electromagnetic shielding techniques in the wireless power

Research ArticleElectromagnetic Shielding Techniques in the Wireless PowerTransfer System for Charging Inspection Robot Application

Chaoqun Jiao ,1 Yang Xu ,1 Xiang Li,1 Xiumin Zhang,1 Zhibin Zhao,2

and Chengzong Pang3

1School of Electrical Engineering, Beijing Jiaotong University, Beijing 100044, China2State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources,North China Electric Power University, Beijing 102206, China3Department of Electrical Engineering and Computer Science, Wichita State University, Wichita, KS, USA

Correspondence should be addressed to Chaoqun Jiao; [email protected] and Yang Xu; [email protected]

Received 9 March 2021; Revised 6 May 2021; Accepted 12 June 2021; Published 5 July 2021

Academic Editor: Giuseppina Monti

Copyright © 2021 Chaoqun Jiao et al.,is is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Aiming at eliminating the leakage of magnetic fields from the wireless power transfer (WPT) system, the structural and workingcharacteristics of the WPTsystem for the inspection robot are analyzed and an electromagnetic shielding method combining passiveshielding and active shielding is proposed in this paper. Firstly, we simulated the magnetic field distribution of the WPT system inMaxwell. Secondly, passive shielding is configured in theWPTsystem, and the material, size, and position of the passive shielding arestudied. ,en, we add active shielding to areas where passive shielding cannot achieve a good shielding effect. Based on the analysisand summary of the two methods, we shield the WPT system in the horizontal direction with the appropriate size and distance ofaluminum plate, and in the vertical direction, we use the active shielding coils. Simulation and experimental results show that thescheme only slightly reduces the transmission efficiency of the system (from 80.2% to 77.6%), but the shielding ability is 34.06%higher than that of only aluminum plates. ,e excellent effect of the proposed shielding method is verified in our experiment.

1. Introduction

,e traditional wired charging technology is still the mainway of power transmission at present, and the powertransmission is realized by the physical connection of plugand socket. However, it also has a variety of problems, suchas safety problems caused by abrasion and aging, the in-convenience of charging portable electronic equipment andimplantable medical equipment, and the danger of powertransmission in mine and underwater working environ-ments. Wireless charging technology, as a new powertransmission mode, can effectively realize noncontact powertransmission and meet the needs of safety, efficiency, andconvenience in some specific scenarios.

Because the WPT system needs to use the magnetic cou-pling mechanism to transmit energy with the high-frequencymagnetic field (MF), the electromagnetic radiation problemwillbring panic to the public and even a real security threat [1–3].

Inspection robot is the product of power system auto-mation in recent years. Its working environment is complex.,e traditional charging mode of the power supply line andelectric shock occupies its working time and reduces pro-duction efficiency. ,e wireless charging strategy allows theinspection robot to complete the inspection task along aspecific track and replenish electric energy. However, theinspection robot will have an adverse effect on the elec-tromagnetic environment during wireless charging, so it isnecessary to design a shielding method for the WPTsystem.

Passive shielding is a common electromagnetic shieldingstrategy. Passive shielding refers to the optimization of themagnetic channel by using the magnetic conductivity ofmagnetic materials or the formation of reverse MF by eddycurrent produced by conductive metal materials in high-frequency MF, so as to suppress the leakage of MF [4–6].Attaching an aluminum (Al) plate to the transmit (Tx) coiland/or the receive (Rx) coil is the most prevalent measure,

HindawiInternational Journal of Antennas and PropagationVolume 2021, Article ID 9984595, 15 pageshttps://doi.org/10.1155/2021/9984595

mailto:[email protected]

mailto:[email protected]

https://orcid.org/0000-0002-4311-9947

https://orcid.org/0000-0002-9891-5967

https://creativecommons.org/licenses/by/4.0/



https://doi.org/10.1155/2021/9984595

which exploits the eddy current induced in the high-con-ductivity metal to block the leakage field. For stationaryelectric vehicle (EV) wireless charging applications, re-gardless of whether the coil structure is a basic circular pad[7], DD pad [8, 9], DDQ pad [10], or even bipolar pad [11],the Al-plate passive shielding [12–14] is widely used.

Active shielding is relatively novel, but it can also play agood shielding effect in the WPTsystem. In some cases witha large air gap, the leakage MF in the horizontal directioncannot be ignored. ,e suppression coil with excitationsource produces the offset MF opposite to the original MF toclear or weaken the leakage MF [6, 15–17]. ,e shieldingeffect is obvious in a dynamic wireless charging system, butthis method is unfavorable to the overall efficiency of thesystem, so it is necessary to design a suitable shieldingscheme [18, 19].

Previous studies to reduce the EMF by using ferrite tileor another material to cover the WPT system are heavy andcause additional costs [20]. And those works by usingshielding coil lack experimental verification usually [21, 22].Moreover, the coil form of the WPT system is limited to flattype in previous works, and the research on electromagneticshielding of I-shaped WPT system is lacking. In this paper,we use Simplorer and Maxwell to conduct a field-circuitcosimulation for the WPT system at first. On this basis, thecorresponding shielding scheme is designed according to themagnetic field distribution characteristics of the primaryside and secondary side, and the advantages of passive andactive shielding in the scheme are fully utilized. Afterward,the position, material, and size of the shielding device arediscussed and optimized. Our research shows that thevertical shield at the receiving side is more favorable toreduce the magnetic field around the system, and the activeshielding coil is better than the aluminum plate in thevertical direction. ,e simulation and experimental resultsshow that the method can effectively reduce the adverseeffects on the electromagnetic environment during theworking process of the magnetic coupling mechanism andensure the working efficiency of the system.

2. Magnetic Coupling Mechanism of theInspection Robot

2.1. Selectionof theSecondaryCoil. According to the workingcharacteristics of the slow-moving inspection robot, itsresonant coil should be designed in accordance with therequirements of dynamic wireless charging. ,e secondarycoil is located under the robot chassis. Due to the limitedinstallation area of the robot chassis, space must be used asmuch as possible, and the skin effect, eddy current effect,magnetic saturation, and other factors must be fully con-sidered to obtain a higher coupling coefficient and trans-mission efficiency. For the selection of resonant coils, thereare more research studies on the flat circular, E-typemagnetic core skeleton, double D type, and so on [13]. ,eflat circular shape is used as the secondary side although itcan be charged efficiently under ideal conditions facing theprimary side, but considering that the installable area of theactual inspection robot is a rectangular space of about

300mm× 150mm, the design size of this shape will belimited and the charging effect will be greatly reduced. As thesecondary side, the E-shaped core frame can be designed tobe rectangular with multiple windings. ,e core frame usedcan also improve the antioffset ability of the coil, but thestructure is bulky and difficult to flatten the design. DoubleD type is formed by two circular coils connected in reverseseries, generating MF in opposite directions. Figure 1 showsthe effect diagram of different winding methods on thesecondary side. Compared with the other two coils, double Dwinding is lighter and has a larger coupling coefficient andstronger antioffset ability.

2.2. Selection of the Primary Coil. As to the primary side ofthe dynamic wireless charging system, early studies haveadopted the long rail type and the segmented rail type. ,elong rail type is characterized by its simple coil structure, butit is restricted by the length of the rail when designing thesize, number of turns, and self-inductance. It is not flexibleenough and is not suitable for scenarios with long charginglines. Segmented guide rails make up for the shortcomings oflong guide rails, and the combination method is flexible, buteach part requires independent compensation devices andswitch control strategies, and the system is more compli-cated and costly.,is paper adopts the primary coil layout ofthe I-shaped magnetic core skeleton shown in Figure 2. ,etwo magnetic cores are a set, the Litz wire is wound on theskeleton clockwise and counterclockwise in the drivingdirection, and the size and winding method of the secondaryside are the same. Correspondingly, this double-magnetic-pole coupling mechanism has the following advantages:

(1) ,e coils on a single set of magnetic poles are ar-bitrarily superimposed within the range of themagnetic pole height, which can increase the in-ductance and improve the vertical transmissioncapacity

(2) Since the primary coil needs to be buried below theground, the H-shaped coil has a narrow width, whichis beneficial to reduce the track excavation width

(3) It has the characteristic of modularization, con-necting multiple groups of primary coils with wires,which can extend the inspection path and facilitatetroubleshooting

(4) ,e adjacent magnetic poles wind in opposite di-rections so that the polarity of the magnetic in-duction intensity B of each pole is reversed accordingto the current direction. ,erefore, the main mag-netic flux circulates between the nearest poles toform a closed MF and reduce the leakage MF aroundthe power rail

2.3. Analysis of the Equivalent Circuit. According to thedifferent numbers and composition modes of resonantcapacitors and inductors, the tuning circuit of the wirelesscharging system can be divided into four basic resonanttopologies and six composite topologies. Among them,

2 International Journal of Antennas and Propagation

the series-series (SS) compensation topology is simple instructure and convenient in parameter design. It issuitable for the voltage type system in this study. ,eimpedance model of the compensation structure is shownin Figure 3.

Lp and Ls are the inductances of the primary coil andsecondary coil; meanwhile, Cp and Cs are the resonant ca-pacitances of Lp and Ls. Rp and Rs represent coil resistances.Because of the large transmission air gap, the primary andsecondary sides are weakly coupled, and the secondary sideis reduced to the primary side in the form of reflectedimpedance.

2.4. Simulationof theMagneticCouplingMechanism. We usethe finite element simulation software ANSYS to simulatethe ideal working environment of the system. Figure 4shows the situation where the primary coil and the sec-ondary coil are completely aligned. ,e charging guideline is long, and only a small section is selected formodeling. According to the robot’s chassis size, launchtrack size, and transmission requirements, the selectedresonant coil parameters are as follows: the number ofturns of the double D coil on the secondary side is24mm × 2mm; using the dense winding method, the totallength is 300mm; the total width is 150mm; the trans-mitter side is H type. ,e number of turns of the coil is16mm × 2mm, and the dense winding method is alsoadopted. ,e total length is 300mm, the total width is100mm, and the total height is 55mm; the closest air gapbetween the primary side and the secondary side is 50mm.

We select a horizontal observation line 1 which is 50mmabove the secondary side with a length of 300mm and avertical observation line 2 which is 50mm to the right of thesecondary coil and 200mm above the ground for obser-vation (the plane with Z� −50mm is specified as theground). Because there are communication, control, andother hardware equipment 50mm above the secondary side,

there may be electrical equipment being inspected andmaintenance workers 50mm to the right of the secondaryside. ,ese areas need to be protected. In addition, a hor-izontal observation line 3 with a length of 125mm is set onthe right side of the primary coil, which is used to grasp thespread of the MF above the ground. Observing the distancefrom the resonant coil can ensure that there is no electro-magnetic interference. Using Simplorer and Maxwell toconduct field-circuit cosimulation, when the magneticcoupling mechanism is under the rated operating conditionsof 2 kW and 85 kHz, the simulation results of each obser-vation line are shown in Figure 5. According to ICNIRP[23, 24], the magnetic induction intensity in the nonworkingarea of the magnetic coupling mechanism seriously exceedsthe standard, and shielding measures are needed to reducethe magnetic induction intensity in the designated area tothe limit of 27 μT.

(a) (b) (c)

Figure 1: Effect diagram of different winding methods on the secondary side. (a) Flat circular winding. (b) Double D winding. (c) E-typecore skeleton winding.

Figure 2: Effect diagram of I-shapedmagnetic core skeleton coil onthe primary side.

Vac

Cp Cs

RL

Ip M Is

Lp

Rp

Ls

Rs

Figure 3: Impedance model of SS type compensation structure.

Z

Y

X

O

50m

m

200m

m

50m

m

125mm

50mm

300mmLine 1

Line

2

Line 3

Figure 4: Simulation model of wireless charging magnetic cou-pling mechanism.

International Journal of Antennas and Propagation 3

3. Passive Shielding

3.1. Passive Shielding Mechanism. ,e wireless chargingsystem can be shielded by magnetic materials. Commonmagnetic materials such as ferrite and iron-nickel alloy havethe characteristics of high permeability and high saturationmagnetic flux. ,e internal magnetic resistance is extremelysmall, and the magnetic flux circuit will be guided. It ispossible to achieve the purpose of improving the MF dis-tribution and weakening the local field strength throughmagnetic materials.

,e alternating magnetic flux density B generated by theresonant coil will induce an electromotive force in thesurrounding space, the induced electromotive force canform an eddy current in the conductor, and a counteractingMF will be generated to opposite to the original one. ,eeffect of eddy current is shown in Figure 6. Accordingly, theeddy current can be used to reduce the leaked magnetic field.

3.2. Material Optimization of the Passive Shielding Device.In order to analyze the shielding effects of different magneticmaterials, this paper selects iron-nickel alloy and ferriteshields for simulation. We add iron-nickel alloy and ferriteplates of the same thickness and same size 3mm above thesecondary side coil, and the magnetic induction intensitycloud diagram of the YOZ plane is as follows. Comparedwith the case without shielding in Figure 5, it can be seenintuitively that the magnetic induction intensity value on theupper and left sides becomes low, and the magnetic

induction intensity value of the magnetic coupling areabetween the primary and secondary coils becomes high.

,e result in Figure 7 shows the shielding effect of thetwo materials. On observation line 1 above the secondaryside, the ferrite shield plate represented by the blue curvedoes not change the MF change law of the couplingmechanism, while the iron-nickel alloy represented by thered curve fully reflects its advantage of high magneticpermeability. ,e level of magnetic induction on the ob-servation line is greatly reduced. Considering that bothmaterials can achieve a good shielding effect, and the metalnickel in the alloy shielding plate is more expensive, wechoose a ferrite shielding plate that is cheaper and has awider operating frequency range.

7.5402E + 002B (uTesla)

Observation line 17.0375E + 0026.5349E + 0026.0322E + 0025.5295E + 0025.0268E + 0024.5242E + 0024.0215E + 0023.5188E + 0023.0161E + 0022.5134E + 0022.0108E + 0021.5081E + 0021.0054E + 0025.0275E + 0017.4677E – 003

(a)

150

140

130

120

B (u

T)

110

100

90

800 50 100 150

Y (mm)200 250 300

(b)

45

40

35

30

B (u

T)

25

20

15

100 4020 60 80 100

Z (mm)120 180160140 200

(c)

Figure 5: MF distribution curve during normal operation. (a) MF distribution of the YOZ plane. (b) Distribution of observation line 1.(c) Distribution of observation line 2.

9.3332E + 003Z

J (A/m2)

8.7110E + 0038.0888E + 0037.4665E + 0036.8443E + 0036.2221E + 0035.5999E + 0034.9777E + 0034.3555E + 0033.7333E + 0033.1111E + 0032.4888E + 0031.8666E + 0031.2444E + 0036.2221E + 0028.9304E – 019

Figure 6: Eddy current effect.


According to impact factors of the eddy current, whenthe resonance frequency, shielding plate thickness, volume,and relative permeability are the same, the conductivitybecomes the determining factor of the eddy current loss.,ehigher the conductivity, the stronger the induced eddycurrent, and the more obvious the shielding effect. It isimportant to note that for the resonant wireless chargingsystem, the intensity of the induced eddy current is closelyrelated to the resonant frequency. When the frequency is toolow, conductive metal materials cannot guarantee a goodelectromagnetic shielding effect.

,rough simulation and comparison of the shieldingconditions of aluminum and SUS306 (stainless steel) plates,it can be seen from Figure 8 that the MF in the couplingregion of the primary and secondary coils is weaker thanbefore shielding. Compared with the cloud image withmagnetic material, the MF in the coupling region is greatlyweakened. ,at is, when a metal material is used forshielding, the induced eddy current can weaken the mainmagnetic flux while canceling the leakage MF, resulting inthe reduction of the self-inductance, mutual inductance, andcoupling coefficient of the magnetic coupling mechanism.

It can be seen from the simulation results of the ob-servation line that the shielding effects of the two metalshielding plates at 85 kHz frequency are very similar, andaluminum is the most abundant metal element in the earth’scrust. It is cheap, light, and conductive. In addition, the eddycurrent loss is only 0.48597W, which is much smaller thanthe 2.288W in the stainless steel plate, so an aluminum plateis selected as the metal material.

After understanding the mechanism of magnetic ma-terials and metal materials as the shielding layer, we try tocombine the advantages of the two to form a compositeshielding layer. We place an iron plate directly above thesecondary side and then place an aluminum plate of thesame size close to the iron plate to obtain the following cloudmap distribution in Figure 9.

,e composite shielding layer combines the effects ofmagnetic materials and metal materials, which not onlyenhances the MF coupling between the primary and the

secondary coils and increases the coupling coefficient to0.27593 but also reduces the leakage MF at the edge of thesecondary coil, and the overall shielding effect is better.

In Figure 10, all values at observation line 1 have beenreduced below standard, and the minimum value is 7 μT,which is 42.3% of theminimum shielding value of aluminumplate and 56.7% of the minimum value of ferrite shielding.On the vertical observation line 2, the standard value ofmagnetic flux density can be reached only 8.2mm above theground, and the exceeding range is 10.22% of the aluminumshielding.

In order to minimize the number of shielding materialsandmeet the expected shielding requirements, it is necessaryto simulate and optimize shielding devices with differentthicknesses. According to the material selection above,keeping the distance between the shielding layer and thesecondary coil constant, the simulation analysis of ferriteplates and aluminum plates of different thicknesses is per-formed, and the results are as shown in Figure 11.

For the ferrite shielding board, the thicker the thickness,the more the MF lines pass through the board, the less themagnetic leakage, and the better the shielding effect.However, manganese-zinc ferrite is heavy. Considering thatits shielding effect has met the standard value of ICNIRP-2010 when 2mm is considered, a ferrite plate of thisthickness can be used. ,e thickness has little effect on theshielding effect of the aluminum plate. Choosing a 0.5mmaluminum plate can ensure the shielding effect and savematerial consumption.

3.3. Passive Shielding Position Optimization. On the onehand, considering that the Mn-Zn ferrite may have magneticsaturation, it is difficult to transport due to its brittleness andfragility, and it is not suitable to be too large, etc.; on theother hand, the magnetic induction intensity within thespecified observation range already dropped below thestandard owing to the ferrite, aluminum plates of differentthicknesses were placed in the existing position, the MF onboth sides of 33mm near the edge of the secondary coil still

7.5309E + 002B (uTesla)


(a)

18

B (u

T)

468

10121416

20

0 50 100 150Y (mm)

200 250 300

FerriteAlloy

(b)

Figure 7: MF distribution curve in the presence of magnetic materials. (a) MF distribution of YOZ plane. (b) Distribution curve ofobservation line 1.


obviously exceeded the standard, and the MF within 80mmon the ground did not fall below the safe value. ,erefore,the use of an aluminum plate that is easy to cut should beconsidered as much as possible for shielding, and the po-sition of the shielding plate should be optimized.

3.3.1. Secondary Side Horizontal Shielding. Based on theabove conclusions on the shielding mechanism and thick-ness optimization of the aluminum plate, we can observe therelationship between the distance of the shield coil and the

coupling coefficient, eddy current loss, and magnetic in-duction intensity of the system.

Table 1 shows the simulation data of the horizontalaluminum plate in different positions. As the distancebetween the aluminum plate and the coil increases, thecoupling coefficient between the coils increases signifi-cantly, but at the same time, the shielding effect decreasesslightly. Taking into account the transmission efficiencyand shielding effect, 7 mm is selected as the distancebetween the aluminum shielding plate and the secondarycoil.

7.4932E + 002B (uTesla)


(a)

3836

B (u

T)

22242628303234

2018

0 50 100 150Y (mm)

200 250 300

A1SUS306

(b)

Figure 8: MF distribution curve in the presence of metallic materials. (a) MF distribution of YOZ plane. (b) Distribution curve ofobservation line 1.

7.5467E + 0027.0436E + 002

B (uTesla)

6.5405E + 0026.0374E + 0025.5343E + 0025.0312E + 0024.5260E + 0024.0249E + 0023.5218E + 0023.0187E + 0022.5156E + 0022.0125E + 0021.5094E + 0021.0863E + 0025.0318E + 0017.4791E – 003

Figure 9: MF distribution of YOZ plane in the presence of composite materials.

12.512

B (u

T)

8.59

9.510

10.511

11.5

87.5

0 50 100 150Y (mm)

200 250 300

(a)

30

25

B (u

T)

5

10

15

20

00 4020 60 80 100

Z (mm)140120 160 180 200

(b)

Figure 10: MF distribution curve in the presence of composite materials. (a) Observation line 1. (b) Observation line 2.


3.3.2. Vertical Shield on Secondary Side. ,e inspectionrobot must move along the launching side track, so there is acertain air gap between the secondary side and the primaryside. We choose aluminum plates with a length of 300mmand a width of 20mm and install them, respectively, on bothsides of 20mm from the secondary coil, as shown inFigure 12.

According to the simulation results, after installing thevertical shield, the range of the MF below the safe value inthe space has not been significantly expanded. It can be seenthat the passive shielding method is not suitable for verticalshielding on the secondary side of the magnetic couplingmechanism.

3.3.3. Horizontal Shielding on the Primary Side.Installing passive shielding devices on the primary side canalso weaken the local MF strength. We place a300mm× 55mm aluminum plate horizontally on theground, keeping a distance of 40mm from the primary coil,as shown in Figure 13.

On observation line 1 above the secondary side, therange below the safe value of the MF is 233.4mm. On thevertical observation line 2 beside the secondary coil, thevalue of the magnetic flux induction drops to 27 μT, and therange exceeding the standard value is 2.39% smaller than themethod with only the horizontal aluminum plate in thesecondary coil. On horizontal observation line 3 beside theprimary coil, the overstandard MF is also improved com-pared to the situation where only the secondary sideshielding is applied.

30B

(uT)

10

15

20

25

50 50 100 150

Y (mm)200 250 300

1mm2mm3mm4mm

(a)

0.5mm1mm

2mm4mm

36

B (u

T)

2224

20

26

3032

28

34

180 50 100 150

Y (mm)200 250 300

(b)

Figure 11: Shielding effect of different thickness plates. (a) Shielding effect of ferrite plates. (b) Shielding effect of aluminum plates.

Table 1: Simulation data of the horizontal aluminum plate on different positions.

Distance with the coils (mm) k Eddy current loss (w) ,e range over the standard (mm)3 0.054414 0.53951 785 0.073893 0.48597 80.27 0.090421 0.46328 83.89 0.10493 0.44639 87.6

7.4932E + 002B (uTesla)


Figure 12:MF distribution of theXOZ plane with vertical shieldingon the secondary side.

7.4932E + 002B (uTesla)


Figure 13: MF distribution of the XOZ plane with horizontalshielding on the primary side.


3.3.4. Primary Side Vertical Shield. An aluminum plate isplaced vertically on the primary side, and the direction isparallel to the direction of movement of the inspectionrobot. ,e simulation results of the XOZ plane MF distri-bution map are shown in Figure 14.

On the observation line 1 300mm above the secondaryside, there is almost no change in theMF distribution, but onthe observation line 2 on the right side of the secondary coil,it drops to 27 μT at 80.2mm above the ground, and theoverstandard range is lower than that when only the hori-zontal aluminum plate is added. On observation line 3 on theright side of the primary coil, the area where the MF in-tensity exceeds the standard is reduced by 8.31% comparedwith only the horizontal aluminum plate. It can be seen thatthis installation method can affect the distribution ofmagnetic induction in a wide range above the ground be-cause it restricts the diffusion of the magnetic induction lineto the surroundings on the primary side so that the directionbecomes upwards along the aluminum plate, which is moredirectly related to the secondary coil. ,is method can alsoimprove the coupling coefficient of the system.

4. Active Shielding

4.1. Active Shielding 8eory. ,e difference between activeshielding and passive shielding is that passive shielding usesconductive or magnetic materials to change the MF of theleakage area, while active shielding uses the MF generated bythe energized coil to change the MF of the leakage area.Among them, the current of the active shielding coil or themagnetic flux formed by it should be at the same frequencyas the resonance coil and have a phase difference of 180°.However, if the shielding coil is powered by an externalsource, it is difficult to sense and control the frequency andphase modulation process; on the other hand, it also reducesthe overall efficiency of the system. ,is paper uses a series-type infinite shielding method, which can be used withoutany external source. Under the circumstances, activeshielding is realized. ,e circuit impedance model is asshown in Figure 15.

From Figure 15, by changing the active suppression coilLSC with secondary side coil LSM, the relative position of theMF is generated by the resonant coil LSM. ,e generatedleakage MF is the same frequency and opposite phase, butthe active shielding coil should be placed as far as possibleoutside the normal coupling path of the resonant coil toreduce the impact on the transmission effect. In order tostudy the influence of the added active suppression coil onthe parameters in the original circuit, the current on theprimary side is reduced to the secondary side using the idealtransformer rule, and the following ,evenin equivalentcircuit of the secondary side is shown in Figure 15(b).

Due to the existence of an active shielding coil, therequired matching capacitance becomes smaller. On thepremise of achieving the purpose of offsetting the leakageMF, reducing the self-inductance Lsc of the suppression coilcan reduce the impact on the matching capacitance in theoriginal circuit. Based on this, we can optimize the structureand position of the active shielding coil.

4.2. Position Optimization of the Active Shielding Coil

4.2.1. Horizontal Suppression Coil on the Secondary Side.,e above optimized results of passive shielding show thatthe MF values on both sides of the 33mm edge of thesecondary side coil still need to be reduced. ,erefore, twohorizontal suppression coils with opposite directions areadded above the secondary side aluminum plate to offset theleakage MF on the left and right sides in different directions.

Figure 16 shows the connectionmethod of the horizontalsuppression coil in the simulation. ,ere is still a distance of50mm between the secondary coil and the chassis of thepatrol robot. Here, the number of turns of the suppressioncoil is tentatively set to 2. We change the height of thesuppression coil to obtain the magnetic induction intensitychange at the leftmost end of horizontal observation 1 abovethe secondary side coil.

,e horizontal axis of Figure 17 represents the dis-tance between the secondary coil and the active sup-pression coil. It can be seen that before 14mm, the fartherthe distance is, the better the shielding effect is. Afterward,the shielding effect deteriorates because the influence ofthe reverse MF generated by the suppression coil on theobservation point becomes stronger, which causes the MFat the observation point to increase. ,e effect of thenumber of turns of active suppression shielding effect isdiscussed hereinafter.

,e result of Figure 18 shows that, within a certain range,the active suppression coil is farther from the main magneticcircuit; the more turns, the better the shielding effect. In thisstructure, the increase in the number of turns has a littleeffect on the MF of the lateral vertical observation line 2, butit has a significant impact on the upper horizontal obser-vation line 1. When the number of turns is 4, the 50mmabove the receiving side is below the national standard limit.And its average value of 16.4 μT is 36.92% lower than the26 μT of aluminum only. ,erefore, we choose 4 as thenumber of turns of the suppression coil.

4.2.2. Vertical Suppression Coil on the Secondary Side.Previous studies have shown that the passive shielding andactive horizontal suppression coils on the secondary side canhardly affect the magnetic induction intensity on the sideobservation line 2 of the robot, so we try to install a vertical

7.4932E + 002B (uTesla)


Figure 14:MF distribution of theXOZ plane with vertical shieldingon the primary side.


suppression coil shown in Figure 19 to improve the MFdistribution here.

In order to compare with the case of adding verticalshielding on the secondary side in passive shielding, theposition, length, and width of the active suppression coil arekept consistent with the previous aluminum plate, whilechanging the number of the coil turns to suppress thecorresponding design optimization.

It is reported in Figure 20 that with the change of thenumber of the active coil turns, the magnetic inductionintensity on observation line 2 is significantly reduced. With3 turns, it can be guaranteed that all above 51.3mm on theground falls below the standard value. ,e shielding effect of

the vertical suppression coil under the same size is betterthan that of the aluminum plate. ,erefore, it is moresuitable for small magnetic coupling mechanismapplications.

5. Experiment and Method Validation

5.1. Experimental Platform. Figure 21 shows our MF mea-surement experimental platform. ,e frequency-trackinghigh-frequency inverter power supply is selected, which canprovide 85 kHz rated working power.,e power supply doesnot need to accurately input the signal of the specified

Vac

Cp Cs

RL

Ipkm

kc

Is

Lpm

Rp

Lsm

LscRs

(a)

V

Cs

RL

Lsm

Lsc Rs

(b)

Figure 15: Topology models of active shielding. (a) Impedance model of active shielding. (b) ,e equivalent circuit diagram.

Figure 16: Connection method of horizontal suppression coil.

35

34.5

34

33.5

Mag

netic

flux

den

sity

(μT)

33

32.5

32

31.5

318 9 10 11 12 13

Distance between primary coil and secondary coil (mm)14 15 16 17 18

Figure 17: Magnetic flux density on the left side of observationline1.

2 turns3 turns4 turns

35

30

B (u

T)15

20

10

25

50 50 100 150

Y (mm)200 250 300

Figure 18: MF distribution curve of observation line 1.

Figure 19: Connection method of vertical suppression coil.


frequency and can be automatically adjusted to ensure thatthe system is always in resonance. ,is can avoid the sit-uation that the output power and efficiency of the system aredrastically reduced due to the shift of the resonance pointwhen the system structure changes.

,e resonant coil is wound according to the simulationmodel parameters. ,e coil wire adopts Litz wire with aspecification of 0.2mm× 75mm strands, and the maximumfluid resistance value is 8.856A. ,e wire diameter is 2mmmeasured with a vernier caliper. ,e 75 strands of wiresinside the coil are insulated from each other. When thecurrent is applied, the stranded multistrand enameled wirecan effectively suppress the skin effect and reduce the wireloss. ,e I-type magnetic core is spliced by manganese-zinc

ferrite of different sizes. When splicing, attention should bepaid to reducing the gap to prevent the distortion of the MFlines.

As the most critical part of resonance, capacitance needsto match the actual measured value of coil inductance. Here,the impedance analyzer NF ZGA5920 is used to accuratelymeasure the self-inductance, and then the capacitance valueis calculated through the resonance condition of the seriescompensation.

,e received high-frequency energy is converted to DCthrough a rectifier module, and the load is composed of 72LED bulbs. ,e load can be more directly observed ex-perimental phenomenon when shielding device is added,and it will not affect the coupling state of the coil.

50

2530354045

B (u

T)

101520

50 4020 60 80 100

Z (mm)140120 160 180 200

1 turn coil2 turns coil3 turns coil

(a)

7.5488E + 0027.0456E + 002

B (uTesla)

6.5423E + 0026.0391E + 0025.5358E + 0025.0326E + 0024.5293E + 0024.0261E + 0023.5228E + 0023.0196E + 0022.5163E + 0022.0131E + 0021.5098E + 0021.0066E + 0025.0332E + 0017.4264E – 003

(b)

Figure 20: Shielding effect of vertical suppression coil. (a) MF distribution of different turns of the coil. (b) MF distribution diagram of XOZplane.

Magnetic-fieldmeasuringinstrument

Oscilloscope

Host computer

Robotic arm

Console

Resonant coil

High-frequencyinverter power

supply

Figure 21: MF measurement experimental platform.


5.2. Experimental Data and Analysis. Table 2 shows theelectromagnetic configuration used in our simulation (sim)and experiment.

5.2.1. Magnetic Shielding Device Verification. First, we testthe situation without a shielding device and with magneticmaterial above the secondary side. ,e magnetic material,manganese-zinc ferrite, is located 3mm above the secondarycoil and is spliced with a square core of 2mm thickness. ,eposition of the magnetic material is shown in Figure 22.

After setting the initial measurement position and themeasurement length of the two axial directions on the hostcomputer interface, we power on the wireless chargingsystem to make it reach the resonance state. We use the hostcomputer to drive the console and the robotic arm to makethe three-dimensional MF measuring instrument collect theMF value on the specified plane and then output it throughthe connected USB interface to draw the following three-dimensional MF distribution map in Figures 23 and 24.

X� 75mm in the X-Y measurement plane andY� 65mm in the Y-Z measurement plane correspond toobservation lines 1 and 2, respectively. ,e test result of themeasuring instrument shows that without any shielding, themaximum magnetic induction intensity at 50mm above thesecondary side can reach 140 μT, and the minimum value is60 μT, far exceeding the corresponding ICNIRP-2010. Afteradding the ferrite shielding device, the MF values in the twomeasurement planes dropped sharply, the maximum drop inthe X-Y measurement plane was 88.57%, and the maximum

drop in the Y-Zmeasurement plane was 34.88%. In addition,the ferrite shielding device increases the system efficiencyfrom 80.2% to 82.4%.

5.2.2. Metal Shielding Device Verification. As is shown inFigures 25 and 26, the experimental results are consistentwith the simulation MF distribution law. ,e aluminumplate minimizes the central MF of the X-Y measurementplane, which is 85.71% lower than that without anyshielding.,e closer to the surroundings, the greater the MFvalue. Compared with the case without shielding, more than85% of the area in the plane is replaced by the blue areabelow 25 μT, and the magnetic induction intensity of the testplane is basically up to the standard.

5.2.3. Active Shielding Device Verification. On the basis ofpassive shielding, two types of active suppression coils areadded. ,e horizontal suppression coil selects 4 turns and isplaced 14mm above the secondary coil, as shown in Fig-ure 27. ,e measured self-inductance is 12.93 μH; the ver-tical suppression coil selects 3 turns and places it. At 20mmto the right of the secondary coil, the total length and totalwidth of the two vertical suppression coils with oppositewinding directions are consistent with the vertical alumi-num plate on the secondary side, and the self-inductance ismeasured to be 4.13 μH. We connect the two active sup-pression coils to the circuit, respectively, and continue themeasurement after the system is in resonance.

Table 2: ,e parameters of the experiment.

Parameters Primary side Secondary sideFrequency 85 kHzDistance 50mmReal power 2.46 kW 1.968 kWSelf-inductance (sim) 360.81 μH 91.476 μHSelf-inductance 354.76 μH 94.7 μHResonant capacitance 10.1 nF 37 nFResistance 0.341Ω 0.407ΩNumber of turns 16× 2 24× 2

Figure 22: Position of magnetic material shielding.


,e results in Figure 28 show that the active shield has amore significant shielding effect than the horizontal alu-minum plate, which can reduce the magnetic inductionintensity to a minimum of 8 μT, and the edge value of theplane also drops below the safe value. ,e Y-Z plane resultsshow that the vicinity of the vertical suppression coil can be

placed below the safe value, which ensures the electro-magnetic safety of electronic equipment and staff in non-working areas for a long time. ,rough the measurement ofthe oscilloscope, it is found that the transmission efficiencyhas dropped from 80.2% without the shielding device to77.6%, making the system efficiency slightly lower.

B (u

T)

160140120100

806040

300200

100Y (mm) X (mm)0 050

100150

5060708090100110120130140

(a)

B (u

T)

30

25

20

15

10

50

100200Y (mm) Z (mm)300 100 80 60 40 20 0

10121416182022242628

(b)

Figure 23: 3D MF distribution without shielding. (a) X-Y measurement plane. (b) Y-Z measurement plane.

B (u

T)

17

16

15

14

13

12300

200100Y (mm) X (mm)0 0

50100

15013

12.5

13.5

14

14.5

15

15.5

16

(a)

B (u

T)30

25

20

15

10

50

100200Y (mm) Z (mm)300 100 80 60 40 20 0

10121416182022242628

(b)

Figure 24: 3D MF distribution with magnetic shielding. (a) X-Y measurement plane. (b) Y-Z measurement plane.

(a) (b)

Figure 25: Position of metal material shielding. (a) 0.5mm aluminum plate on the receiving side. (b) 2mm aluminum plate on the receivingside.


6. Conclusions

In this paper, we proposed an electromagnetic shieldingmethod combining passive shielding and active shielding. Asa result of Maxwell simulation and experiment, it is con-firmed that the shielding method significantly reducedleaked magnetic field by almost 85% for the WPT system

with a decrease in efficiency by 2.6%. Moreover, the fol-lowing conclusions can be drawn from our study:

(1) Passive shielding can achieve shielding effect on theprimary or secondary sides of the WPT, but theshielding ability in the vertical direction is not asgood as the active shielding, and a large area of

B (u

T)

40

35

30

25

20

15300200

100Y (mm) X (mm)0 0

50100

150

20

25

30

35

(a)

B (u

T)

50

40

30

20

100

100200Y (mm) Z (mm)300 100 80 60 40 20 0

20

25

30

35

40

45

(b)

Figure 26: 3D MF distribution of X-Ymeasurement plane. (a) 0.5mm aluminum plate on the receiving side. (b) 2mm aluminum plate onthe receiving side.

(a) (b)

Figure 27: Physical picture of active shielding. (a) Horizontal suppression coil. (b) Vertical suppression coil.

B (u

T)

302520

5

1510

300200

100Y (mm) X (mm)

0 050

100150

8101214161820222426

(a)

B (u

T)

50

40

30

20

100

100200Y (mm) Z (mm)300 100 80 60 40 20

020

25

30

35

40

45

(b)

Figure 28: 3D MF distribution of the system with active shielding. (a) X-Y measurement plane. (b) Y-Z measurement plane.


shielding board will hinder the mobility of the WPTsystem

(2) Magnetic material and conductive material have anopposite effect on coupling coefficient so that we cancombine those two materials in passive shielding.But it should be pointed out that the thickness offerrite has a great influence on the shielding effect,and the thickness of the aluminum plate has a littleeffect on the shielding effect, but all of them arepositively correlated

(3) Under the same proportion of space, the MFweakening ability of active shielding is 34.06% higherthan that of only passive shielding

(4) ,e secondary side is the best mounting position foractive shielding

For future works, a more realistic multiphysical fieldmodel will be established to analyze the electromagnetic lossof the WPTsystem and the resonant frequency offset causedby the shielding design. In addition, the saturation phe-nomenon of the I-shaped core skeleton and optimizing itsstructure will be studied in further work.

Data Availability

,e data that support the findings of this study are availableon request from the corresponding author.

Conflicts of Interest

,e authors declare that there are no conflicts of interest.

Acknowledgments

,is work was supported by Joint Funds of the NationalNatural Science Foundation of China (U20A20305).

References

[1] P.-P. Ding, L. Bernard, L. Pichon, and A. Razek, “Evaluationof electromagnetic fields in human body exposed to wirelessinductive charging system,” IEEE Transactions on Magnetics,vol. 50, no. 2, pp. 1037–1040, 2014.

[2] C. Chen, X. Huang, and L. Tan, “Electromagnetic environ-ment and security evaluation for wireless charging of electricvehicles,” Transactions of China Electrotechnical Society,vol. 30, no. 19, pp. 61–67, 2015.

[3] G. Xu, C. Li, and J. Zhao, “Electro-magnetic environmentsafety study of wireless electric vehicle charging,” Transactionsof China Electrotechnical Society, vol. 32, no. 22, pp. 152–157,2017.

[4] L. Zhai, G. Zhong, Y. Cao, G. Hu, and X. Li, “Research onmagnetic field distribution and characteristics of a 3.7 kWwireless charging system for electric vehicles under offset,”Energies, vol. 12, no. 3, p. 392, 2019.

[5] M. Mohammad, E. T. Wodajo, S. Choi, and M. E. Elbuluk,“Modeling and design of passive shield to limit EMF emissionand to minimize shield loss in unipolar wireless chargingsystem for EV,” IEEE Transactions on Power Electronics,vol. 34, no. 12, pp. 12235–12245, 2019.

[6] H. Kim, C. Song, D. H. Kim et al., “Coil design and mea-surements of automotive magnetic resonant wireless chargingsystem for high-efficiency and low MF leakage,” IEEETransactions on Microwave 8eory and Techniques, vol. 64,pp. 383–400, 2016.

[7] Q. Zhu, Y. Guo, L. Wang, C. Liao, and F. Li, “Improving themisalignment tolerance of wireless charging system by op-timizing the compensate capacitor,” IEEE Transactions onIndustrial Electronics, vol. 62, no. 8, pp. 4832–4836, 2015.

[8] M. Budhia, J. T. Boys, G. A. Covic, and C.-Y. Huang, “De-velopment of a single-sided flux magnetic coupler for electricvehicle IPT charging systems,” IEEE Transactions on Indus-trial Electronics, vol. 60, no. 1, pp. 318–328, 2013.

[9] A. Zaheer, H. Hao, G. A. Covic, and D. Kacprzak, “Investi-gation of multiple decoupled coil primary pad topologies inlumped IPT systems for interoperable electric vehiclecharging,” IEEE Transactions on Power Electronics, vol. 30,no. 4, pp. 1937–1955, 2015.

[10] V. De Santis, T. Campi, S. Cruciani, I. Laakso, andM. Feliziani, “Assessment of the induced electric fields in acarbon-fiber electrical vehicle equipped with a wireless powertransfer system,” Energies, vol. 11, no. 3, p. 684, 2018.

[11] A. Zaheer, D. Kacprzak, and G. A. Covic, “A bipolar receiverpad in a lumped IPT system for electric vehicle chargingapplications,” in Proceedings of the IEEE Energy ConversionCongress and Exposition (ECCE), pp. 283–290, Vancouver,Canada, September 2012.

[12] J. Kim, J. Kim, S. Kong et al., “Coil design and shieldingmethods for a magnetic resonant wireless power transfersystem,” Proceedings of the IEEE, vol. 101, no. 6,pp. 1332–1342, 2013.

[13] H. Hwansoo Moon, S. Sungkyu Kim, H. H. Hyun Ho Park,and S. Seungyoung Ahn, “Design of a resonant reactive shieldwith double coils and a phase shifter for wireless charging ofelectric vehicles,” IEEE Transactions on Magnetics, vol. 51,no. 3, pp. 1–4, 2015.

[14] Y. Jing, B. Sun, and Z. Zhu, “Shielding effectiveness of infinitemetal plate with finite thickness at oblique incident and ar-bitrary polarizations,” in Proceedings of the Conference onEnvironmental Electromagnetics (CEEM), pp. 241–245,Hangzhou, China, November 2015.

[15] S. Cruciani, T. Campi, F. Maradei, and M. Feliziani, “Activeshielding design for wireless power transfer systems,” IEEETransactions on Electromagnetic Compatibility, vol. 61, no. 6,pp. 1953–1960, 2019.

[16] S. Cruciani, T. Campi, F. Maradei, and M. Feliziani, “Activeshielding applied to an electrified road in a dynamic wirelesspower transfer (WPT) system,” Energies, vol. 13, no. 10,p. 2522, 2020.

[17] J. Ren, B. Wang, and Z. Chen, “Density functional theorystudy of the magnetic properties of rare earth complexes: themagnetic coupling mechanism in YIII and GdIII complexeswith nitronyl nitroxide,” Science in China Series B: Chemistry,vol. 52, no. 11, pp. 1961–1968, 2009.

[18] T. Campi, S. Cruciani, V. De Santis, and F. Maradei, “Wirelesspower transfer (WPT) system for an electric vehicle (EV): howto shield the car from the magnetic field generated by twoplanar coils,” Wireless Power Transfer, vol. 5, no. 1, pp. 1–8,2018.

[19] Q.-F. Li, S.-B. Chen, W.-M. Wang, H.-W. Hao, and L.-M. Li,“Improving the efficiency of magnetic coupling energytransfer by etching fractal patterns in the shielding metals,”Frontiers of Information Technology & Electronic Engineering,vol. 17, no. 1, pp. 74–82, 2016.


[20] Y. Kitano, H. Omori, T. Morizane, N. Kimura, andM. Nakaoka, “A new shielding method for magnetic fields of awireless EV charger with regard to human exposure by eddycurrent and magnetic path,” in Proceedings of the Interna-tional Power Electronics and Application Conference andExposition, pp. 778–781, Shenzhen, China, November 2014.

[21] J. Park and S. Ahn, “A novel shielding coil for electromagneticfield (EMF) reduction of wireless power transfer in laptopcomputer,” in Proceedings of the Wireless Power TransferConference IEEE, 2014, pp. 235–238, Jeju Island, South Korea,May 2014.

[22] H. Kim, J. Song, and J. Kim, “Shielded coil structure sup-pressing leakage magnetic field from 100W-class wirelesspower transfer system with higher efficiency,” in Proceedingsof the IEEE MTT-S International Microwave Workshop Serieson Innovative Wireless Power Transmission: Technologies,Systems, and Applications, pp. 83–86, Kyoto, Japan, May 2012.

[23] International Commission on Non-Ionizing Radiation Pro-tection, “Guidelines for limiting exposure to time-varyingelectric and MFs for low frequencies (1Hz–100 kHz),” HealthPhysics, vol. 99, pp. 818–836, 2010.

[24] International Commission on Non-Ionizing Radiation Pro-tection, “Guidelines for limiting exposure to electromagneticfields (100 kHz to 300GHz),” Health Physics, vol. 118, no. 5,pp. 483–524, 2020.


electromagnetic shielding techniques in the wireless power

Documents