Over the last few years, the use of renewable power generation techniques has expanded drastically, thus it is critical to develop a mechanism in order to ease the implementation of Renewable Energy Resources (RER) so that the overall efficiency, safety and dependability of the grid is enhanced. As there is a drastic increase in electricity demand all over the world which makes it crucial to use new sources of power generation which include solar, wind fuel cells etc. [1]. These RERs have been found very effective in order to meet the rising demand of energy while also addressing some major environmental issues. Out of the all the Renewable power generation sources, Solar PV is the most common and effective one because it is cost effective, highly efficient with low maintenance charges [2].
In a typical solar power generation system, solar PV panels are installed serially or parallelly to maximize power generation. The power generated by the solar panel directly depends on the intensity of sunlight which means if intensity of sunlight is more, more power is generated and vice versa [3]. Solar panels generate DC electrical power. To make this solar energy usable, it must first be converted fromDirect Current to Alternating Current with the help of an inverter. The AC electrical energy thus generated can be used to operate local electronics or sent to the electrical grid to be used elsewhere. The major drawback of using solar power generation systems is its dependency of atmospheric factors such as solar irradiance and temperature. This leads to inefficiency as the solar panels are unable to extract maximum power [4]. Owing to their dynamic design, solar power faces difficulties in interacting with automotive systems too. As a result, it is important to have a battery in EVs in order to address the issue.Different MPPT (Maximum Power Point Tracking) techniques are used, toimprove system efficiency and to obtain maximum power from the panels, Some commonly used techniques are: Fuzzy logic, Perturb and Observe (hill climbing method), Neural Network, Fractional open circuit voltage, Incremental Conductance method, Fractional short circuit current.To extract maximum power from the solar PV panels, many techniques have been proposed by several researchers in this field [5][6][7][8][9][10][11][12].
This paper proposes an electric vehicle battery charger using PV array with FOPID(fractional order PID)controller.Experiments were performed in MATLAB platform and a comparative performance analysis with the results obtained by using PI and PID controller is done to illustrate the effectiveness of the proposed method. It uses controller having two stages using a FOPID witha MPPT algorithm to extract maximum power from the PV system. The FOPID controller is used to vary the duty cycle of boost converter.
Fractional-order calculus is amathematical tool fordealing with derivatives and integrals from noninteger orders. In recent times lots of research have been done both by the academicians andindustrialists dealing with Fractional-order proportional-integral-derivative (FOPID) controllers. Fractional system provides a better understanding of system characteristic like system response, rejection of disturbance , better and improved capability of handling model uncertainties in nonlinear as well as real time applications. The 's' domain representation of PID controller is:
Where 'C(s)' represents the output of the system, 'E(s)'is error and K p ,K d and K i represent theproportional, derivate and integral parameters of the control system. And the 's' domain representation of FOPID controller is:
Thus, in case of FOPID the power of 's' is fraction compared to PID controller where it is integer. The aim thus is to optimize the value of the two additional parameters 'μ' and 'λ' in addition to the three K p ,K d and K i parameters. Thus, they have additional flexibility in the controller design, compared to the standard PID controller, because they have five degrees-of-freedom (DOF), compared to three DOFs of its integer-order counterpart. Higher DOF provides better time and frequency responses of the control system.
The proposed model consists of solar PV panels, sepic dc-dc boost converter, MPPT charging with FOPID controller, alternate battery bank. The proposed model works in following three modes, Thesepic converter is used to regulate the voltage and current supply. The MPPT technique is thus used to obtain maximum power from solar panels by usingthe FOPID controller. Mode 1: When the sunlight is at peak, the sunlight falls on the solar panel which converts it into the electrical energy that is capable of charging the battery bank as well as the battery of electric vehicle.
In the next mode, when the input irradiance of sun is very low in such case, the supply of solar panel is cut off and the battery bank is used as a charging source for EV. The battery bank starts getting discharged slowly. Mode 3: In the 3 rd mode of operation, the irradiance is increased so that EV is charged by the solar panel but the battery bank is not charged in this case. Another major advantage of using the Sepic converter is that it can operate in boost and buck modes depending on the duty ratio cycle. (6) Where 𝑉𝑉 𝑃𝑃𝑉𝑉 𝑚𝑚𝑚𝑚𝑚𝑚 represents the minimum voltage generated by PV panels, Δi PV represents the input current ripple, f SW represent the switching frequency, I dc represents the dc link current, ΔV C1 represents the voltage ripple of capacitor C1, ΔV dc represents the output voltage ripple, and D max represents the maximum duty ratio which can be calculated by the equation 7.
𝑉𝑉 𝑑𝑑𝑑𝑑 +𝑉𝑉 𝐷𝐷 𝑉𝑉 𝑃𝑃𝑉𝑉 𝑚𝑚𝑚𝑚𝑚𝑚 + 𝑉𝑉 𝑑𝑑𝑑𝑑 +𝑉𝑉 𝐷𝐷 (7) where V D represents the voltage drop in diode.
Bidirectional dc-dc converter
The process opted by the proposed FOPID system to charge vehicles by using solar PV panels are explained briefly here;
1. Initially, when the sunlight falls on the PV panels that are converted to the electrical energy in order to charge the batteries of EVs. As the proposed model is working in three modes, the first mode is when battery bank and EV battery are getting charged by the solar PV panel. A number of parameters are defined such as input solar irradiance, temperature, open circuit voltage etc. Other than this there are some other important parameters which are given in table1. 2. Once the voltage is generated, DC-DC converter requires a duty cycle to perform effective operations. For this a MPPT technique is designed which would assist the converter to produce duty cycle, whichin this case is generated using the FOPID converter (taking λ= 0.0675 and μ= 0.5). along with their values [8].
Table2.Table 3.
The third mode of the proposed system is when alternate charging source i.e., battery bank is turned off and battery of EV is getting charged by the solar PV panels. Table 3. below represents the different EV parameters along with their configurational values. From the graph, it is observed that initially the battery is getting charged by the solar PV panel in mode 1. But as soon as the next mode starts the battery starts to discharge slowly up to 1.3 sec. After 1.3 sec, the third mode starts in which the battery is neither getting charged nor discharged i.e., it remains constant. 6. Likewise, the SOC of EVs battery is obtained along with its current and voltage waveform which are shown in figure 9 below. Figure represents the state of charge of the EV batteries. From the graph, it is analyzed that the EV is getting charged in all the three modes. In the first mode, EV battery is getting charged by the Solar PV panel and battery bank, in the second mode, it's getting charged by the battery bank and in the third mode, the battery of EV is getting charged by the PV panel.
The performance of the proposed FOPID model is analyzed and compared with the conventional PI and PID models in terms of the voltage generated by solar PV panels. From the graph, it is observed that the battery bank SOC in the proposed model is higher than the traditional two approaches which lasts more than 70.00155%, thus making it long lasting and efficient .
Lastly, the performance of the proposed model is also determined in terms of the state of charge of EV battery. Figure 11. demonstrates the comparison graph of the proposed FOPID model and traditional PI and PID models in terms of the EV battery SOC. The performance of the traditional PI and PID model is represented by the blue-and orange-colored lines and the performance of the proposed FOPID model is represented by the yellow-colored line. From the graph, it is observed that in all the three modes of operation the EV is getting charged constantly. However, after analyzing the graph closely it is observed that the EV battery of the proposed FOPID model lasts longer than the conventional two approaches.
From the graphs, it is observed that the proposed FOPID model outperforms the classical PI and PID model in all the factors and is more effective and efficient in charging EV through solar panels. The simulation outcomes are obtained in terms of the voltage and current generated. In case of the proposed FOPID model, the voltage and current doesn't fluctuate much and can effectively charge and battery bank of EVs. Similarly, the performance of the proposed FOPID model is compared with the conventional PI and PID models in terms of the voltage and current generated by the sepic dc-dc converter. The voltage generated by the sepic dc-dc converter remains constant in all three operating modes which ultimately provides sufficient amount of current to the EV for charging. Lastly, the charging state of the battery bank and EV battery is determined in which the proposed model outperforms the classical PI and PID models. All these factors, make the proposed FOPID model more efficient and effective for charging the EV.
FOPID has more number of parameters to be tuned, thus providing better and finer tuning than PID controllers to meet the system target. In spite of the technical advantages offered by FOPID over its integer-order counterparts the adoption of the fractional order controller in the industry is slow. The cost of producing such controllers, the cost-benefit to the end user and the complexity of implementation of FOPID controllers with respect to its extra tuning flexibility verses the extent of performance improvement are certain factors which needs to be investigated to make it readily acceptable to the industry.