Method for Calculating Coupling Coefficients in

Dynamic Energy Transfer for Electric Vehicles

John Linden1, Yasha Nikulshin 1, Shuki Wolfus1, Hanan Rumbak2, Oren Ezer2, Yosi Yeshurun1
1
Institute of Superconductivity, Department of Physics, Bar-Ilan University, Ramat-Gan, Israel
2
ElectRoad Inc., Caesarea, Israel

Abstract—Optimizing the efficiency of primary and secondary coil I. CURRENT TECHNOLOGICAL CHALLENGES
configurations for Dynamic Wireless Power Transfer (DWPT) in
Electric Vehicles (EVs) requires means for accurate calculation of
Although it seems that technology is in a fast-moving
the mutual inductance in an array of coils. Based on finite element direction towards achieving efficient, game-changing EVs, there
simulation method, we present a quick and accurate method for are still several problems that must be solved.
calculating the energy transfer capabilities of a given DWPT These include:
array. By consecutively switching on and off every coil and • Ensuring electrical safety in handling high power
driving them with a constant current ramp rate, mutual lines, especially in extreme weather
inductances and a coupling coefficient matrix of the whole • Increasing the speed of recharging
configuration is easily calculated. This method allows for
relatively easy optimization and up-scaling of DWPT systems to
• Decreasing the EV system's weight
multiple arrays of primary and secondary coils as well as Wireless energy transfer is an emerging technology that
implementing various coil designs and configurations. The data presents a solution to these problems [1], [4] and [5].
acquired using this method may also be used during real-time The DWPT solution is well based and is presented as one of
applications providing indications of vehicle relative alignment. the most promising technologies for EV solutions [6].
However, in order for this technology to be adopted, the
I. INTRODUCTION optimization process needs to overcome cost and efficiency
As electric vehicles (EVs) are becoming more popular, challenges.
several technologies are competing to make their imprint in the Several methods for calculating optimization parameters for
market. Each EV technology has its advantages and DWPT systems have been proposed [7] and [8], focusing
disadvantages that may promote their use for different mainly on the calculation of the magnetic flux density, where
applications, optimizing their efficiency. the coils are chosen with common geometrical topologies in
A comprehensive assessment of modern inductive power order to ease the analysis [9].
transfer systems for transportation may be found in [1]. We propose to focus on the coupling parameter between the
ElectRoad Ltd. [2] is one of the few firms that are working on primary and secondary coils, while also being able to calculate
developing a system based on Dynamic Wireless Power the self-inductance of each coil with relative ease, unreliant on
Transfer (DWPT) [3]. using common geometrical coil topologies.
The DWPT technology is predominantly dependent on the It is apparent that the coupling of the coils is not the sole
magnetic coupling achieved between primary transmitting coils parameter optimized to achieve a cost efficient product.
within the road, and secondary coils situated under the vehicle However, using this parameter and applying the simple
chassis. method of calculation we describe below, more complex and
This makes optimization of the coupling between the coils a innovative designs can be tested with ease to determine the
decisive factor in the efficiency of the technology and its efficiency of the coupling factors and allow for a relatively
ultimate success in implementation. quick and simple optimization process.
It is the purpose of this article to introduce a method that
allows simulating the coupling coefficients between multiple
arrays of coils, regardless of their topology, yielding the
essential information needed to determine the design's power
transfer capabilities.
The method also takes into consideration the exact location of
the coils and effects of other structures/materials in the vicinity.

978-1-5386-2382-4/17/$31.00 ©2017 IEEE

EV2017 www.evshow.ro

II. MUTUAL INDUCTANCE AND COUPLING COEFFICIENTS

Mutual inductance between two coils is defined by equation
(1):
di1
L 2,1 = V2 / (1)
dt
where V2 is the induced voltage on the secondary coil due to the
change in current i1 in the primary coil.
Figure 1. FEM model as simulated. Includes 12 primary coils below (orange),
Note that the equation (1) is independent of any spatial 9 interlapping secondary coils above (light blue) and a ferrite plate on top.
constrictions and considers the flux leakage that can be caused
by various system configurations.
The amount of mutual inductance between coupled coils is
expressed as a fractional number between zero (no mutual
inductance/loosely coupled) and 1 (maximum mutual
inductance/tight coupling).
This parameter for two coils is given as:
L 2, 1
K 2,1 = (2)
L1, 1 ⋅ L 2,2 Figure 2. Magnetic Field lines as simulated within the wireless power transfer
structure. Top coils are to be embedded in chassis with ferrite plate above,
where L1, 1 and L2, 2 are the coils’ self-inductance. while bottom coils are embedded in the road.
In an array of multiple coils, we can calculate all the coupling In order to calculate the inductance matrix of such an array,
coefficients between each coil as: finite element method simulation software, COBHAM OPERA
^ L i, j was used.
K = ¦i =1 ¦
N N
(3) To simplify the simulation process, the calculation of the
i =1
L j, jLi, i induced voltage in the receiving coils and the transformation to
the inductance matrix of the entire array, we have used a linear
where N is the sum of the number of coils in the total system current ramp in every coil and recorded the induced voltage in
(primary and secondary combined). all other coils.
Thus, from a matrix representation of L (calculated or A constant din/dt is fed to each coil consecutively while
measured) we get a symmetrical matrix of the coupling switching on and off the other coil circuits. A schematic
coefficients where the main diagonal Ki,i (self-coupling example for each coil switching circuit is displayed in figure 3.
parameter) equals 1.

§ L1, 1 " L N, 1 · § K1, 1 " K N, 1 ·

¨ ¸ ¨ ¸
¨ # % # ¸→¨ # % # ¸ (4)
¨L ¸ ¨ ¸
© 1, N " L N, N ¹ © K1, N " K N, N ¹
III. SIMULATION METHOD Figure 3. Simulated electrical circuit for each coil.

A specific DWPT design may consist of a complex array of

transmitting and receiving coils, metallic components and
shields, and magnetic flux concentrators and diverters.
An example of such a possible design is shown in figure 1.
The figure displays an array of 12 transmission coils
(orange), which represents a short section from a long
underground transmission track. The receiving coils array (light
blue in the figure 1), is embedded in the EV's chassis above the
transmitting track.
The array of secondary coils in the figure is composed of 9
coils at different height levels. Also shown here is a ferrite plate
mounted above the receiving array.
Each receiving coil captures a portion of the magnetic flux
generated by the primary coils below. The supporting
electronics integrate the induced voltage in the receiving array
to maximize energy transfer.
Figure 2 displays the magnetic layout of such a design along
with the magnetic field lines obtained when only the center coil
in figure 2 is excited.
Figure 4. Switch Data Logic compared with current sources during simulation
in two circuits. Each circuit has a time step to calculate the self-inductance Nn, n
and the coupling inductance Ln, m.
EV2017 www.evshow.ro

Figure 4 demonstrates the principle of coil switching in the

simulation. Only two coils are presented here but the coil which
is switched off represents all coils except for the excited one. In
this example, S1 is first switched on for a period of 1 μsec
during which the current is linearly ramped up to a value of 1 A
and back down to zero.
During this time S2 (and all other switches) are in the off
position, preventing current flow in the associated coils. Voltage
is measured on every coil in the “off” state and the mutual
inductance is then calculated for each pair of coils. For the ‫ܮ‬ேǡே
matrix, N2 simulation steps are required. After the inductance
matrix is calculated, a simple numerical analysis is needed to
compute the ‫ܭ‬ ෡ coupling coefficients for every coil in the system
as presented in figure 5.
The top left square in white presents primary windings 1-12, Figure 6. Coupling coefficients are presented with a colormap scale during
where the bottom and right are secondary windings 13-21. vehicle approach (60 cm before maximum coupling matching).
Coefficients are calculated for the configuration as presented
in figure 1 (Cross-diagonal represents Ki, i = 1 but is set to zero
to allow proper viewing of the scaling).

Figure 7. Coil configuration for displacement of an approaching vehicle

(same as in figure 6).

Through these figures is it easy to visualize the advantage of

using a system of multiple arrays of primary and secondary
coils.
As the vehicle approaches/leaves an array segment, the
previously strongly coupled coils #14 and #2 show a decrease in
there coupling factors and have been replaced in magnitude in
figure 6 with #15 and #1 respectively.
A sensory device for vehicle alignment would be able to
Figure 5. Coupling coefficients are presented with a colormap scale. Cross- consecutively switch to transferring the power primarily
diagonal value is set to zero to allow proper viewing of the scaling through the most strongly coupled coils, thus providing a more
Clearly, the coupling coefficient matrix of figure 5 is efficient power transfer system with less radiative losses.
position dependent. As the secondary receiving coils pass over The method of sensing vehicle alignment may be done as
the transmission track, the matrix changes according to the well by constantly measuring the coupling coefficients and
momentary coil coupling. The matrix of figure 5 is, therefore, a alerting the vehicle of the proper corrections to be made.
snapshot of the specific state described in figure 1 and figure 2. Constant measurement of the coefficients can be very
Here, the secondary coils 14, 17 and 20 are situated just straightforward when using the already known waveforms and
above primary coils 2, 6 and 10 respectively. Therefore, the frequencies used for the wireless power transfer, these values
substituted in equations (1) and (2) produce the desired ‫ܭ‬ ෡
coefficient matrix shows maximum coupling exactly for these
pairs of coils. parameters.
In the specific application where importance is placed on
dynamic coupling coefficients due to the vehicle motion,
multiple simulations can provide the coupling information to
determine the desired switching of the primary coils to be the
most power efficient.
Such analysis could make use of the most strongly coupled
coils rather than the whole array which consists of a broad range
of coupling coefficients.
Figures 6 - 9 show the results of displacements in the vehicle
alignments.
Figures 6 - 7 model the expected motion of the vehicle as it
approaches a new set of primary coils, while figures 8 - 9
present a misalignment of the vehicle traveling off of the
primary coil’s main axis.
Figure 8. Coupling coefficients are presented with a colormap scale during
vehicle misalignment (20 cm off of maximum coupling matching axis).
EV2017 www.evshow.ro

ACKNOWLEDGMENT
J. Linden acknowledges a fellowship granted by the Israeli
Ministry of Science allowing his active participation in this
project.
REFERENCES
[1] Grant Anthony Covic, Senior Member, IEEE, and Talbot Boys "Modern
Trends in Inductive Power Transfer for Transportation Applications,"
IEEE Journal of Emerging and Selected Topics in Power Electronics, vol.
Figure 9. Coil configuration for misalignment of the vehicle (same as in 1, no. 1, pp. 28-41, 2013.
figure 8).
[2] "Electroad," [Online]. Available: https://www.electroad.me/.
Figures 8-9 demonstrate the case of a misaligned vehicle
[3] Rumbak, Hanan, "System and method for powering on-road electric
moving at a distance of 20 cm from the center of the primary vehicles via wireless power transfer". Patent US20160339785 A1, 24
array (similar to not driving exactly in the middle of a lane). November 2016.
This may be considered a common event, either while switching [4] Dr. Aviva Brecher, Mr. David Arthur, "Review and Evaluation of Wireless
lanes or due to other driver activity. Power Transfer for Electric Transit Applications," Federal Transit
This case provides an example when the coupling between Administration, US Department of Transportation, August 2014.
two neighboring secondary or primary coils far outdoes the [5] Keshri, Ravikiran Vaka and Ritesh Kumar, "Review on Contactless Power
coupling strength between any set of secondary and primary Transfer for Electric Vehicle Charging," Energies, vol. 10, no. 636, pp. 2-
coils. In figure 8 this is shown by consecutive coils #13 and #14 20, 2017.
having a high coupling value compared to other pairs. [6] Leandros A. Maglaras, Jianmin Jiang, Athanasios Maglaras, Frangiskos V.
As this study presents, this coupling factor is in direct relation Topalis, Sotiris Moschoyiannis, "Dynamic wireless charging of electric
vehicles on the move with Mobile Energy Disseminators," International
to the energy transfer between coils, therefore any undesired Journal of Advanced Computer Science and Applications, 2014.
coupling between coils may lead to a loss in efficiency or other
[7] Jesús Sallán, Member, IEEE, Juan L. Villa, Andrés Llombart, and José
problems in the application’s performance.
Fco. Sanz, "Optimal Design of ICPT Systems Applied to Electric Vehicle
This analysis shows the capabilities of identifying the exact Battery Charge," IEEE Transactions on Industrial Electronics, vol. 56, no.
points when events like these may occur and will provide 6, pp. 2140-2149, 2009.
valuable information to limit undesired results. [8] Vaez-Zadeh, Saeed Hasanzadeh, and Sadegh, "Design of a Wireless Power
Several solutions may be optimizing the sizes and distanced Transfer System for High Power Moving Applications," Progress In
between consecutive coils to achieve the desired results. Electromagnetics Research, vol. 28, pp. 258-271, 2013.
[9] Adeel Zaheer, Hao Hao, Grant A. Covic, Senior Member, IEEE, and
IV. CONCLUSIONS Dariusz Kacprzak, "Investigation of Multiple Decoupled Coil Primary Pad
In this paper, various examples were presented to prove the Topologies in Lumped IPT Systems for Interoperable Electric Vehicle
Charging," IEEE Transactions on Power Electronics, vol. 30, no. 4, pp.
method of concept for the analysis of wireless power transfer 1937-1955, 2015.
configurations with relative ease and short simulation times.
The goal in mind is to enhance the process needed to come
up with new and better topologies and allow relatively easy
optimization and up-scaling the system to multiple arrays of
primary and secondary coils.