Optimizing wireless charging for electric vehicles via a novel coil design and artificial intelligence techniques

1 Introduction

The rapid growth and supremacy of electric vehicles (EVs) has seen a substantial surge over the last five years. The simplicity of usage and socioeconomic advantages received are identified as key factors contributing to the significant demand. Developed countries have recognized the need for EVs and the forthcoming industrial revolution. In contrast, developing countries are transitioning their transportation industry from traditional internal combustion (IC) engines to EVs in alignment with the sustainable development goals of the United Countries. Recent developments in the field of EVs have been pushed forward by carbon emission reduction mandates because of the benefits of EVs over vehicles fueled by internal combustion engines. EVs obtain their energy from enormous rechargeable batteries, necessitating the availability of charging stations. One of the major challenges to the widespread adoption of EVs is the extended battery charging time. This issue leads to dissatisfaction among vehicle owners, who may face challenges in completing their tasks due to battery depletion. Consequently, this reduces the tendency towards purchasing EVs and may lead consumers to revert to internal combustion engine vehicles, which contribute significantly to CO₂ emissions. Therefore, the implementation of a comprehensive charging station infrastructure, equipped with specialized equipment assumes significant importance. There are currently two rapidly developing technologies that employ powerful electric motors as their primary propulsion sources: plug-in hybrid electric cars (PHEVs) and battery electric vehicles (BEVs). Energy storage in massive battery packs of EVs is limited [1]. Despite numerous available sources in the literature on EVs, the research remains a significant trend in scientific discourse due to its necessary role in converting environmentally harmful energies into clean, renewable ones. However, there is a pressing need for further research to explore various aspects comprehensively. Typically, they must be recharged by getting to the electrical grid at home or in public places. A charger must convert the power supplied by an external source to match the specific electrical demands of the EV battery. Furthermore, to enhance the efficiency of power transmission between the linked coils, it is necessary to raise the operating frequency for wireless power transfer applications. Power converters effectively adapt voltage levels, frequencies, currents, and impedances with decreased losses by converting between different voltages and currents. In a power converter, the semiconductors are arranged in a way that allows them to transition between two states, open and closed at a certain frequency [2,3]. Fast charging aims to shorten the charging period of EV batteries, but it causes several battery degradation mechanisms like energy loss, temperature rise, life cycle, and reduction. Therefore, there is significant demand to investigate and alleviate the adverse battery degradation mechanisms to develop safe and efficient chargers for durable batteries and to meet end-user expectations. Fast charging with the CC–CV charge control strategy is the best option for lithium-ion batteries because these batteries can be damaged if the voltage becomes too high [4]. The CC–CV technique is to charge the battery with a constant rated charging current till the voltage reaches the cut-off value. Beyond that, the voltage is maintained at a constant level while the current gradually decreases to its minimum value, as explained in Figure 1 [5].

Figure 1

Schematic diagram of the CC–CV [5].

The implementation of constant voltage (CV) and constant current (CC) control for a wireless charging system is proposed. The CV mode is operational when the current is under the set current value, while the CC mode is active otherwise. These control modes are necessary for a battery charging system to protect the battery and maximize system efficiency. To regulate the output voltage and current of the charger, the wireless charging system does not use a post-regulator step. However, an ordinary PI control is combined with a variable frequency control. The proposed method is a viable option for implementing the wireless charging controller. Furthermore, under full load conditions and with an air gap of 200 mm, the direct current (DC)–DC stage’s hardware efficiency may reach 97.9% [6].

A wireless charging model involving two stages of power electronic converters is proposed [7]. The popular configuration in AC–DC converters is the neutral boost power factor correction (PFC) in Vienna rectifiers, which extracts minimal harmonic content and reactive power from the electrical grid while maintaining a constant DC bus voltage. The phase-shifted full bridge (PSFB) converter is one of the DC–DC converter topologies used in EV charging. PSFB converters offer advantages for EV charging, such as high efficiency, low electromagnetic interference (EMI), and high power density. The simulations show that DC fast chargers are most inefficient in low voltage–high current or high voltage–low current situations. Due to energy passing through two power converters, DC fast charger efficiency decreases significantly. This problem is solved by electric vehicle charge-responsibility sharing. The suggested research shows that DC fast charging systems are more efficient at higher output voltages [8].

The selection of the proposed method is based on minimizing the charging time, the harmful spikes in charging current, and efficiency enhancement and reflecting researchers’ focus on advancing efficiency and sustainability in energy conversion. This underscores the importance of continued research in this critical field. This study endeavors to introduce an innovative design for electric vehicle charging, offering both regular and rapid charging capabilities. Through the addition of multi-coils, these coils are designed to be closed together to ensure a minimal air gap between them, allowing for higher power transfer with lower losses. To enhance the performance of the fast wireless charging system for EVs, we propose the addition of multi-coils to the main receiving coil, controlled by an artificial intelligence circuit. These coils are engaged as needed based on the actual condition of the battery. The charging time is significantly reduced, which is particularly beneficial during urgent situations. This advancement aims to address complaints from EV owners regarding charging durations and alleviate congestion at standard charging stations.

2 Proposed schematic of EV wireless charging

Different types of converter concepts can be used in charging battery systems. An AC/DC rectifier, DC/DC converter, resonant transformer, and AC/DC rectifier must create the DC voltage needed by the battery pack in an EV’s charging system [9]. Figure 2 shows the block diagram of the wireless charging of EVs.

Figure 2

The block diagram of wireless charging of EVs.

2.1 Three-phase controlled rectifier modeling

The wireless charging system model uses a voltage source current-controlled PWM rectifier. The control is accomplished by detecting the instantaneous phase currents and requiring them to adhere to a sinusoidal current reference template, I__ref. The amplitude of the current reference template, I _MAX, is calculated as follows [10]:

I MAX = G C · e = G C · ( V REF − V D )

where G C is defined as a PI controller. As illustrated in Figure 3, the sinusoidal waveform of the template is produced by multiplying I MAX by a sine function, the required phase-shift angle φ, and the same frequency as the mains [11]. Furthermore, there is a need to synchronize the template with the power supply. Following that, the template was made and prepared to generate the PWM pattern [12,13]. One issue with the rectifier is that the feedback control loop on the voltage V _C can produce instability. Subsequently, this issue needs to be examined when designing the rectifier [14].

Figure 3

Voltage source current-controlled PWM rectifier [14].

2.2 DC–DC converter

The DC signal is converted using a DC–DC converter. The main function of a DC–DC converter in the charging system is to charge the battery of EVs in a broad voltage range and to be produced in various voltage levels; EV batteries also vary depending on the state of charge (SoC). The power converter and its semiconductor components need to be switched at a high frequency, in the range of 30 kHz, to accommodate the magnetic resonance charger operations. Higher frequencies can be switched by conventional silicon MOSFETs. Figure 4 shows the basic circuit of the DC–DC converter. Note that the converter’s input is fed by switching pulses produced by the converter’s pulse generator. Thus, the converter creates a square wave that is fed into the transmitting end of the coil [15,16].

Figure 4

DC–DC converter [17].

2.3 Input voltage of resonant tank

The output of the converter on the transmitter side feeds the resonant tank input. In high-power WCS for EV applications, the most popular DC–DC converter architecture is shown in Figure 5. The DC power is provided to the converter input, either directly or via a low-frequency rectifier from the AC grid.

Figure 5

Fundamental design of the DC–DC converter [17].

3 Proposed wireless charging of EV model

This study investigates wireless fast charging of EVs to reduce charging time and congestion at charging stations. We proposed a model for a fast-charging system based on Level 3 charging standards according to the SAE J1772 standard for electric and hybrid vehicles [18]. Figures 6 and 7 show the input/output voltage of the controlled rectifier and the output voltage of the DC–DC converter. Advanced smart controllers such as adaptive neuro-fuzzy inference system (ANFIS) were utilized to evaluate their impact on improving charging system performance using a constant current–constant voltage (CC–CV) method. Using MATLAB/Simulink, the simulation for the EV charging system was modeled, as shown in Figure 8.

Figure 6

Input/output voltage waveforms of three phase-controlled rectifiers.

Figure 7

The output voltage of the DC–DC converter, with T s = 1 / 30 kHz .

Figure 8

EV wireless fast charging system modeled in MATLAB/Simulink.

4 Modeling of wireless charging system with adaptive neuro-fuzzy control circuit

4.1 Close-loop battery charging system

The phase (A) voltage and current waveform-controlled AC–DC converter simulation are shown in Figure 9. The corresponding phase voltages and currents are clearly in perfect sync. Based on the fast Fourier transform (FFT) shown in Figure 10, the phase current’s harmonic content is about 3.4%. The PFC satisfies the criteria since it is less than the 5% limit set out by IEEE Std 519-2014.

Figure 9

The phase voltage and input phase A current waveforms of the controlled AC–DC converter.

Figure 10

The harmonic content of the phase current.

The current is controlled at a fixed level during the CC charging process until the voltage of the battery cells reaches an established level. Following that, the charging method is changed to CV charging, which charges the battery by applying a current through the charger’s constant voltage output [19]. The ANFIS controller is integrated with a PWM DC–DC converter, which is responsible for the charging process [20]. This means that the purpose of this controller is to regulate and maintain a constant current value of 100 A during the battery charging process, specifically up to 80% of the charge. In this scenario, the controller maintains the voltage level at a stable value during the charging period Figure 11 shows the ANFIS controller circuit in an overall wireless fast charging system.

Figure 11

Block diagram of MATLAB simulink of ANFIS controller circuit.

The adaptive current controllers in this study, which have the structure depicted in Figure 12, perform the fuzzy controller’s function with the help of two inputs and three MFs from each input’s triangle, and Figure 13 shows its rules.

Figure 12

Neuro-fuzzy network structure.

Figure 13

Neuro-fuzzy rules.

The output MFs and rules were modified using a neural training system with a large number of inputs and outputs data sets. Figure 14a and b illustrates MFs, and the ANFIS input/output surface performance is illustrated in Figure 15.

Figure 14

Input/output membership functions: (a) input membership functions (error), and (b) output membership functions (integral error).

Figure 15

ANFIS input/output surface performance.

4.2 Intelligent multi-coils fast wireless charging of EVs

The charging time reduction is an important goal in the effort to make EVs more user-friendly. Here, fast charging presents a promising opportunity. It may cut down charging times to around 10–20 min. According to the DC Level 3 [15], 800 V DC is proposed to offer up to 30 kW (100 A). The entirely new design presented in this work for the receiving part (vehicle) will include multiple coils that are connected in parallel and operate simultaneously in smart charging systems. The proposed new model, in Figure 16, enables the vehicle’s battery to draw energy from the charging station based on the battery’s state of charge.

Figure 16

Block diagram of the proposed intelligent multi-coils fast wireless charging of EVs.

The adaptive intelligent wireless charging systems use multiple interconnected coils to satisfy the amount of battery charging current requirement. Neural networks are an artificial intelligence technique that can be constructed of one input, as in the symmetric sigmoid transfer function, and of one output, as a linear transfer function. Ten hidden layers can be used, as shown in Figure 17, to be trained on various datasets related to the state of charge of various cases and durations. Such a network can adapt and determine the optimal number of coils contributing to the shortest charging time based on the battery’s condition.

Figure 17

Intelligence neural network structure.

Figure 18 shows the flow chart of a fast wireless charging circuit.

Figure 18

Charging control flow chart.

The neural network closely and continuously monitors the battery’s state during charging. It compares the charging voltage and the state of charge with the stored and trained data. This allows us to select the number of coils and increase the charging voltage. This ultimately reduces the charging time. Meanwhile, the current remains constant at 100 A until the battery reaches 80% SoC. Then, it gradually decreases until it reaches zero when the battery is fully charged. The adaptive intelligence wireless charging system differs from the traditional wireless charging systems, which typically control the lithium battery charging process through a modulation index, thereby regulating the charging voltage. This requires an integrated communication system and feedback between the vehicle and the charging station. In the proposed model, a charging control is managed entirely by the vehicle through a neural network trained on various current, voltage, and charging state scenarios. Overall multi-coil of wireless fast charging of EVs is shown in Figure 19.

Figure 19

Overall simulation of proposed intelligent multi-coils fast wireless charging of EV.

5 Simulation results

The CC–CV has been identified as the main factor in centralized EV charging stations to reduce the waiting period for each EV, particularly during busy periods. Figure 20 shows the current, voltage, and SoC of the battery and the CC–CV charging curve of the lithium battery using the ANFIS.

Figure 20

(a) current battery charging waveform (b) voltage battery charging waveform (c) state of charge (SoC%) (d) CC–CV charging curve of Li-ion battery used ANFIS.

This suggests the design of a series of multi-coils linked serially with the main receiving coil within the EV. This arrangement aims to increase the efficiency of the electrical energy transfer from the charging unit to the vehicle, thus enabling swift or fast charging contingent on the different battery’s state and requirements. The model will demonstrate the fluctuations in current and voltage waveforms, as well as the charging status for various configurations of the auxiliary coils that are connected to the main receiving coil. This approach seeks to optimize energy transfer and to adapt charging rates to extend battery life and performance. After adding three coils in series with the main receiver coil, the voltage will increase until the battery reaches full charge, as shown in Figure 21a.

Figure 21

(a) Voltage battery charging. (b) Current battery charging. (c) SoC% (3 coils).

The controller sequentially selects coils to increase the magnetic field strength and the energy transferred to the battery, as indicated by the current peaks. Every peak is a charging rate step, as shown in Figure 21b. The staged approach allows the system to adjust the battery acceptance rate while charging. Peaks must not exceed the battery’s thermal or electrochemical limits to avoid possible damage.

A high current and its possible spike are unsafe if the battery is near its full capacity because of the decreases in the acceptance rate. To determine the timing and amplitude of the current peaks, the smart controller’s charging protocol considers different factors like battery chemistry, state of charge, internal resistance, temperature, and the desired final voltage. The state of charge of the battery with three auxiliary coils is shown in Figure 21c. To mitigate the risks associated with current spikes and their adverse effects on the electronic circuit components and to accelerate the charging process whenever necessary, a strategy of increasing the number of auxiliary coil series connected to the main coil in the wireless charging system in EVs is adopted. Increasing the number of coils contributes to minimizing the spike current to 122 A. This process acts to distribute the transferred energy across multiple points, rather than concentrating it at a single point, to reduce the possibility of reaching magnetic saturation in the coils and preventing thermal stress caused by high spike currents. Consequently, this helps protect electronic components and generally increases the efficiency of the charging process. Figure 22 shows the charging voltage, current, and SoC with four auxiliary coils.

Figure 22

(a) Voltage battery charging. (b) Current battery charging. (c) SoC% (4 coils).

To improve and develop EVs, fast charging infrastructure and advanced technologies, including coil structure modification and control methods, are used to increase charging efficiency while minimizing waiting times. The goal is to obtain more rapid, safer, and more efficient charging while preserving components. Five sequential auxiliary coils can increase the EV’s wireless charging current and reduce the spike current. Also, using increased current can accelerate battery charging, achieving optimal charge status faster. Reducing spike currents minimizes stress on electronic and electrical components and the vehicle’s charging system control panel, extending their lifespan and reducing damage risk. This improved charging efficiency and reduced charging time is the major step toward fast and reliable EV charging systems. Figure 23 shows the voltage, current, and the State of Charge (SoC) by adding five coils. In the current waveform, it is observed that the spikes are about 115 A, and the peak current is 112 A, which leads to increasing the state of charge more rapidly.

Figure 23

(a) Voltage battery charging. (b) Current battery charging. (c) SoC% (5 coils).

Figure 24 illustrates the battery status, in response to increasing the number of coils, for the cases studied and simulated using MATLAB’s Simulink. The figure presents a comparison of the evolution of the battery’s charge state over time for various scenarios with differing coil counts. The study indicates that an increase in the number of coils contributes to achieving optimal battery charging in a reduced timeframe. This suggests that enhancing the coil configuration can positively affect the efficiency and speed up the charging.

Figure 24

A comparison of SoC battery cases.

6 Experimental test results

6.1 A novel coaxial nested coil (CNC) of the wireless charging system of EV

The proposed prototype model, as shown in Figure 25, relies on advanced inductive charging technology, wherein the secondary coil is centrally positioned within the primary coil, with a precise air gap not exceeding 1 mm. This arrangement ensures high and efficient energy transfer with minimal losses. The secondary coil in this system is not singular but comprises a main coil and a series of secondary coils connected in series. These coils work harmoniously to meet the changing needs of the battery and to provide a rapid charging option for the vehicle owner when needed. Figure 26a and b shows a novel CNC design of wireless EV charging in simulation design and prototype.

Figure 25

Prototype wireless charging of EV with novel CNC coils.

Figure 26

(a) A novel CNC design of wireless EV charging. (b) Coils in the prototype.

The integration of a charging received coil system to match the battery’s needs and to achieve the necessity of fast charging represents a significant advancement through the operation of a prototype. Testing has shown that the use of mono and multi-coils, as shown in Figure 27a and b, can achieve a full charge.

Figure 27

State of charge of the EV battery in (a) CNC-mono coil. (b) CNC-multi-coil.

This innovative design provides the capability to offer diverse charging options, both soft and fast, to satisfy the required conditions. The importance of this design is highlighted by the relationship between the charging power, which is a function of the charging speed, and the state of charge, as shown in Figure 28. This presents a flexible and efficient approach to meet the changing demands of EV charging.

Figure 28

Speed charging effects on the state of charge with CNC mono/multi-coil.

7 Conclusions

This study examined and implemented the series–series wireless charging of EVs. This is to resolve the long charging time problem, which is one of the major challenges EV users and owners face. This issue has led researchers to reduce charging duration while ensuring system components and battery lifespan integrity. Power electronics converters are employed for the power transmission and the receiving side. The presented work aims to introduce a novel design capable of regular and fast EV charging. By increasing the number of added coils, charging time decreases during urgent situations, consequently reducing complaints from EV owners about charging durations and reducing queues at regular charging stations. This, in turn, encourages more individuals to adopt EVs. The AC voltage source is boosted and converted to DC voltage using a three-phase full-wave controlled rectifier. This achieved 800 V DC with minimal harmonics, around 3.47%, within the acceptable range set by the IEEE standards. The CC–CV is an essential approach in centralized EV charging stations that reduces EV queues, especially during busy times. The smart charging systems are needed for fast-changing EVs. EV adoption requires reliable, fast, battery-adaptable charging. Traditional EV charging uses CC–CV, which is effective but hard to control for battery life and performance. ANFIS uses neural network learning and fuzzy logic. This controller regulates and maintains 100 A during battery charging up to 80%, which gradually decreases to 0 A until the battery reaches its fully charged state. The controller stabilizes voltage during charging. An innovative design is proposed for the main charging coil. CNC is proposed as a model that relies on advanced inductive charging technology, wherein the receiver coil is centrally positioned within the transmitter coil with a precise air gap of about 1 mm. This arrangement ensures highly efficient energy transfer with minimal loss. A receiver coil consisting of a main coil with a series of auxiliary coils connected sequentially is designed for use in EVs to control the charging process according to the required electromagnetic field and to minimize the charging time in urgent cases. An intelligent selecting switching circuit, based on a neural network, estimates the necessary charging rate based on the state of charge to control the whole charging process. In a series configuration, the neural network activates the auxiliary coils on demand. To mitigate the negative impacts of these spikes on electrical and electronic components and to minimize the charging time, it is proposed to increase the auxiliary charging coils to five within the main receiving unit. This reduces current spikes by approximately 50% to ensure a more stable, efficient, and reliable charging system. This work is valuable for automotive companies, EV users, and industry professionals. It addresses common barriers to EV adoption compared to traditional internal combustion engine vehicles, providing short- and long-term benefits. Our findings offer insights applicable across diverse contexts within the EV industry, benefiting a wide range of interested parties.