0% found this document useful (0 votes)
26 views6 pages

Electric Field Analysis of 11 kV Insulators

Uploaded by

JONATHAN MORALES
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
26 views6 pages

Electric Field Analysis of 11 kV Insulators

Uploaded by

JONATHAN MORALES
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

2021 lit International Conference on Power Electronics and Energy

(ICPEB-2021)

Electric Field and Potential Distribution Evaluation


of Environmentally Polluted 11 kV Polymeric
Outdoor Insulators
H.P. Shrimathi and Mithun Mondal"
Department of Electrical & Electronics Engineering
Birla Institute of Technology & Science (BITS) Hyderabad
Hyderabad, India
"mithun@[Link]
2021 1st International Conference on Power Electronics and Energy (ICPEE) | 978-1-7281-8774-7/21/$31.00 ©2021 IEEE | DOI: 10.1109/ICPEE50452.2021.9358530

Abstract-This paper analyses the potential and electric field The non-linear field distribution combined with environmental
distribution of three different variations of 11 kV polymeric insu- stresses like pollution, UV radiation etc. results in partial
lators under the clean, uniform, and non-uniform dry pollutiou discharges and flashovers, which speeds up the ageing process
conditions. The variation of the considered insulators is due to
the change in the shed's shape and width. The first insulator and degrades the polymeric material [12]. It is essential to
model is a uniform one with aU the sheds of the same width and analyse polymeric insulators' field distribution patterns under
shape. The second model terminal end sheds are of the same environmental stresses to predict its petformance.
nature, and the remaining sheds are of the same type. The third Hence, numerical methods are used to estimate the elec-
model has an alternate shed configuration type. The pollutions tric field distribution of polymeric insulators [13]. Numerical
are artIftclaI1y stimulated by varying their conductivities and
thickness. Simulations are carried out on the 2D polymeric analysis based on the finite element method is gaining its
insulator models using COMSOL Multi-physil2l software, and importance due to its ability to model complex geometries,
the variation of electric field and potential distributions w.r.t incorporating boundary conditions and a higher degree of
creepage distance are evaluated. The obtained field and potential accuracy [14]. Most of the electric field distribution studies of
distributions of the three dJfferent insulator models are compared polymer insulators focus on the effect of Uniform pollution.
to obtain the best configuration which would outperform others
in these environmentally polluted conditions. Electric field studies focusing on the effect of Non-uniform
Index 1enns-Polymeric Insulators, Profile Configurations, pollution are relatively rare.
Uniform Pollution, Non-uniform Pollution, Electric Field Com- In this paper, the potential and electric field distribution of
putation, Potential Distn"bution, COMSOL Multi-physics. three different variations of 11 kV polymeric insulator has
been analysed under clean, uniform pollution and non-uniform
I. INTRODUCTION pollution surface conditions using COMSOL Multi-physics.
Polymeric insulators are gaining importance over porce- Simulation results are compared to obtain a suitable profile
lain and glass insulators due to various advantages like less configuration that performs better under clean and pollution
weight, low installation cost, hydrophobic nature, good tensile conditions. Section II of the paper discusses modelling of
strength, and higher withstand voltage [1] [2]. However, the polymeric insulators, Section ill is dedicated to electric field
electric field distribution along the polymeric insulator surface computation. The simulation results are explained in Section
is non-linear due to the absence of intennediate metal parts IV. Discussion and conclusion are done in Section V.
that provides stress grading [3]. Factors like creepage length, II. MODELLING OF POLYMERIC INSULATOR
shed dimensions affect the electrical petformance of polymeric
insulators [4]. So, proper composite insulation design that A. Insulator Geometry and Material Properties
provides uniform potential and electric field distribution of A typical polymeric insulator shown in Fig.I consists of
composite insulators is still the area of concern. three main components:
Environmental pollution is one of the severe threats to 1) The central core made up of Fibre-Reinforced Plastic
polymeric insulators. The pollution deposit on the insulator (FRP) that provides the required mechanical strength to
varies according to the location, season , and altitude [5]. bear the load,
Generally, the contamination deposit is classified into two 2) End terminals made up of forged steel that transmits the
main types: active pollution like salt fog, fertilizers, and mechanical load to the core
inert pollution like sand and dust [6] [7]. The pollution 3) Silicone Rubber (SiR) sheath and weather sheds
deposits on the insulator surface as a solid layer and alters wrapped around the FRP core that protects the rod and
the field distribution [8] [9]. Various factors like wind and provides the required creepage distance.
surface gravity make this pollution deposit more non-uniform, Electrical properties such as relative permittivity e r and con-
which highly distorts the electric field distribution [10] [11]. ductivity (J' for different materials of polymeric insulator are

978-1-7281-8774-71111$31.00 © 2021 IEEE

Authorized licensed use limited to: Carleton University. Downloaded on June 16,2021 at [Link] UTC from IEEE Xplore. Restrictions apply.
+ -... HV TERMINAL
-++---1~ FRP CORE

SiR SHED

I+---+ SiR SHEATH

I+----.,~ SHED W IDTH

(a) Modell (b) Model 2 (c) Model 3


....++--++--..~ CREE PAGE
DI STANCE Fig. 2: Three different configurations of 11 kV polymeric
insulator (a) Model-I: uniform sheds (b) Model-2: smaller
-+--.....~ GROUND TERMI NAL middle sheds (c) Model-3: alternate sheds
Fig. 1: Structure of polymeric insulators

TABLE I: Material properties of polymeric insulator

Material Permittivity, er Conductivity, a (S/m)

Silicone rubber 4.3 1 X 10- 14


FRP rod 7.1 1 X 10- 14
Forged steel 1 5.9 X 10 7
Air 1 X 10- 15

given in Table I. The electric field analysis that has been


carried out on three different variations of 11 kV insulators are
shown in Fig. 2 and its profile parameters are given in Table II.
Here, N 1 and N 2 are the number of bigger and smaller sheds, (a) Dry uniform pollution (b) Dry non-uniform pollution
likewise, D 1 and D 2 are diameter of bigger and smaller sheds.
Fig. 3: Simulated uniform and non-uniform pollution deposited
on the insulator surface
B. Insulator Surface Condition
When atmospheric pollution like smoke, dust, sand, etc.,
settle on the insulator surface, it forms a layer that covers
The dust particles carried by wind tend to settle more on
the entire structure. To analyse the impact of dry uniform
the insulator surface which is prone to exposure, like the
contamination on 11 kV insulators, a pollution layer of 0.5
upper part of the sheds and near voltage terminals. Regions
mm thickness is modelled over the SiR housing of all three
models and it is shown in Fig. 3(a). Practically, the pollution like shaft and bottom part of sheds are moderately or less
deposit on the surface of outdoor insulators is non-uniform. polluted. To simulate the effect of this uneven distribution of
contamination, the uniform pollution layer along the creepage
length of all the models are divided into three main regions:
TABLE II: Profile Parameters of polymeric insulators High, medium and low, depending upon the pollution deposit
as shown in the Fig. 3(b). Region high is assigned to the top
Parameters (mm) Modell Model 2 Model 3 surface of the shed and shank part close it. Region medium
Creepage Length 372.72 360.88 382.06 is assigned to half of the bottom part of shed close to the tip.
Shed Number, N1/N2 4 2/4 3/2 Region low is assigned to the remaining part of the bottom
Shed Diameter, DtiD2 90 90/68 90/68
shed and shank part close to it, here the pollution deposit is less
Shed spacing 45
due to the protection by upper sheds. The electrical properties
33 33
of the pollution layer are given in Table m.

Authorized licensed use limited to: Carleton University. Downloaded on June 16,2021 at [Link] UTC from IEEE Xplore. Restrictions apply.
TABLE III: Electrical properties of pollution layer equation 3 using time step domain to obtain potential and
electric field distributions.
Pollution type Permittivity, e r Conductivity, a (S/m)

6 X 10- 7
a
-'ii'. at (E:oE:r'ii'V) - 'ii'. (O''ii'V) = 0 (4)
Uniform 7.2
Non-Uniform 7.2 4.2 X 10- 3 X 10- 7, 2 X 10- 7
7,
Where V is the potential difference, 0' is the electrical con-
ductivity, E:o is permittivity of vacuum and e; is the relative
300lij~m~8:l2~ permittivity of the dielectric material.
280 EffklSl2lSkt'
260 IV. SIMULATION RESULTS
240 DkffflasK A. Electric Field Distribution of Clean Insulator
220
200-f't'JI2tl'lS8ISk'f'k The potential and E-field distribution of clean polymeric
18 0 insulators that has been measured along the creepage distance
16 0 1t;tm~ti?i (from ground terminal to HV terminal) are shown in Fig. 5
140
and Fig. 6, respectively. The potential distribution curve of
120
10 0
mSkfiasKfi1< all the three models shows an increasing trend when moving
80
towards the HV terminal. Higher voltage gradients close to
60
U/K.I/KI /K IOKI )K
the end terminals contribute to excessive field stress. The
J'1Sk'fJ<tSkfi2t: Electric field strength of the shank region is found to be greater
40 Effffim *
than that of the weather sheds, due to the flow of electric
2: ~mm~ field in tangential direction. The flow of leakage current on
-20 DKIOKI)K.I/KI)K
_ 4 0 -=C'<.L>;'w~JOL;J.-"""C\L
the insulator surface is highly influenced by this tangential
electric field. Due to sharper edges, the field stress peaks to
Fig. 4: Insulator model with refined mesh higher values at the intersection points of shed and sheath.
The maximum magnitude of electric field stress at the shed
and sheath junction, close to the HV end fitting is found to
III. ELECTRIC FIELD COMPUTATION be 2.45 kV/cm. The field distribution pattern near the ground
end is similar to that of HV end. In the shed region, higher
In this study, a voltage of 11 kV is applied to the top field stress occurs at the tip of weather sheds. It can be seen
HV terminal and the bottom electrode is grounded, 0 V. The that the shed very near to the [Link] and the ground end is
simulation has been carried out in ACIDC module using a subjected to maximum stress compared to that of the middle
quasi-static electric current solver, which allows the user to sheds.
specify both conductivity and permittivity of the material. In All the three models has a similar electric field distribution
quasi-static, the current and electromagnetic fields vary slowly, pattern near the terminals and shank regions. However, at
which is valid for insulators operating at power frequencies middle weather sheds the distribution pattern varies due to
50-60 Hz. the difference in shed diameter and spacing. The electric field
Manually refined triangular mesh elements shown in Fig. 4 of model 1 with greater shed spacing is found to be 16.44%
is used for the entire model and the surrounding air region to less than that of the model 2 and 3 with smaller shed spacing.
enhance the computational accuracy. The electric field in larger diameter sheds are less compared
Electric field along the polymeric insulator surface is calcu- to that of sheds with smaller diameter. The decrease in electric
lated from the potential distribution by subtracting its gradient. field stress is about 18.92%.
E = -'ii'V (1) B. Electric Field Distribution Under Dry Uniform Pollution
The induced current from magnetic fields, B and electric fields, Condition
E is neglected in the computation and this approximation is The potential and E-field distribution of uniformly polluted
represented by mathematical expression, polymeric insulators are shown in Fig. 7 and Fig. 8. It can
be seen that the potential distribution of all the three models
aB (2) are more smoother compared to that of clean insulators. The
'ii' x E=-~
at presence of resistive pollution layer has reduced the electric
By applying divergence to Maxwell-Ampere's law, field stress throughout the insulator surface. The maximum
magnitude of electric field stress at the shed and sheath
v.v x H = v. + [J ~~]
=0 (3) junction, close to the HV end fitting has been reduced to
1.17 kV/cm. The field stress at the weather sheds close to
Here, the term J and an/at represents conduction current the terminals has also been reduced. Electric field strength at
density (J = O'E) and displacement current density (D = E:oE:rE) the middle part of shank regions is found to be greater than the
respectively. With Gauss's law, 'ii'.n = p the COMSOL solves shed and sheath junction points due to the pollution deposit.

Authorized licensed use limited to: Carleton University. Downloaded on June 16,2021 at [Link] UTC from IEEE Xplore. Restrictions apply.
18
- Mod el l -
- ---, -----, -- - - ,
Mo deI 2 - Mod el 3 2O,r==""",~=~~=""",=""'i---;---'
- MOd eI 1 - M OdeI 2 { MOd eI 3
18
~ 16

..
16
" ~ 14
-----t--
.."•
E 12 -- ---T -- ---,

'0 10
0 8


1l
iii
6

50 100 150 200 250 300 350 4


Creepage distance (mm) 2

0
Fig. 5: Potential Distribution w.r.t creepage distance of differ- 0 50 100 150 200 250 300 350
ent clean insulator models Creepage distance (mm)

Fig. 7: Potential distribution w.r.t creepage distance of different


uniformly polluted insulator models

'·'~PT~F:~~r:~F~~Tt~
,.,
z.e - Model 1 - Model 2 - Model 3

~:~l :.: :••••: .: ~ ••••t •••••t I


E
~ 1.841
0;
~
"'0 1.6~I + llFijJ,i il iiJlj
1.4
1.41;=========:::;-:,
1.2
- Mod el 1 - Mod el 2 - Mo del 3
u 1.2
~ t. 0
E
iii 0.8 (J 1.0
0.'
0.' ~ 0.8
0.'
0.0 .:I-J-;-L-~l.;.I~;:OJ -'-.-y.u..u,---.-'
o 50 100 150 200 250 300 350 0.6
Creepage distance (mm)
0.4
Fig. 6: Electric field distribution w.r.t creepage distance of
0.2
different clean insulator models
0.0 +-~T--r-'---;e----T~--T~--r--"';"-'
o 50 100 150 200 250 300 350

All lbe three models bas a similar electric field distribution Creepage distance (mm)
pattern near lbe terminals. However, at middle wealber sheds
and sbank region lbe distribution pattern varies according to Fig. 8: Electric field distribution w.r.t creepage distance of
lbe shed diameter and spacing. In the shank region, lbe electric different uniformly polluted insulator models
field of model 2 and 3 wilb lesser sbed spacing is found
to be 5% less than lbe model I wilb larger sbed spacing.
Under uniform pollution condition lbe decrease in electric field decreases lbe E-field magnitude near terminals, at shed and
stress by larger diameter sheds is about 61.89%. In model shealb junctions and at middle of lbe shealb region. However,
2 and model 3, lbe electric field stress of lesser diameter lbe electric field stress at lbe wealber sheds has been increased.
middle sheds exceeded lbe field strenglb of sheds close to The field strength at lbe middle part of lbe sheath region is
lbe terminals. However, in model I, the middle sheds have found to be greater lban lbe shed and sheath junction points at
lesser field strength compared to lbe sbeds closer to the end 0.5 mm and 1 mm pollution thickness. But, lbe field stress of
terminals because of uniform shed pattern. middle shealb region is less at 1.5 mm and 2 mm thickness.

C. Influence of Pollution Layer Thickness on E-jield Distribu- D. Influence of Pollution Layer Conductivity on E-field Dis-
tion tribution
To analyse lbe impact of pollution layer thickness on E-field The effect of electrical conductivity on lbe field distribution
distribution, lbe permittivity and conductivity values of lbe has been analysed by varying the conductivity of Pollution
Pollution layer in model I are kept constant and its thickness layer in model I, keeping its permittivity value constant.
is varied between 0.5 mm and 3 mm. Fig. 9 shows the E-field Simulations were carried out for three different conductivities
distribution of Model I under different pollution thicknesses. It 60-6 SIm, 6e-7 Slm, and 60-8 Slm of lbe pollution layer. Fig.
can be seen thatan increase in the thicknessof pollutionlayer, 10 shows lbe E-field distribution of Modell under different

Authorized licensed use limited to: Carleton University. Downloaded on June 16,2021 at [Link] UTC from IEEE Xplore. Restrictions apply.
20
1.4T;==============;]
- - O.5mm - - 1mm - - 2mm - - 3mm 18
1.2

..
16
E(.) 1.0 , , ~ 14
s=.
~ ~ l _

.
E 12
"C 0.' ~
'0 10
0;
'"
U 0.'
c
'~

U

'~

U ~
~
0.4 iii
iii 4

0.2 2

50 100 150 200 250 300 350


50 100 150 200 250 300 350 Creepage distance (mm)
Creepage distance (mm)
Fig. 11: Potential distribution w.r.t. creepage distance of non-
Fig. 9: Electric field distribution w.r.t creepage distance for uniformly polluted insulator models
pollution of different thickness

E. Esfield Distribution Under Dry Non-uniform Pollution Con-


dition
2.4
2.2
1 - 60-6 - 60-7 - 60-8 1-- The potential and electric field distribution of non-uniformly
polluted polymeric insulators are shown in Fig. II and Fig.
2.0
Ec 12, respectively. The potential distribution curve is similar to
s=.
1.'
the uniform pollution condition, but wilb slight increase in
1.'
magnitude. It has been found that the electric field stress peaks
"C 1.4
0;
1.2
to higher values at the regions wilb low surface conductivity,
'"c
'~ 1.0
like the ground terminal, shed and shealb junctions at the bot-
U
~ 0.'
tom part of the shed. Electric field strength at the intersection
iii 0.'
part of two different conductivities (low and medium) at the
0.4 bottom part of the shed is found to be greater than any other
0.2 field stress that occurs at various locations wilb a maximum
0.0 value of 1.64 kVIcm. The shed and sheath junctions closer
0 50 100 150 200 250 300 350 to the ground terminal has field stress of about 1.62 kV/cm.
Creepage distance (mm) The electric field at high surface conductivity regions like HV
terminal and shed and shealb junctions at the upper part of the
Fig. 10: Variation of electric field plot w.r.t creepage distance shed is found to be less than the stress observed in uniform
for pollution of different conductivities pollution condition. The shed tip in the non-uniform model
is an intersection point of the high and medium conductivity
region. The magnitude of electric field stress at the shed tips
pollution conductivities. It has been found that an increase in are lesser than the uniform pollution model, except wealber
conductivity decreases the E-field stress at terminals, shed and shed tip that is close to the HV terminal.
shealb junctions, and at middle of the shealb region. However, Similar to the uniform pollution condition, the distribution
the distribution of E-field in the shed region is non-linear for pattern at middle weather sheds and shank region varies
three different conductivities. It has been observed that the according to the shed diameter and spacing. In the shank
pollution layer wilb a conductivity of 00.6 Slm has high field region, the electric field at higher conductivity region of model
strength at all the shed regions (top and bottom of the shed, tip 2 and 3 wilb lesser shed spacing is found to be 2.3% less
of wealber shed) compared to olber two conductivities. The than the model I wilb larger shed spacing. But, at lower
pollution layer wilb a conductivity of 6007 Slm has higher conductivity region the lesser shed spacing increases the stress
field strength at the top and bottom region of middle sheds to 3%. The electric field in middle smaller diameter sheds
and shed tip close to the ground terminal. In all olber middle of model 2 and model 3 are very high compared to that of
shed tips and at the top and bottom region of shed closer to sheds wilb larger diameter in model I. The bigger sheds has
the terminals, the pollution layer wilb a conductivity of 6e-8 decreased the electric field stress in model I by 33% and 47%
Slm has higher electric field stress. compared to model 2 and model 3 respectively.

Authorized licensed use limited to: Carleton University. Downloaded on June 16,2021 at [Link] UTC from IEEE Xplore. Restrictions apply.
REFERENCES

2.0,-;============;---,
- Model 1 - M odel 2 - Model 3
[1] J. F. Hall, "History and bibliography of polymeric insulators for outdoor
applications," IEEE Transactions on Power Delivery, vol. 8, no. 1, pp.
1.8
376-385, 1993.
[2] J. S. Looms, Insulators for high voltages. lET, 1988, no. 7.

1
1.6
[3] T. Doshi, R. Gorur, and J. Hunt ''Electric field computation of composite
' 4 line insulators up to 1200 kv act" IEEE Transactions on Dielectrics and
- 1.2 Electrical Insulation, vol. 18, no. 3, pp. 861-867,2011.
[4] N. Murugan, G. Sharmi1a, and G. Kannayeram, "Design optimization
1.0
of high voltage composite insulator using electric field computations,"
0.8 Internationm conference on circuits, power and computing technologies
(ICCPCT), pp. 315-320, 2013.
0.6 [5] S. Venkataraman and R. Gomr; "Prediction of flashover voltageof non-
0.4 ceramic insulators under contaminated conditions," IEEE Transactions
on Dielectrics and Electrical Insulation, vol. 13, no. 4, pp. 862--869,
2006.
0.0 o 50 100 150 200 250 300 350
[6] H. Su, Z. Jla, Z. Guan, andL. Li, ''MechaniBm of contaminant accumu-
lationand flashover of insulator in heavily polluted coastal area," IEEE
Creepage distance (mm)
Transactions on Dielectrics and Electrical Insulation, vol. 17, no. 5, pp.
1635-1641, 2010.
[7] M. I. Qureshi et al., ''Effect of contamination on the leakage current
Fig. 12: Electric field distribution w.r.t creepage distance of of inland desertinsulators," IEEE Transactions on Electrical Insulation,
non-uniformly polluted insulator models no. 4, pp. 332-339. 1984.
[8] E.-S. M. El-Refaie, M. Abel Elrahman, and M. K. Mohamed, ''Electric
field distribution of optimized composite insulator profiles under differ-
ent pollutionconditions," Ain Shams Engineering Journal, vol. 9, no. 4,
pp. 1349-1356, 2018.
V. DISCUSSIONS & CONCLUSION [9] A. Nekahi, S. McMeekin, M. Farzaoeh et aI., "Effect of pollution
severity on electric field distribution along a polymeric insulator," 11th
The potential and electric field distribution of three different international IEEE conference on the properties and applications of
variations of 11 kV polymeric insulators has been analysed dielectric materials (ICPADM), pp. 612--615, 2015.
under the clean, uniform, and non-uniform pollution condi- [10] [Link], A. Ahmed, andH. Singer,"Electricfield distortion caused.
by asymmetric pollution on insulator surfaces," IEEE transactions on
tions using the COMSOL Multi-physics software. The electric dielectrics and electrical insulation. vol. 6. no. 2, pp. 175-180. 1999.
field stress along the insulator surface under the non-uniform [11] Z. Zhang, J. You,D. Wei,X. Jiang, D. Zhang, andM. Bi, "Investigations
pollution condition is observed to be higher than that of the on ac pollutionfiashover performance of insulator string under different
non-uniform pollution conditions," IEr Genuation, Transmission d:
uniform pollution. In all conditions, the electric field stress Distribution, vol. 10, no. 2, pp. 437-443, 2016.
is found to be maximum at the intersection point of shed [12] R. Chakraborty and B. S. Reddy, "Studies on high temperature vul-
and sheath close to the terminals. Insulator shanks and end canized silicone robber insulators under arid climatic aging," IEEE
Transactions on Dielectrics and Electrical Insulation, vol. 24, no. 3,
fittings are highly stressed than the weather sheds. In all three pp. 1751-1760, 2017.
clean insulator models, the sheds close to the end fittings have [13] B. Zhang, S. Han, J. He, R. Zeng, and P. Zhu, "Numerical analysis
a higher electric field than the middle sheds. However, the of electric-field distribution around composite insulator and head of
transmission tower," IEEE Transactions on power delivery, vol. 21,
field stress in model 2 and model 3 is found to be greater no. 2, pp. 959-')65, 2006.
than that of model 1. Under the pollution conditions, field [14] J. Li, Z. Peng, and X. Yans, "Potential calculation and grading ring
stress at the smaller sheds exceeds the field strength of sheds design for ceramic insulators in 1000 kv uhv substations," IEEE Trans-
actions on Dielectrics and Ekctricallnsulation, vol. 19, no. 2, pp. 723-
closer to the terminals. The uniform resistive pollution layer on 732,2012.
the insulator surface smoothen the potential and electric field
distribution. The non-uniform pollution distributes the field
stress according to the conductivity of surface pollution. The
low conductivity regions are highly stressed as compared to
the high conductivity regions. The field strength at the junction
point of low and high conductivity region on the shank is
higher than the terminals' stress. The smaller shed spacing in
model 2 and model 3 decreases the uniformly polluted shank
region's field. stress. In non-uniform pollution conditions, the
stress at the high conductivity region has also been reduced.
However, the field stress at the low conductivity region and
the clean surface condition has been increased. Although the
higher electric field stress values of smaller diameter sheds are
not critical under dry pollution conditions, when the insulator
becomes wet, this higher stress and the leakage current could
initiate surface discharge and damage the insulator. Model
I, with uniform sheds and larger shed spacing, shows better
performance as compared to others.

Authorized licensed use limited to: Carleton University. Downloaded on June 16,2021 at [Link] UTC from IEEE Xplore. Restrictions apply.

You might also like