase, the voltage is maintained at a bulk voltage level

for specific time while the current is gradually dropping. After the absorption time passes, the float stage begins. The voltage is lowered to the float level and the battery draws a very small current [35]. In the proposed controller, a voltage and current regulator circuits are utilized in order to achieve the charging process. 2.2. System hardware structure Fig. 4 shows the circuit block diagram of the proposed controller. It is composed of five main boards. The control board contains the 16F877A microcontroller and the LCD. The main function of this board is to generate the PWM pulses, which drive switches of the converter. The power board contains the DC–DC boost converter. It is designed according to the standard boost converter model equations in [36]. The input design parameters of the DC–DC boost converter are listed in Table 2. The calculated output parameters are depicted in Table 3. In addition, the practical values used in the proposed controller are listed in Table 4. The sensor board contains the system sensors and their peripheral circuit components. It comprises the temperature sensor LM35DZ, the light to frequency converter sensor TSL235, panel voltage sensor and battery voltage sensor. The final board is the battery charging current sensor board, which converts the sensed current to voltage signal acceptable as input signal to the microcontroller. The used S320P36 PV module with power of 80 W has the main parameters listed in Table 5. 2.3. I–V and P–V characteristics of the PV module It is very important to fully characterize the PV module by their I–V and P–V curves as an input power source to battery charging system. That will be taken as a guide in order to verify and validate the experimental results. Figs. 5 and 6 show the measured I–V curves of the panel, while Figs. 7 and 8 show P–V characteristic curves respectively. One of every pair is taken at fixed temperature and different irradiances, while the other at fixed irradiance and different temperatures.

650

Y.E. Abu Eldahab et al. / Renewable and Sustainable Energy Reviews 58 (2016) 646–655

Table 2 The input design parameters of the proposed boast converter. The The The The The The The

boost converter frequency lowest expected input voltage highest expected input voltage lowest desired output voltage highest desired output voltage output current maximum allowable voltage ripple

15,625 Hz 10 V 18 V 59.5 V 60.5 V 1.3 A 1V

Table 3 The output design parameters of the proposed boast converter. The minimum duty cycle The maximum duty cycle Minimum inductor size Peak inductor current Minimum capacitor Minimum schottky diode

69.97% 83.34% 93.09 mH 8.66 A 832 mF 60.05 V and 8.66 A Fig. 6. The I–V characteristic curves at fixed irradiance and different temperatures W at G ¼ 1000 m 2.

Table 4 The practical values used in the proposed boast converter. Duty cycle range Inductor size Peak inductor current Capacitor Schottky diode

70–84% 150 mH 10 A 1000 mF 100 V and 20 A

Table 5 The SOLARA S320P36 ULTRA module parameters. Parameter Peak power System voltage Voltage at peak power Current at peak power Open circuit voltage Short circuit current Operating temperature Number of cells

Value 80 W 12 V 17.9 4.48 A 21.8 V 4.82 A  40 to þ 85 °C 36

W Fig. 7. The P–V characteristic curves at fixed irradiance (G ¼ 1000 m 2 ) and different temperatures.

Fig. 5. The I–V characteristic curves at fixed temperature and different irradiances at T ¼25°.

2.4. System software The controller software program was written in ANSI C language using mikroC PRO for PIC microcontroller devices from Microchips. The program code size is 2.1 Kbytes. The software

program is based on the logic of the new MPPT tracking technique, the logic of the GA algorithm, and the constant current–constant voltage charging algorithm. In order to compare the results of the proposed controller and the conventional controllers, two versions of the software program were developed. One represents the proposed controller logic while the other carries out the traditional controller of perturb and observe technique. Each version is loaded and executed separately in order to compare the results. The flowchart of the software is described in Fig. 9. Firstly, all system parameters such as the output current and voltage of the PV array, the output current, and voltage of the battery, the ambient temperature of PV array and battery and irradiance are initialized. Then, the main program loop is started. Every time the sensors are scanned for updating system variables. After that, it checks for any changes occurred in irradiance and/or temperature. If there is a change, it searches the learning database for the nearest values to the optimum MPP values of current, voltage, and

Y.E. Abu Eldahab et al. / Renewable and Sustainable Energy Reviews 58 (2016) 646–655

Fig. 8. The P–V characteristic curves at fixed temperature (T ¼ 25°) and different irradiances.

651

Fig. 10. The PV current of MPPT versus the current of MPPT optimized by GA.

Fig. 11. The PV voltage of MPPT versus the voltage of MPPT optimized by GA.

Fig. 9. The software flowchart of the proposed controller.

power. Otherwise, it goes through the new MPPT tracking logic, and updates all system parameters for every scan cycle. It then checks the battery voltage and current to decide the stage of charging. If battery voltage does not reach the voltage set point, then it enables the constant current charging mode, else it changes over to the constant voltage charging mode. Then, it updates all system parameters accordingly and process is repeated for every scan cycle.

3. Experiment results The hardware setup described above is used in order to experimentally test the new controller, to work out its main performance parameters, and to compare it with the conventional

Fig. 12. The PV power of MPPT versus the power of MPPT optimized by GA.

MPPT controllers. Three modes of operation are planned and investigated. In each mode of operation, a specific version of the software program is loaded into the microcontroller. The first

652

Y.E. Abu Eldahab et al. / Renewable and Sustainable Energy Reviews 58 (2016) 646–655

mode, when the controller uses the conventional MPPT logic. The second mode, when it uses a GA based MPPT logic. The last mode, when it uses the new MPPT logic. To cover all practical operating conditions, three test cases are planned and executed. Test case I, where the temperature and irradiance are fixed, test case II, where the temperature and/or irradiance are changed, test case III, where the new MPPT technique is used. In addition, there is a separate test for the charging process.

system takes only three seconds searching the database to obtain the optimum power value. Then it goes forwards toward the MPP, W W and when irradiance changes from 1000 m 2 to 750m2 , it takes only two seconds searching the database to get the new optimum value of power. However, in case of using the traditional MPPT logic, the controller was not able to track the MPP.

3.1. Test case I, fixed temperature, and irradiance In this test, the temperature is fixed at 25 °C and the irradiance W is fixed at 1000 m 2 . There are two test results to be compared. The first, when the controller runs the traditional MPPT logic, and the other, when the controller executes the GA based MPPT logic. Figs. 10, 11, and 12 represent the output current, voltage, and power curves of the PV with time respectively together with the ideal response values indicated in orange. They show that in the case of conventional MPPT, the response steeply goes to the maximum power point. For every step, the duty cycle changes by 1.0%. It takes about 200 s to reach the MPP. However, it takes only few seconds to do the same in case of MPPT with GA. It takes 3 s searching the learning database to get the optimum value, which is the nearest point to the MPP. 3.2. Test case II, varying temperature, and/or irradiance with time W W In this test, the irradiance is varied from 1000 m 2 to 750 m2 . This test case looks like the event of partial shading, which may occur many times daily. The controller uses the GA based MPPT logic. W Fig. 13 demonstrates that, when the irradiance is 1000 m 2 , the

Fig. 14. The PV current curve of the new MPPT versus the traditional MPPT.

W W Fig. 13. The PV power curve when irradiance changes from 1000 m 2 to 750 m2 :.

Fig. 15. The PV voltage curve of the new MPPT versus the traditional MPPT.

Table 6 W W Sample records of the system log showing the case of MPPT with GA controller and irradiance varying from 1000 m 2 to 750 m2 . P old (W)

V old (V)

T (°C)

T bat (°C)

V bat (V)

Ibat (A)

V out (V)

Iout (A)

P out ðWÞ

W G(m 2)

P pv (W)

V pv (V)

Ipv (A)

Command flag

Cycle no.

46.96 51.598 60.135 63.536 79.245 79.293 42.955 45.943 59.128 59.149

8.86 10.8 12.7 13.4 17.6 17.7 12.1 13 17.7 17.6

26 26 26 26 26 26 26 27 27 26

26 26 26 26 26 26 26 27 27 26

5.6 5.6 5.6 5.6 5.7 5.6 5.7 5.7 5.7 5.7

1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2

60 59 60 59 60 60 60 59 60 59

0.82 0.97 1.01 1.28 1.26 0.68 0.73 0.95 0.94 0.95

49 57 61 76 76 41 44 56 56 56

1000 1000 1000 1000 1000 750 750 750 750 750

51.598 60.135 63.536 79.245 79.293 42.955 45.943 59.128 59.149 59.136

10.8 12.7 13.4 17.6 17.7 12.1 13 17.7 17.6 17.6

4.76 4.75 4.75 4.5 4.49 3.55 3.55 3.35 3.36 3.36

1 1 1 0 0 1 1 0 0 0

145 146 147 148 149 150 151 152 153 154

Y.E. Abu Eldahab et al. / Renewable and Sustainable Energy Reviews 58 (2016) 646–655

Fig. 16. The PV power curve of the new MPPT versus the traditional MPPT.

Fig. 17. The PV efficiency curve of the new MPPT versus the traditional MPPT.

Fig. 18. The first 10 s of the PV current curve of the new MPPT versus the traditional MPPT.

653

Fig. 19. The first 10 s of the PV voltage curve of the new MPPT versus the traditional MPPT.

Fig. 20. The first 10 s of the PV power curve of the new MPPT versus the traditional MPPT.

Fig. 21. The three stage constant current constant voltage charging process.

654

Y.E. Abu Eldahab et al. / Renewable and Sustainable Energy Reviews 58 (2016) 646–655

Table 7 Sample of the system logic showing the battery charging process transition from constant current to constant voltage. P old (W)

V old (V)

T (°C)

T bat (°C)

V bat (V)

I bat (A)

V out (V)

I out (A)

P out ðWÞ

W G (m 2)

P pv (W)

V pv (V)

I pv (A)

Command flag

Cycle no.

79.293 79.245 79.293 79.245 79.293 79.245 79.293 79.245 79.293 79.245 79.293 79.245

17.7 17.6 17.7 17.6 17.7 17.6 17.7 17.6 17.7 17.6 17.7 17.6

26 27 26 26 26 26 27 27 26 26 26 27

26 27 26 26 26 26 27 27 26 26 26 27

7.2 7.2 7.2 7.2 7.2 7.2 7.2 7.2 7.2 7.2 7.2 7.2

1.2 1.2 1.2 1.2 1.2 1.1 1.1 1.1 1.1 1 1.1 1

60 60 60 60 60 59 59 60 59 60 59 60

1.25 1.26 1.25 1.26 1.25 1.28 1.28 1.26 1.28 1.26 1.28 1.26

75 76 75 76 75 76 76 76 76 76 76 76

1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000

79.245 79.293 79.245 79.293 79.245 79.293 79.245 79.293 79.245 79.293 79.245 79.293

17.6 17.7 17.6 17.7 17.6 17.7 17.6 17.7 17.6 17.7 17.6 17.7

4.5 4.49 4.5 4.49 4.5 4.49 4.5 4.49 4.5 4.49 4.5 4.49

0 0 0 0 0 0 0 0 0 0 0 0

211 212 213 214 215 216 217 218 219 220 221 222

Table 8 Sample of system logic showing the third stage is finished and the battery voltage is fixed at 6.8 V, which is the battery nominal voltage. P old (W)

V old (V)

T (°C)

T bat (°C)

V bat (V)

I bat (A)

V out (V)

I out (A)

P out ðWÞ

W G (m 2)

P pv (W)

V pv (V)

I pv (A)

Flag

Cycle no.

48.90 69.87 77.73 79.24 78.92 79.25 79.29

20.3 19.30 18.42 17.53 18.06 17.81 17.68

25.00 25.00 25.00 25.00 25.00 25.00 25.00

25.00 25.00 25.00 25.00 25.00 25.00 25.00

6.70 6.70 6.70 6.80 6.80 6.80 6.70

0.01 0.01 0.01 0.00 0.01 0.01 0.01

59.00 59.00 59.00 59.00 60.00 60.00 59.00

1.12 1.25 1.28 1.27 1.25 1.26 1.28

66.37 73.85 75.27 74.98 75.29 75.33 75.33

1000.00 1000.00 1000.00 1000.00 1000.00 1000.00 1000.00

69.87 77.73 79.24 78.92 79.25 79.29 79.29

19.30 18.42 17.53 18.06 17.81 17.68 17.68

3.62 4.22 4.52 4.37 4.45 4.49 4.49

0 0 0 0 0 0 0

1 2 3 4 5 6 7

Table 6, lists some records of the system monitoring which is one of the good features of the proposed system. The second column from the right is a flag. It represents any changes occurred in temperature and/or irradiance. The flag is equal to one, it means that, a change is detected, and searching process for the optimum value begins. Once the searching is finished and optimum value is obtained, the flag returns to zero. It is clear that, when irradiance W becomes 1000 m 2 , the command flag has value of one for three iterations or three seconds, then it is changed to zero when searching is done and optimum value is obtained. Also, when the W W irradiance changed from 1000 m 2 to 750 m2 , the flag value changes to one for two iterations or two seconds, then it is changed to zero when searching is finished. 3.3. Test case III, the new MPPT test This test case compares the controller with the new MPPT logic and that with the traditional MPPT. Figs. 14,15,16 and 17 represent the dynamic response curves of current, voltage, power, and efficiency respectively. The figures show that, the controller with the new MPPT reaches the MPP in 4 s or iterations, while it reaches the MPP after 200 s with the traditional MPPT logic. They also show the new method has very limited steady state error and zero oscillation around the target MPP. Figs. 18,19,and 20 show the hidden details of the first 10 s in the PV output current, voltage and power curves respectively. It is clear that, the controller with the new MPPT logic reaches the MPP value in the fourth iteration or it takes only 4 s, while it needs more than 200 s with the traditional MPPT.

Table 9 Comparison between the PV datasheet parameters and the practical values of the experiments. Parameter

S320P36 PV module

Peak power(A) 80 W Voltage at peak 17.9 power(V) Current at peak 4.48 A power(A)

Experiment results MPPT based controller

GA based MPPT controller

New MPPT controller

79.293 W 17.66 V

79.293 W 17.66 V

79.3845 17.7

4.49 A

4.49 A

4.485

7.2 V, while the charging current is dropping gradually. After the absorption time passes, the float stage starts and continues until the charging current reaches its minimum value of 0.034 A. While the voltage drops to the nominal voltage of the battery of 6.8 V. It is clear from the charging curve that, the first stage takes one hour, 1.5 h for the second, and 0.5 h for the last stage. The fifth and sixth columns from the left in Table 7 shows the transition from constant current in the bulk stage to constant voltage in the absorption stage. While the fifth column from the left in Table 8 shows the case when the battery becomes fully charged. It is clear that, the battery voltage reaches its nominal voltage of 6.8 V and the charging current is zero because the battery is disconnected from the charging path. 3.5. Validation

3.4. Charging process test The test result of the charging process is depicted in Fig. 21. It shows that, the bulk stage starts and continues with constant charging current of 1.15 A, while the battery voltage incrementing steadily until it reaches the bulk voltage level at 7.2 V. Then absorption stage begins and continues with a constant voltage of

The practical experiments presented above are the best real way to validate any design or simulation work. However, in order to validate and verify the experimental results, they are compared with the PV module datasheet parameters. Table 9 compares the experimental obtained results to the datasheet parameters of the S320P36 PV module. Comparison includes all modes of operation

Y.E. Abu Eldahab et al. / Renewable and Sustainable Energy Reviews 58 (2016) 646–655

of the proposed controller, which are the MPPT based controller, the GA based MPPT controller and the New MPPT based controller. The comparison shows that the obtained values of the experiments are very close to the datasheet parameters.

4. Conclusion This paper presents a review of the most recent publications in the designing and implementation of solar battery charger controllers. It stresses the recent contributions features, drawbacks, and limitation. In addition, the review is augmented by presenting the design and implementation of a new controller, which preserves the good features and avoids the drawbacks of the conventional controller. The new controller is based on a new MPPT technique, which is optimized using genetic neural algorithm. Moreover, it uses the constant current constant voltage approach as a charging algorithm. All these modifications have been added in order to increase the speed of tracking process to the maximum power point and to reduce the charging period appreciably. One of the value added to the new controller is the feature of remote monitoring and administration of operation parameters. This feature enables monitoring operation parameters such as PV current, voltage, and temperature, and battery temperature, voltage, charging current, and state of charge. In addition, it allows sending commands to the controller in order to control the operation remotely like disconnecting or reconnecting the battery bank, and isolating the controller from PV modules. The experimental results show that, the new controller has a significant improvement compared to the traditional controllers in tracking speed and charging time. It tracks the MPP faster than the conventional controller. In addition, the charging period is reduced thanks to using a smart multi-stage-charging algorithm. In the near future, this controller will be used with higher system power and with different battery types to further evaluate its practical performance in the field. The implementation results of the new controller prove that, the new controller keeps on the features and avoids the drawbacks of the traditional controllers. In addition, it utilizes a new technique of MPPT for fast and accurate tracking, minimizing the steady state errors, and limiting the oscillation around the target MPP. Consequently, it is more efficient than the conventional charging regulator.

References [1] Haq S Shamshul, Wilson Shyam sunder B, Zameer G Mohammad. Design and simulation of MPPT algorithm of photovoltaic system using intelligent controller. Int J Adv Sci Tech Res 2013;3(6):2249–9954. [2] James P. Dunlop batteries and charge control in stand-alone photovoltaic systems. Fundamentals and Application, the Florida Solar Energy Center for Sandia National Laboratories; 1997. [3] Yan Jingyu, Xu Guoqing, Qian Huihuan, Xu Yangsheng, Song Zhibin. Model predictive control-based fast charging for vehicular batteries. Energies 2011;4:1178–96. http://dx.doi.org/10.3390/en408117811. [4] Abu Eldahab Yasser E, Saad Naggar H, Zekry Abdalhalim. Enhancing the maximum power point tracking techniques for photovoltaic systems. Renew Sustain Energy Rev 2014;40:505–14. [5] García-Villalobos J, Zamora I, San Martín JI, Asensio FJ, Aperribay V. Plug-in electric vehicles in electric distribution networks: a review of smart charging approaches. Renew Sustain Energy Rev 2014;38:717–31. [6] Spanos Constantine, Turney Damon E, Fthenakis Vasilis. Life-cycle analysis of flow-assisted nickel zinc-, manganese dioxide-, and valve-regulated lead-acid batteries designed for demand-charge reduction. Renew Sustain Energy Rev 2015;43:478–94.

655

[7] Han Xiaojuan, Ji Tianming, Zhaoand Zekun, Zhang Hao. Economic evaluation of batteries planning in energy storage power stations for load shifting. Renew Energy 2015;78:643–7. [8] Hussein Ala A, Fardoun Abbas A. Design considerations and performance evaluation of outdoor PV battery chargers. Renew Energy 2015;82:85–91. [9] Crossland Andrew F, Anuta Oghenetejiri H, Wade Neal S. A socio-technical approach to increasing the battery lifetime of off-grid photovoltaic systems applied to a case study in Rwanda. Renew Energy 2015;83:30–40. [10] Diouf Boucar, Pode Ramchandra. Potential of lithium-ion batteries in renewable energy. Renew Energy 2015;76:375–80. [11] Monem Mohamed Abdel, Trad Khiem, Omar Noshin, Hegazy Omar, Mantels Bart, Mulder Grietus, Van den Bossche Peter, Van Mierlo Joeri. Lithium-ion batteries: evaluation study of different charging methodologies based on aging process. Appl Energy 2015;152:143–55. [12] Tsang KM, Chan WL. State of health detection for Lithium ion batteries in photovoltaic system. Energy Convers Manag 2013;65:7–12. [13] Goli P, Shireen W. PV powered smart charging station for PHEVs. Renew Energy 2014;66:280–7. [14] Fendri Dalia, Chaabene Maher. Dynamic model to follow the state of charge of a lead-acid battery connected to photovoltaic panel. Energy Convers Manag 2012;64:587–93. [15] Bendib B, Krim F, Belmili H, Almi MF, Boulouma S. Advanced Fuzzy MPPT controller for a stand-alone PV system. Energy Procedia 2014;50:383–92. [16] Guenounou Ouahib, Dahhou Boutaib, Chabour Ferhat. Adaptive fuzzy controller based MPPT for photovoltaic systems. Energy Convers Manag 2014;78:843–50. [17] Ahmed Jubaer, Salam Zainal. A critical evaluation on maximum power point tracking methods for partial shading in PV systems. Renew Sustain Energy Rev 2015;47:933–53. [18] Fathy Ahmed. Reliable and efficient approach for mitigating the shading effect on photovoltaic module based on modified artificial bee colony algorithm. Renew Energy 2015;81:78–88. [19] Rizzo SantiAgatino, Scelba Giacomo. ANN based MPPT method for rapidly variable shading conditions. Appl Energy 2015;145:124–32. [20] Deshkar Shubhankar Niranjan, Dhale Sumedh Bhaskar, Mukherjee Jishnu Shekar, Babu T Sudhakar, Rajasekar N. Solar PV array reconfiguration under partial shading conditions for maximum power extraction using genetic algorithm. Renew Sustain Energy Rev 2015;43:102–10. [21] Dakkak Mohamed, Hasan Adel. A charge controller based on microcontroller in stand-alone photovoltaic systems. Energy Procedia 2012;19:87–90. [22] Nguyen Thanh-Tuan, Kim Hyung Won, Lee Geun Hong, Choi Woojin. Design and implementation of the low cost and fast solar charger with the rooftop PV array of the vehicle. Sol Energy 2013;96:83–95. [23] Sani F, Yahya HN, Momoh M, Saidu IG, Akpootu DO. Design and construction of microcontroller based charge controller for photovoltaic application. IOSR J Electr Electron Eng 2014;9(1):92–7 e-ISSN: 2278-1676, p-ISSN: 2320-3331. [24] Thounaojam Wallies, Ebenezer V, Balekundri Avinash. Design and development of microcontroller based solar charge controller. Int J Emerg Technol Advanced Eng 2014;4(5) ISSN 2250-2459, ISO 9001:2008 Certified Journal. [25] Ingole Jaya N, Choudhary Madhuri A, Kanphade Dr RD. PIC based solar charging controller for battery. Int J Eng Sci Technol 2012;4(02) ISSN 0975-5462. [26] Kulkarni Kaustubh S, Bhagare Sanket U, Shete Deepak P, Kulkarni Deepthi D. A review on maximum power point tracker (MPPT) for photovoltaic (PV) systems. Eur J Adv Eng Technol 2015;2(2):44–5. [27] Kuo J Jian-Long, Chao Kai-Lun, Lee Li-Shiang. Dual mechatronic MPPT controllers with PN and OPSO control algorithms for the rotatable solar panel in PHEV system. IEEE Transactions Ind Electron 2010;57(2):678–89. [28] Tesfahunegn SG, Ulleberg O, et al. A simplified battery charge controller for safety and increased utilization in standalone PV applications. IEEE; 2011. [29] Jiang Yuncong, Hassan Ahmed, Abdelkarem Emad, Orabi Mohamed. Load current based analog MPPT controller for PV solar systems. IEEE; 2012. [30] SreeManju B, Ramaprabha R, Mathur BL. Modelling and control of standalone solar photovoltaic charging system In: Proceedings of ICETECT; 2011. [31] Harish1 A, Prasad MVD. Microcontroller based photovoltaic MPPT charge controller. Int J Eng Trends Technol 2013;4(4):1018. http://www.ijettjournal. org. [32] AbduAllah Zaki Majeed, Mahmood Omar Talal, AL-Naib Ahmed MT Ibraheem. Photovoltaic battery charging system based on PIC16F877A microcontroller. Int J Eng AdvTechnol 2014;3(4):27–31 ISSN: 2249 8958. [33] Hiwale1 Anil S, Patil Mugdha V, Vinchurkar Hemangi. An efficient MPPT solar charge controller. Int J Adv Res Electr Electron Instrum Eng 2014;3(7):10505– 11. [34] Ingole Jaya N, Choudhary Madhuri A, Kanphade RD. PIC based solar charging controller for battery. Int J Eng Sci Technol 2012;4(02):384–90 ISSN 09755462. [35] Chen BY, Lai YS. New digital-controlled technique for battery charger with constant current and voltage control without current feedback. IEEE Trans Ind Electron 2012;59(3):1545–53. [36] Boost Converter Operation. LT1070 Design Manual, Carl Nelson & Jim Williams.