Computer Systems Science & Engineering DOI:10.32604/csse.2022.022927  
Article 
Modeling and Experimental Verification of Electric Vehicles OffGrid Photovoltaic Powered Charging Station
1Department of Electrical Engineering, College of Engineering, Taif University, Taif, 21944, Saudi Arabia
2Department of Electrical Engineering, Faculty of Engineering, University of Tabuk, Tabuk, 47913, Saudi Arabia
3Renewable Energy & Energy Efficiency Center (REEEC), University of Tabuk, Tabuk, 47913, Saudi Arabia
4Department of Electrical Power, Faculty of Engineering, Cairo University, Cairo, 12613, Egypt
*Corresponding Author: Essam Hendawi. Email: essam@tu.edu.sa
Received: 23 August 2021; Accepted: 25 October 2021
Abstract: With the increasing development of EVs, the energy demand from the conventional utility grid increases in proportion. On the other hand, photovoltaic (PV) energy sources can overcome several problems when charging EVs from the utility grid especially in remote areas. This paper presents an effective photovoltaic standalone charging station for EV applications. The proposed charging station incorporates PV array, a lithiumion battery representing the EV battery, and a leadacid battery representing the energy storage system (ESS). A bidirectional DCDC converter is employed for charging/discharging the ESS and a unidirectional DCDC converter is utilized for charging the EV battery. The proposed controllers achieve maximum power extraction from the PV and regulate the DClink voltage. It also controls the voltage and current levels of both the ESS and the EV during the charging/discharging process. The study has been applied to two cases with different power levels. Analysis, simulation, and implementation of the proposed system are presented. A 120 W laboratory prototype is carried out to verify the system performance, experimentally. Design guides for higher power levels are proposed to help in choosing the proper parameters of the converters. Both the simulation and experimental results are matched and verify the high performance of the proposed system.
Keywords: Electric vehicle; lithiumion battery; leadacid battery; photovoltaic; microcontroller; charging station
Recently, EVs are going to replace the classical internal combustion engine (ICE) vehicles [1,2]. This is based on the many advantages of EVs over ICEs such as higher energy efficiency, lower noise, and lower maintenance than the ICE. Moreover, EVs give many benefits to public health and are considered environmentallyfriendly vehicles. Nevertheless, the EV battery charging operation yet has many barriers such as the charging station’s infrastructure, the battery charging time, and the impact of these stations on the utility grid. The fastcharging approaches can be utilized to reduce the charging time to minutes [3–5]. However, these charging stations are connected to the utility grid that is affected adversely when a high number of stations are online. Many problems are generated such as excessive overload, voltage instability, and voltage variation [6–8]. One of the solutions to overcome these problems is promoting the utility grid however this requires extensive costs. Recently, energy storage systems (ESS) are integrated to the EV charging station [9–11] to reduce the utility demand performance. This prevents utility grid overloading while the EV is being charged.
If the EV charging stations are supplied mainly from the utility grid, the concept that the EV is an environmentfriend could not be accepted. Commonly, the utility grids are utilizing fossil fuels to generate the electricity. To ensure the better environmental impacts of the EV, renewable energy sources must be used in the charging stations. However, these sources are characterized by its intermittent nature. The ESSs can be used with these sources to store energy and charge EVs when needed.
The common renewable energy sources used for the EV charging stations are the PV and wind systems. However, the PV systems have many advantages such as the high efficiency and simplicity. Therefore, it is more appealing for EV charging stations. In literature, there are some research have been proposed for the EV charging stations utilizing PV energy [12,13]. Reference [14] has introduced a highenergy charging station using a twoquadrant converter supplied by PV energy. However, the introduced charging station could not be used for AC charging. An EV charging station utilizing PV array and a multiport converter has been suggested in [15]. But the harmonic distortion of the chargingstation current. A zsource converter is utilized to interface the PV energy to the grid and supply an EV charger has been introduced in [16]. The charging station gave better performance but it could not work in the offgrid mode. Reference [17] has introduced a PVpowered EV charging station that the ESS management has been optimized. An optimization of the EV chargingstations scheduling has been proposed by [18]. An EV charging station powered by PV array with an ESS has been presented by [19]. The ESS showed a good support for the system during high demands. The authors in [20] presented an analysis and evaluation of a gridconnected electric vehicle charging station (EVCS) powered by a PV solar system and a pack of batteries as a storage system. The authors also gave only a simulation study and did not introduce the details of controllers. They also focused on the simulation of the system. The authors in [21] presented also a theoretical study focusing on the power management and current scenario of a solarpowered EV charging station with ESS. Additional grid support is also considered to avoid power interruption in the charging station. A PV based offGrid charging station for EV is presented by the authors in [22]. However, the authors focused on the simulation study of the system without mathematical verifications. In addition, the PV panels operate for several periods without achieving maximum power point operation.
In this paper, an efficient EV charging station supplied from a PV panel and operates in the offgrid mode is proposed. The proposed system incorporates a PV array, a lithiumion battery representing the EV battery, and a leadacid battery representing the energy storage system (ESS). Also, a bidirectional DCDC converter is employed for charging/discharging the ESS and a unidirectional DCDC converter is utilized for charging the EV battery. The proposed control system includes three controllers. Each controller has two nested loops. One of the controllers is used to control the charging and discharging of a leadacid battery pack that represents the ESS. Another controller is utilized to regulate the charging process of the EV battery (lithiumion battery). The power extracted from the PV panels is always maximized to meet the required energy of the EV battery and store the excessive energy in the leadacid battery. Design guides for selecting the suitable parameters of the converters are proposed for the low power case of study and for any higher power levels of the charging station. Modeling, analysis, and simulation of the system are presented including two cases: low power (120 W) and high power (3 kW) charging systems. Simulation results in both cases of study and experimental results in a laboratory prototype are very close and verify the effectiveness of the proposed system. A singlechip PIC18F4550 microcontroller is utilized to realize the operation of controllers and generate the control signals of the power switches, experimentally. The system is tested successfully for 120 W and 3 kW power levels. The results show that the system controllers track their references perfectly. Also, the ESS compensates for the drop in the PV energy at low insolation states and keeps the EV battery charge continuously.
The paper structure is as follows. Section 2 presents the proposed system description. Modeling of the system has been described in Section 3. The system controllers are introduced in Section 4. Simulation results are presented in Section 5. Design guides for higher power levels are proposed in Section 6. Experimental verification has been proposed in Section 7 and finally, Section 8 comes with the conclusions.
The proposed system, shown in Fig. 1, is a standalone EVCS, which is powered by a PV panel. The main power source for the system is the PV unit that is normally connected to a DC power regulating capacitor [23]. A boost converter is cascaded to the PV output terminals. The function of the boost converter is to manage the maximum power point tracking (MPPT) condition of the PV panel and provide a voltage level matching between the PV and the system DC bus. The DC bus voltage is regulated through the controllers of charging/discharging the storage battery. The boost converter output terminals are the system DC bus that is linked to the EV charger input. The EV charger is a onequadrant buck converter. Also, the storage system, leadacid batteries connected through a twoquadrant DC/DC converter, is linked to the DC bus.
3 System Modelling and Analysis Layout
Generally, the first step for any research is the modeling of the system. The system model is an important foundation for simulation, implementation, and controllers’ design. The mathematical model assumes that the system has no losses, no voltage drops and leakage currents, and no snubber circuits. Nevertheless, the computer simulations do not use those assumptions. The following paragraphs will explain the model of each part of the proposed system.
The dynamic model of the PV panel is represented by the circuit shown in Fig. 2 [24,25]. The current source value (IS) represents the short circuit current of the PV panel. The resistors (rp, rs) are the model parallel and series resistances. Relations between the model parameters are wellknown in the literature [24,25].
The system converters circuit diagram is shown in Fig. 3. The analysis of the twoquadrant converter applied to the storage leadacid battery “B1” is presented so that the converter parameters can be designed.
The storage battery converter dynamic model in the statespace form is described as follows.
First, the buck operation is represented as:
where il, and vc are the inductor current and the capacitor voltage, respectively. u is a logic number {0:1} that implements the operation of S2 action. C1 and L1 are the filter capacitance and inductance, respectively. Vdc is the DC bus voltage. The matrices A, B, and C are defined respectively as:
Manipulating the previous equations gives:
where ‘
Second, the boost operation is represented as follows. S3 is alternated between ‘on’ and ‘off’ state for periods TonB1 and ToffB1 respectively while S2 is kept in the “off” state. During the ‘on’ period of the boost mode.
where u1 is a logic number {0:1} that implements the operation of S3 action.
To achieve a certain ripple factor “
“R” is the output resistance of the converter, and it can be calculated as:
The value of the converter inductor
To achieve the same ripple factor “
The same procedures are applied to the onequadrant converter of the EV battery to determine the parameters
The relation of the capacitor
The inductance of the boost converter on the PV side (L3) is chosen according to the following equation:
The proposed control system is divided into three controllers, shown in Fig. 4. The first controller is the MPPT controller. This controller is adapted to absorb maximum power from the PV array. It generates the required boost converter duty cycle that forces the PV into the MPPT conditions. The second controller is the EV converter controller. This controller regulates the EV charging process. The third controller is the battery storage, converter controller. This controller adopts the DC link voltage and the storage battery current.
The energy utilization of the PV can be improved using the MPPT controller. Nevertheless, the MPPT has been applied using many algorithms [26–29]. The common technique called “Perturb and Observe” is utilized here. The algorithm of this technique is briefly described as follows:
Step 1: Read the present values of PV voltage Vpvnew and current Ipvnew then calculate the present PV power Ppvnew
Step 2: Calculate the increment values of PV power ΔPpv and PV voltage ΔVpv.
Step 3: Update the duty cycle of the switch S1 and generate the corresponding control signal.
Step 4: Go to Step 1 to repeat the cycle of MPPT control.
4.2 The Charge Controller of the EV Converter
The function of this controller is to control the EV chargingprocess. The common control methods for the charging process are the constant current, constant voltage, and constant voltage/constant current. The last method is usually selected due to its low charging time high current protection. The constant voltage/constant current method is used in the proposed charging station. The controller possesses two nested loops for the current and voltage as shown in Fig. 4b. A simple PI controller is assigned for both loops. The PI controller gains parameters are tuned by Ziegler–Nichols technique.
4.3 The Charge Controller of the ESS Converter
This controller is used to regulate the voltage of the DClink and control the charge/discharge processes of the ESS battery. The utilized controller this converter is the conventional PI controller. If the PV energy is sufficient to charge the EV, the excess PV energy is sent to charge the ESS batteries. When the ESS batteries are fully charged, the ESS charging ends and the MPPT stop.
The proposed system shown in Fig. 1 is simulated using Matlab/Simulink platform. The simulation process is carried out for the following two cases:
5.1 Case (1): (3 kW) Power System
A relatively highpower system is simulated to verify the effectiveness of the proposed system. Four units of PV panels (125 W) are connected in series to get one PV string. Six strings are connected in parallel to get the PV array of 24 PV panels having a total maximum power of 3 kW. The storage battery bank is four, 12 V, 250 Ah leadacid batteries connected in series. The lithiumion battery representing the electric vehicle is a 48 V, 200 Ah. Sun irradiance is assigned three steps. The first period (0 to 5 sec) is 1000 W/m2 corresponding to the maximum power of the PV array. The second (5 to 10 sec) and third (10 to 15 sec) are 700 W/m2 and zero W/m2 corresponding to 70% Pmax and evening time, respectively. The DC bus voltage is set at 100 V. The behavior of the lithiumion battery is shown in Fig. 5. The S.O.C. is initially set at a low value of 30%. Therefore, the voltage controller generates a high reference charging current of 30 A. The current controller follows the reference current during the three periods and the battery is charged from the PV array from t = 0 to t = 10 s and is charged at the same voltage and current levels during the third period from t = 10 s to t = 15 s through the leadacid storage battery bank.
Fig. 6 presents the performance of the PV array and storage battery. The MPPT algorithm always tracks the maximum power during the first two periods while all PV array values are zero during the evening period. The performance of the leadacid storage battery bank (B1) is presented in Figs. 6f–6h. During the first period with maximum PV array power, the leadacid storage battery bank is charged rapidly with a high charging current and high state of charge (S.O.C) rate. During the second period, the rate of charging is reduced due to the reduction in solar irradiance.
After sunset, the PV array is disconnected. The battery current becomes positive, and the battery starts discharging. The battery voltage and S.O.C. starts decreasing. Fig. 6e gives the DC bus voltage and its reference. The voltage controller of the storage battery generates the required charging current of the storage battery during the first two periods and the required discharging current during the third period so that the DC bus voltage is always kept constant.
5.2 Case (2): (120 W) Power System
The system power and parameters are typically the same as the implemented experimental setup. In this case, both the simulation and experimental setup specifications are:
• PV panel (copex model: P120) specifications:
Peak power Pmax 120 W, Voltage at maximum power Vmp 16 V, current at maximum power Imp 7.5 A, opencircuit voltage Voc 19.2 V, and shortcircuit current Is.c. 8.82 A.
• Storage battery (Lead Acid) specifications:
Rating: 65 Ah, nominal voltage: 12 V, fully charged voltage: 13 V, cutoff voltage: 9 V, maximum charging current: 0.2 C = 13 A and maximum discharging current: 0.2 C = 13 A.
• Battery representing EV (Lithiumion):
Construction: four batteries connected in series, rating: 6.5 Ah, nominal voltage: 3.7 V per battery, fully charged voltage: 4.25 V per battery, cutoff voltage: 3 V per battery, maximum charging current = 3 A and maximum discharging current = 5 A. Four seriesconnected Liion batteries form a battery pack having a rating 6.5 Ah, it has 14.8 V rated voltage, the fully charged/cutoff voltage is 17 V/12 V respectively. The parameters of the converters for low power case are listed in Tab. 1. Figs. 7 and 8 present the responses of the system against to step variations in the PV insolation level. The variations of the PV irradiation are shown in Fig. 7a. The response of the voltage of the DC bus precisely follows its reference (25 V), as presented in Fig. 7b. The S.O.C. of the leadacid battery B2 is shown in Fig. 7c.
When the irradiation is greater than (60%), the PV generates enough energy to charge the EV and serve the rest in the ESS battery. This occurs at the first 4 seconds. If the irradiation is less than (60%), the generated PV energy is not sufficient to charge the EV. Therefore, the ESS battery starts discharging to account for the PV energy drop. It is seen from Fig. 7d that the ESSbattery current follows well its reference. Also, the charging and discharging processes track and compensate for the irradiation level. Fig. 7e shows the voltage of B1 battery where it increases with charging and decreases with discharging. The PV current is shown in Fig. 7f. The current level is the same as the MPPT conditions. The variations of the PV irradiation are shown in Fig. 8a. The charging current of the EV follows perfectly its reference value as shown in Fig. 8b. Fig. 8c shows the voltage of the B2 battery which is continuously charged as its S.O.C. can be illustrated in Fig. 8d. Figs. 8e and 8f present the PV power, the EV battery power, and the power of the B1 battery. When the irradiation is greater than (60%), the PV generates enough energy to charge the EV and serve the rest in the ESS battery. This occurs at the first 4 seconds. If the irradiation is less than (60%), the generated PV energy is not sufficient to charge the EV. It is noted that the EV charging is kept constant against all disturbances, as presented in Fig. 8f.
6 Design Guides for Different Power Levels
This section presents design guides for selecting the parameters of the converters assuming that the calculated parameters which are given in Tab. 1 are the base values based on
The following assumptions are considered:


 The storage battery voltage level
 The EV battery voltage level
The power level of the charging station
As the inductor current ripple is usually a percentage value of the average inductor current, then
Therefore Eqs. (3), (7), (9), (10), (12), (15), (16) and (18) are replaced in the higher power charging station by Eqs. (21)–(28) respectively noting that
“R” is the output resistance of the converter and it can be calculated as:
These design equations help for proper selection for inductances and capacitances of the converter keeping into consideration the maximum currents of inductors and maximum voltages of capacitors in high power applications.
An experimental prototype is built to verify the simulation results of the proposed system. The prototype has the same specifications as indicated in case 2. The practical system has the following converters:
• DCDC boost converter with MPPT on the PV panel side
• A bidirectional boostbuck converter on the storage leadacid battery (B1) side.
• A buck converter on the lithiumion battery (B2) side.
A singlechip PIC18F4550 microcontroller is utilized to achieve the operation of the controllers simultaneously. The microcontroller crystal oscillator frequency is set at 16 MHz. The microcontroller receives the signals of Ipv, Vpv, IB1, VB1 Vdc, IB2 and VB2. Then the microcontroller utilizes its modules Timer0, Timer1, Timer2 and CCP to perform the controllers of the system. The outputs of the microcontroller are the signals of switches S1, S2, S3 and S4. The five controllers are:
• DCDC boost converter controller
The microcontroller receives the PV voltage and current, performs the perturb and observe MPPT algorithm, then generates the control signal to S1.
• DC bus voltage controller
This controller regulates the DCbus voltage to track a certain constant value. The stability of the system is directly related to the regulation grade of this controller. The output of this controller is the reference current of B2.
• Leadacid battery current controller
This controller utilizes the reference battery current and generates the control signal to either switch S2 (charging) or to switch S3 (discharging).
• Lithiumion voltage controller and current controller
The lithiumion battery is charged using constant voltage/current method. The outer loop is set for the voltage controller that regulates the battery voltage and produces the reference current for the inner loop. The inner loop (current controller) generates the control signal to S4.
The experimental platform has been implemented in the laboratory to verify the feasibility of the proposed system as shown in Fig. 9. The current controller of the lithiumion battery is tested at different current levels (3A, 2A, and 1A) as illustrated in Fig. 10. The voltage controller forces the battery voltage to change according to each current value. Fig. 11 shows the effects of disconnecting the PV panel. After disconnecting the PV array, the DC link voltage is kept constant by the controller. However, the leadacid battery discharges and its voltage drops. The controller of the buck converter of the lithiumion battery keeps its voltage level and current at the same values before disconnecting the PV panel.
Disconnecting and reconnecting the PV panel are illustrated in Figs. 12a and 12b. In Fig. 12a, if the PV array is removed, the battery B2 charges the EV battery B1. The B1 voltage and the DClink voltage are kept at the set values. However, the leadacid battery voltage decreases. Fig. 12b shows the reconnection of the PV panel. Both batteries are being charged, and the leadacid battery voltage increases. Figs. 13a and 13b show the control pulses of the converter switches in connecting and disconnecting the PV array. The previous results are considered as shorttime results where they can be monitored by oscilloscope. The longtime results are that associated with the charging process. The software in the laboratory from (LabVolt) can store the results at minimum intervals of one second. Fig. 14a presents part of the charging process of EV battery while Fig. 14b gives the performance of the MPPT with a step change in the load at t = 70 s.
An efficient EV charging station based on a PV energy source has been proposed. Analysis and simulation of the system were presented including two cases: low power (120 W) and high power (3 kW) charging system. The PI voltage and current controllers were employed to control the charging and discharging of a leadacid battery pack that represents the energy storage system. Another voltage and current controller were utilized to adjust the charging of an EV battery (lithiumion battery). The power of the PV panel was always maximized to meet the required energy of the EV battery and store the excessive energy in the leadacid battery. Simulation results in both cases of study and experimental results were very close and verify the effectiveness of the proposed system.
A singlechip PIC18F4550 microcontroller is utilized to realize the operation of controllers and generate the control signals of the power switches thus lowering the hardware implementation complexity and cost. The system was tested successfully for 120 W and 3 kW power levels, and it could be extended for higher power charging stations where the proposed design guides can successfully help for selecting the suitable parameters of the converters. Also, it could be applied in urban and remote areas as well.
Funding Statement: This research was funded by the Deanship of Scientific Research, Taif University, KSA (Research project number 144199).
Conflicts of Interest: The authors declare that they have no conflicts of interest to report regarding the present study.
This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 