Ci a ion: Kakouche, K.; Rekioua, T.;
Mezani, S.; Oubelaid, A.; Rekioua, D.;
Blazek, V.; P okop, L.; Misak, S.; Bajaj,
M.; Ghoneim, S.S.M. Model
P edic i e Di ec To que Con ol and
Fuzzy Logic Ene gy Managemen o
Mul i Powe Sou ce Elec ic Vehicles.
Senso s 2022,22, 5669. h ps://
doi.o g/10.3390/s22155669
Academic Edi o : Omp akash
Kaiwa ya
Recei ed: 27 June 2022
Accep ed: 26 July 2022
Published: 28 July 2022
Publishe ’s No e: MDPI s ays neu al
wi h ega d o ju isdic ional claims in
published maps and ins i u ional a il-
ia ions.
Copy igh : © 2022 by he au ho s.
Licensee MDPI, Basel, Swi ze land.
This a icle is an open access a icle
dis ibu ed unde he e ms and
condi ions o he C ea i e Commons
A ibu ion (CC BY) license (h ps://
c ea i ecommons.o g/licenses/by/
4.0/).
senso s
A icle
Model P edic i e Di ec To que Con ol and Fuzzy Logic
Ene gy Managemen o Mul i Powe Sou ce Elec ic Vehicles
Khoudi Kakouche 1, Tou ik Rekioua 1, Smail Mezani 2, Adel Oubelaid 1, Djamila Rekioua 1,
Voj ech Blazek 3,* , Lukas P okop 3, S anisla Misak 3, Mohi Bajaj 4,5 and She i S. M. Ghoneim 6
1Labo a oi e de Technologie Indus ielle e de l’In o ma ion, Facul éde Technologie, Uni e si éde Bejaia,
Bejaia 06000, Alge ia; khoudi [email p o ec ed] (K.K.); [email p o ec ed] (T.R.);
[email p o ec ed] (A.O.); [email p o ec ed] (D.R.)
2Uni e si éde Lo aine, GREEN, F-54000 Nancy, F ance; [email p o ec ed]
3ENET Cen e, VSB—Technical Uni e si y o Os a a, 708 00 Os a a, Czech Republic;
[email p o ec ed] (L.P.); s anisla [email p o ec ed] (S.M.)
4Depa men o Elec ical Enginee ing, Na ional Ins i u e o Technology, Delhi 110040, India;
[email p o ec ed]
5Depa men o Elec ical Enginee ing, G aphic E a (Deemed o be Uni e si y), Deh adun 248002, India
6Depa men o Elec ical Enginee ing, College o Enginee ing, Tai Uni e si y, P.O. Box 11099,
Tai 21944, Saudi A abia; [email p o ec ed]
*Co espondence: [email p o ec ed]
Abs ac :
This pape p oposes a no el Fuzzy-MPDTC con ol applied o a uel cell ba e y elec ic
ehicle whose ac ion is ensu ed using a pe manen magne synch onous mo o (PMSM). On he
ac ion side, model p edic i e di ec o que con ol (MPDTC) is used o con ol PMSM o que, and
gua an ee minimum o que and cu en ipples while ensu ing sa is ac o y speed acking. On he
sou ces side, an ene gy managemen s a egy (EMS) based on uzzy logic is p oposed, i aims o
dis ibu e powe o e ene gy sou ces a ionally and sa is y he load powe demand. To assess hese
echniques, a d i ing cycle unde di e en ope a ing modes, namely c uising, accele a ion, idling
and egene a i e b aking is p oposed. Real- ime simula ion is de eloped using he RT LAB pla o m
and he ob ained esul s ma ch hose ob ained in nume ical simula ion using MATLAB/Simulink.
The esul s show a good pe o mance o he whole sys em, whe e he p oposed MPDTC minimized
he o que and lux ipples wi h 54.54% and 77%, espec i ely, compa ed o he con en ional DTC
and educed he THD o he PMSM cu en wi h 53.37%. Fu he mo e, he p oposed EMS based
on uzzy logic shows good pe o mance and keeps he ba e y SOC wi hin sa e limi s unde he
p oposed speed p o ile and in e na ional NYCC d i ing cycle. These a o emen ioned esul s con i m
he obus ness and e ec i eness o he p oposed con ol echniques.
Keywo ds:
uzzy logic; model p edic i e di ec o que con ol; uel cell; ba e y; pe manen magne
synch onous mo o ; elec ic ehicle
1. In oduc ion
The use o elec ic ehicles (EVs) in he anspo a ion sec o has become a necessi y
in he las decade o deal wi h he ene gy c isis and en i onmen al pollu ion p oblems,
as hey o e many ad an ages such as high e iciency, no ca bon dioxide emissions, low
main enance, and no pollu ion [
1
]. EVs a e mainly composed o uel cells, supe capaci o s,
and ba e ies o he ene gy supply and s o age pa , as well as an elec ic mo o o he
ac ion pa . As i is known, achie ing he bes pe o mance EVs, equi es an adap ed EMS,
o e ec i ely egula e he low o powe be ween he di e en sou ces [
2
]. High-e iciency
elec ic mo o s assis ed by high-accu acy con olle s a e also s ill needed [3].
A p o on exchange memb ane uel cell (PEMFC) is a p omising sou ce used o powe
ehicles due o i s small size, low ope a ing empe a u e, as s a -up, and high e i-
ciency [
4
]. To o e come he d awback o slow dynamic esponse and o implemen ene gy
Senso s 2022,22, 5669. h ps://doi.o g/10.3390/s22155669 h ps://www.mdpi.com/jou nal/senso s
Senso s 2022,22, 5669 2 o 22
eco e y, i is necessa y o in eg a e ene gy s o age sou ces, such as a supe capaci o and/o
a Li-ion ba e y [
5
,
6
]. EMS is e y impo an o manage ene gy alloca ion e ec i ely, and
i s choice should be deeply in es iga ed. In his con ex , se e al me hods in he ecen
li e a u e ha e been s udied and e alua ed, such as op imiza ion me hods, il e -based
me hods, con olle me hods, and ule-based me hods. Op imiza ion-based s a egies ha
include Model P edic i e Con ol [
7
], G ey Wol Op imize [
8
], Pa icle Swa m Op imiza-
ion [
9
], e c, ha e been in es iga ed in o de o deal wi h complex managemen objec i es
(e iciency, cos , li e ime, e c). On he o he hand, hese s a egies a e complex and impose
an impo an compu a ion bu den. Fu he il e -based EMSs can be ound in he li e a u e,
including Low-Pass il e s [
10
], and Wa ele T ans o m [
11
]. The il e -based managemen
s a egy aims o di ide he equi ed powe in o high- and low- equency powe as his
s a egy can imp o e he li e ime o he PEMFC s ack. Ne e heless, he pe o mance o
he wo EMSs depends s ongly on he design o he il e which is a complica ed ask
o be pe o med. Con olle me hods such as Backs epping [
12
], Passi i y Con ol [
13
],
P opo ional-In eg al Con ol [
14
], Sliding Con ol [
15
], e c, can ob ain an exac calcula ion
o he e e ence while aking in o accoun he losses o he sys em. Rule-based s a egies
a e mainly Fuzzy Logic, A i icial Neu al Ne wo k [
16
], and S a e Machine [
17
]. Fuzzy
logic-based EMS is widely used in uel cell hyb id powe sys ems [18–22]. This echnique
has he ad an age o p o ide high pe o mance and suppo ing imp ecise sys em modeling.
Fo hese easons, his echnique is adop ed in he cu en wo k.
Pe manen magne synch onous mo o s a e widely conside ed he bes ype o elec ic
mo o s ha can be used o d i e EVs. This is due p ima ily o hei high e iciency, high
powe densi y, ligh weigh , and wide speed ange [
23
]. Wi h p ope o que con ol, hey
can easily mee all o he ehicle equi emen s. Many o que con ol s a egies, such as
Field-O ien ed Con ol (FOC) and Di ec To que Con ol (DTC), ha e been ex ensi ely
esea ched in he li e a u e [
24
,
25
]. DTC has a simple con ol s uc u e and p o ides a as
dynamic o que esponse [
26
]. Howe e , because i is based on hys e esis compa a o s,
his echnique has some d awbacks such as impo an o que, lux ipples, and a iable
swi ching equency. Many me hods ha e been p oposed o mi iga e hese d awbacks.
In [
27
], he au ho s p oposed a s a egy named Space Vec o Modula ion DTC (SVM-
DTC) based on a cons an swi ching equency, and he esul s showed good pe o mance.
Howe e , his s a egy equi es a p ecise design o he PI con olle and sys em pa ame e s.
In [
28
], a mul ile el in e e was used, which inc eases he numbe o ol age ec o s.
The simula ion esul s indica ed good pe o mances by educing o que and lux ipples.
Ne e heless, issues wi h complexi y and swi ching losses appea ed. Fuzzy logic and
a i icial neu al ne wo ks a e also used as a i icial in elligence con olle s [
29
,
30
]. Au ho s
in [
31
–
34
] ha e p oposed a Model P edic i e Di ec To que Con ol s a egy ha is based
on p edic ing he con ol a iables such as lux and o que while minimizing he e o
o he p edic ed con ol a iables; his imp o es con ol accu acy while main aining he
con ol sys em’s simplici y.
To cope wi h he a o emen ioned challenges and imp o e EV, wo con ol s a egies
a e p oposed in his wo k. The i s one is based on uzzy logic, applied o he Li-ion
ba e y-PEMFC hyb id powe sys em. The second one is based on o que p edic i e con ol
applied o he PMSM. Resul s ob ained using Ma lab/Simulink and hose ob ained using a
RT LAB simula o ha e clea ly shown he e ec i eness o he p oposed con ol echniques
unde di e en d i ing modes (c uising, accele a ion, idling, and egene a i e b aking). In
o de o p ope ly si ua e his wo k, he main con ibu ions made h ough his pape a e:
•
An adequa e EMS s a egy based on uzzy logic con ol is de eloped o ensu e ehicle
p opulsion powe and o egula e e icien ly he ene gy low o he powe sou ces.
•
A model p edic i e di ec o que con ol s a egy is p oposed o con ol he ehicle
ac ion machine wi h he objec i e o minimizing o que and lux ipples and ensu ing
sa is ac o y speed acking.
•
A de ailed physical model o he EV ( he ehicle dynamics sys em, he elec ic powe
sys em, and con ol sys em) is es ablished unde Ma lab/Simulink en i onmen .
Senso s 2022,22, 5669 3 o 22
•
Real- ime simula ion using he RT LAB pla o m is pe o med o con i m he ob ained
simula ion esul s.
The esul s ob ained using Ma lab/Simulink as well as he expe imen al ones ob-
ained using he RT LAB simula o a e p esen ed, and he main conclusion o his wo k
summa izes and p o es p oposed s a egies.
2. Elec ic Vehicle Desc ip ion and Modeling
The EV gene al con igu a ion is shown in Figu e 1. I consis s o a hyd ogen ank wi h
a low egula o , a PEMFC s ack as a p ima y powe sou ce, a Li-ion ba e y as a seconda y
powe sou ce, a PMSM, and a con ol sys em. The en i e VE sys em’s cons uc ion can be
di ided in o wo pa s:
Senso s2022,22,56693o 23


 Anadequa eEMSs a egybasedon uzzylogiccon olisde eloped oensu e ehi‐
clep opulsionpowe and o egula ee icien ly heene gy lowo  hepowe sou ces.
 Amodelp edic i edi ec  o quecon ols a egyisp oposed ocon ol he ehicle
ac ionmachinewi h heobjec i eo minimizing o queand lux ipplesandensu ‐
ingsa is ac o yspeed acking.
 Ade ailedphysicalmodelo  heEV( he ehicledynamicssys em, heelec icpowe 
sys em,andcon olsys em)ises ablishedunde Ma lab/Simulinken i onmen .
 Real‐ imesimula ionusing heRTLABpla o mispe o med ocon i m heob‐
ainedsimula ion esul s.
The esul sob ainedusingMa lab/Simulinkaswellas heexpe imen alonesob‐
ainedusing heRTLABsimula o a ep esen ed,and hemainconclusiono  hiswo k
summa izesandp o esp oposeds a egies.
2.Elec icVehicleDesc ip ionandModeling
TheEVgene alcon igu a ionisshowninFigu e1.I consis so ahyd ogen ank
wi ha low egula o ,aPEMFCs ackasap ima ypowe sou ce,aLi‐ionba e yasa
seconda ypowe sou ce,aPMSM,andacon olsys em.Theen i eVEsys em’scons uc‐
ioncanbedi idedin o wopa s:
On hesou cesside, he low a e egula o adjus s hepowe o  hePEMFCs ack
and egula es hep essu eo  hehyd ogen low.ThePEMFCs ackisconnec ed o heDC
bus ol age iaaunidi ec ionalin‐cu en DC‐DCboos con e e  oDCbus ol age.The
excessi eelec ici ygene a edby hePEMFCs ackisused ocha ge heba e y.TheLi‐
ionba e yisconnec ed o heDCbus ol age iaabidi ec ionalDC‐DCbuck‐boos con‐
e e  o eco e  heb akingene gyandsupplypowe .
On he ac ionside,PMSMo 50kW a edpowe , edbya wo‐le elin e e ,con‐
e s heelec icpowe coming om he wopowe sou cesin omechanicalpowe .
Thecon olsys emin eg a es hecon olo bo h ac ionmachineandpowe 
sou ces.

Figu e1.Elec ic ehiclecon igu a ion.

Figu e 1. Elec ic ehicle con igu a ion.
On he sou ces side, he low a e egula o adjus s he powe o he PEMFC s ack
and egula es he p essu e o he hyd ogen low. The PEMFC s ack is connec ed o he DC
bus ol age ia a unidi ec ional in-cu en DC-DC boos con e e o DC bus ol age. The
excessi e elec ici y gene a ed by he PEMFC s ack is used o cha ge he ba e y. The Li-ion
ba e y is connec ed o he DC bus ol age ia a bidi ec ional DC-DC buck-boos con e e
o eco e he b aking ene gy and supply powe .
On he ac ion side, PMSM o 50 kW a ed powe , ed by a wo-le el in e e , con e s
he elec ic powe coming om he wo powe sou ces in o mechanical powe .
The con ol sys em in eg a es he con ol o bo h ac ion machine and powe sou ces.
2.1. Fuel-Cell Model
The ol age o he PEMFC s ack VFC is gi en by [18,35]:
VFC =E−RIFC (1)
E=EOC −NA lnIFC
i0·1
sTd
3+1(2)
whe e
VFC
and
IFC
a e he ol age and cu en o he PEMFC s ack, espec i ely,
R
is he
in e nal esis ance,
EOC
is he open ci cui ol age,
i0
is he exchange cu en ,
N
,
A
and
Td
a e he cells numbe , he a el slope, and he esponse ime, espec i ely.
EOC
,
i0
and
A
a e
gi en by:
EOC =KCEN(3)
Senso s 2022,22, 5669 4 o 22
i0=zFkPH2+PO2
Rh e−∆G
RT (4)
A=RT
zαF(5)
whe e
KC
and
EN
a e he ol age cons an a nominal condi ion o ope a ion and Ne ns
ol age espec i ely,
z
is he numbe o mo ing elec ons
(z=2)
,
F
,
R
,
k
,
h
and
T
a e he
Fa aday’s cons an , he ideal gas cons an , he Bol zmann’s cons an , he Planck’s cons an
and he empe a u e o ope a ion espec i ely,
∆G
is he ac i a ion ene gy ba ie , and
α
is
he cha ge ans e coe icien .
The PEMFC s ack includes hyd ogen con olle and oxygen con olle , which egula e
he lows a e o
H2
and
O2
, espec i ely. The u iliza ion a es o
H2
and
O2
a e calcula ed as:
U H2=60000RTNIFC
zFP uelV uel x%(6)
U o2=60000RTNIFC
2zFPai Vai y%(7)
whe e
P uel
and
Pai
a e he absolu e supply p essu e o uel and absolu e ai , espec i ely,
V uel
and
Vai
a e he uel low a e and he ai low a e, espec i ely,
x
% and
y
% a e he
pe cen age o H2in he uel and O2in he oxidan , espec i ely.
The pa ame e s o he used PEMFC s ack a e gi en in Table 1.
Table 1. PEMFC s ack pa ame e s.
Pa ame e Value Uni e
Nominal powe 50 kW
Peak powe 60 kW
Numbe o cells 358 Cell
Nominal s ack e iciency 55 %
Ope a ing empe a u e 65 ◦C
Nominal Ai low a e 2100 Ipm
Fuel supply p essu e 1.5 ba
Ai supply p essu e 1 ba
2.2. Ba e y Model
Li-ion ba e ies a e used in his wo k due o hei high ene gy densi y, high e iciency,
and long li e ime when compa ed o o he ba e y ypes such as (NiCd, lead-acid, o
NiMH) [36].
The Li-ion ba e y ol age can be calcula ed using wo di e en equa ions [36,37].
Vdischa ge=E0−R·i−KQ
Q−i ·(i +i∗) + Aexp(−B·i )(8)
Vcha ge=E0−R·i−KQ
i −0.1Q·i∗−KQ
Q−i ·i +Aexp(−B·i )(9)
whe e
R
,
K
,
Q
,
E0
,
i
,
i∗
,
A
and
B
a e he Li-ion ba e y in e nal esis ance, he pola iza ion
cons an , he Li-ion ba e y capaci y, he Li-ion ba e y cons an ol age, he ac ual Li-ion
ba e y cha ge, he il e ed Li-ion ba e y cu en , he exponen ial zone ampli ude, and he
exponen ial zone ime cons an in e se, espec i ely.
The Li-ion ba e y s a e o cha ge can be de e mined using Equa ion (10).
SOCba =1001−Zi( )d
Q(10)
Senso s 2022,22, 5669 5 o 22
2.3. Pe manen Magne Synch onous Mo o Model
The ma hema ical model o PMSM in he d-q o o e e ence ame can be exp essed as
ollows in Equa ions (11) and (12) [23]:
Vsd =RsIsd +dφsd
d +ωφsq (11)
Vsq =RsIsq +dφsq
d +ωφsd (12)
whe e he o alized lux φsd and φsq a e gi en by:
φsd =Lsd Isd +φ (13)
φsq =Lsq Isq (14)
The elec omagne ic o que exp ession is gi en by Equa ion (15):
Te=3
2pIsq(Lsd −Lsd)Isd +φ (15)
The PMSM mechanical equa ion is gi en by:
JdΩ
d =Te−T − Ω(16)
The es ima ions o he o que, lux, and he load angle can be exp essed by he
ollowing se o equa ions:
∧
Te=3
2pIsqφsd −Isdφsq(17)
∧
φs=q(φsd)2+φsq2(18)
∧
θ= an−1φsq
φsd
(19)
Mo o pa ame e s [38] a e summa ized in Table 2.
Table 2. PMSM pa ame e s.
Pa ame e Value Uni e
Ra ed powe (P )50 kW
DC ol age (Vdc )500 V
S a o esis ance (Rs)0.0065 Ω
S a o induc ance (Ld,Lq)8.35 mH
PM magne ic lux (φ )0.17566143 Wb
Numbe o pole pai s (p) 4 -
Mo o ine ia (J) 0.089 kg·m2
Viscous damping ( ) 0.005 N·m·s
2.4. Vehicle Dynamics Sys em
The dynamic model o he elec ical ehicle is depic ed in Figu e 2. The en i e mechan-
ical pa (longi udinal ehicle dynamics, iscous ic ion, di e en ial, i es, and educ ion
gea ) o he ehicle dynamic sys em is modeled by Souleman Njoya Mo apon and Louis-A.
Dessain [
35
]. The longi udinal ehicle dynamics block akes in o accoun body mass,
ae odynamic d ag, and weigh dis ibu ion be ween axles. Meanwhile, wind speed and
oad inclina ion a e no conside ed in his model.

Senso s 2022,22, 5669 6 o 22
Senso s2022,22,56696o 23


Table2.PMSMpa ame e s.
Pa ame e ValueUni e
Ra edpowe (𝑷𝒓)50kW
DC ol age(𝑽𝒅𝒄)500V
S a o  esis ance(𝑹𝒔)0.0065Ω
S a o induc ance(𝑳𝒅,𝑳𝒒)8.35mH
PMmagne ic lux(𝝓
𝒇
)0.17566143Wb
Numbe o polepai s(p)4‐
Mo o ine ia(J)0.089kg∙m
2

Viscousdamping( )0.005N∙m∙s
2.4.VehicleDynamicsSys em
Thedynamicmodelo  heelec ical ehicleisdepic edinFigu e2.Theen i eme‐
chanicalpa (longi udinal ehicledynamics, iscous ic ion,di e en ial, i es,and e‐
duc iongea )o  he ehicledynamicsys emismodeledbySoulemanNjoyaMo aponand
Louis‐A.Dessain [35].Thelongi udinal ehicledynamicsblock akesin oaccoun body
mass,ae odynamicd ag,andweigh dis ibu ionbe weenaxles.Meanwhile,windspeed
and oadinclina iona eno conside edin hismodel.
Thepa ame e so  heused ehicle[38]a egi eninTable3.
Table3.Elec ic ehiclepa ame e s.
Pa ame e ValueUni e
Vehicle o almass󰇛𝑴󰇜 1325kg
Gea  a ioo  he inald i e󰇛𝑮󰇜 5.2‐
Numbe o wheelspe axle2‐
F on ala ea(
𝑨
𝒇
)2.57m
2

Ti e adius( )0.3m
D agcoe icien (𝑪𝒅)0.3‐

Figu e2.Vehicledynamicssys em.
3.Sys emCon ol
3.1.Ene gyManagemen S a egy
Tosa is y heloadpowe demandand oensu eane icien powe dis ibu iono 
heelec ic ehiclepowe sys em,anapp op ia eEMSis equi ed[2].Theseobjec i escan
onlybeme bycon olling hepowe  esponseo eachene gysou ceacco ding o heload
Figu e 2. Vehicle dynamics sys em.
The pa ame e s o he used ehicle [38] a e gi en in Table 3.
Table 3. Elec ic ehicle pa ame e s.
Pa ame e Value Uni e
Vehicle o al mass (M)1325 kg
Gea a io o he inal d i e (G)5.2 -
Numbe o wheels pe axle 2 -
F on al a ea (A )2.57 m2
Ti e adius ( ) 0.3 m
D ag coe icien (Cd)0.3 -
3. Sys em Con ol
3.1. Ene gy Managemen S a egy
To sa is y he load powe demand and o ensu e an e icien powe dis ibu ion o
he elec ic ehicle powe sys em, an app op ia e EMS is equi ed [
2
]. These objec i es
can only be me by con olling he powe esponse o each ene gy sou ce acco ding o he
load demand. We used a uzzy logic con ol-based EMS because i is lexible, e icien , and
wo ks well wi hou exac ma hema ical models.
3.1.1. Inpu and Ou pu Pa ame e s
The inpu s pa ame e s o he uzzy logic con olle a e he Li-ion ba e y s a e-o -
cha ge SOC and he load powe (Pload) ob ained by mul iplying he mo o speed and he
equi ed mo o o que, and he ou pu pa ame e is he e e ence powe o he uel cell, as
illus a ed in Figu e 3.
Senso s2022,22,56697o 23


demand.Weuseda uzzylogiccon ol‐basedEMSbecausei is lexible,e icien ,and
wo kswellwi hou exac ma hema icalmodels.
3.1.1.Inpu andOu pu Pa ame e s
Theinpu spa ame e so  he uzzylogiccon olle a e heLi‐ionba e ys a e‐o ‐
cha geSOCand heloadpowe (Pload)ob ainedbymul iplying hemo o speedand he
equi edmo o  o que,and heou pu pa ame e is he e e encepowe o  he uelcell,
asillus a edinFigu e3.
The uzzyse  o Li‐ionba e ySOCisdi idedin o“Low”(L),“Medium”(M),and
“High”(H).The uzzyse  o Ploadisalsodi idedin o“Nega i e”(N),“Ve yLow”(VL),
L,M,H,“Ve yHigh”(VH),indica ing hepowe demand omlow ohighle els.The
uzzyse  o P cisclassi iedas“Ze o”(ZE),VL,L,M,H,VH.Thegapbe ween(ZE)and
(VL)o P cinFigu e4cis he uelcellsys em’slow‐e iciencyzone(including hecooling
an,humidi ie ,ando he accesso ies), he e o e hePEMFCs ackshouldno ope a ein
his ange.The iangula and apezoidalmembe ship unc ions(MFs)a eusedin his
caseasshowninFigu e4.Thechoiceo  he iangula and apezoidalshapesis o educe
he a ia ionsin hegene a edpowe  e e ence.

Figu e3.Theblockdiag amo  he uzzylogic‐basedene gymanagemen s a egy.

(a)(b)

(c)
µ (SOC)
µ (P c)
Figu e 3. The block diag am o he uzzy logic-based ene gy managemen s a egy.
Senso s 2022,22, 5669 7 o 22
The uzzy se o Li-ion ba e y SOC is di ided in o “Low” (L), “Medium” (M), and
“High” (H). The uzzy se o Pload is also di ided in o “Nega i e” (N), “Ve y Low” (VL),
L, M, H, “Ve y High” (VH), indica ing he powe demand om low o high le els. The
uzzy se o P c is classi ied as “Ze o” (ZE), VL, L, M, H, VH. The gap be ween (ZE) and
(VL) o P c in Figu e 4c is he uel cell sys em’s low-e iciency zone (including he cooling
an, humidi ie , and o he accesso ies), he e o e he PEMFC s ack should no ope a e in
his ange. The iangula and apezoidal membe ship unc ions (MFs) a e used in his
case as shown in Figu e 4. The choice o he iangula and apezoidal shapes is o educe
he a ia ions in he gene a ed powe e e ence.
Senso s2022,22,56697o 23


demand.Weuseda uzzylogiccon ol‐basedEMSbecausei is lexible,e icien ,and
wo kswellwi hou exac ma hema icalmodels.
3.1.1.Inpu andOu pu Pa ame e s
Theinpu spa ame e so  he uzzylogiccon olle a e heLi‐ionba e ys a e‐o ‐
cha geSOCand heloadpowe (Pload)ob ainedbymul iplying hemo o speedand he
equi edmo o  o que,and heou pu pa ame e is he e e encepowe o  he uelcell,
asillus a edinFigu e3.
The uzzyse  o Li‐ionba e ySOCisdi idedin o“Low”(L),“Medium”(M),and
“High”(H).The uzzyse  o Ploadisalsodi idedin o“Nega i e”(N),“Ve yLow”(VL),
L,M,H,“Ve yHigh”(VH),indica ing hepowe demand omlow ohighle els.The
uzzyse  o P cisclassi iedas“Ze o”(ZE),VL,L,M,H,VH.Thegapbe ween(ZE)and
(VL)o P cinFigu e4cis he uelcellsys em’slow‐e iciencyzone(including hecooling
an,humidi ie ,ando he accesso ies), he e o e hePEMFCs ackshouldno ope a ein
his ange.The iangula and apezoidalmembe ship unc ions(MFs)a eusedin his
caseasshowninFigu e4.Thechoiceo  he iangula and apezoidalshapesis o educe
he a ia ionsin hegene a edpowe  e e ence.

Figu e3.Theblockdiag amo  he uzzylogic‐basedene gymanagemen s a egy.

(a)(b)

(c)
µ (SOC)
µ (P c)
Figu e 4.
Membe ship unc ions: (
a
) inpu pa ame e (SOC); (
b
) inpu pa ame e (Pload); (
c
) ou pu
pa ame e (P c).
3.1.2. Fuzzy In e ence Rules
The uzzy logic ules co esponding o his ene gy managemen a e designed and
p esen ed in Table 4. The choice o he ules is acco ding o he desi ed ope a ion on he
PEMFC s ack. Fo example, when he SOC o he Li-ion ba e y is low and he powe
demanded by he ehicle is e y high, hen he powe ha he PEMFC s ack mus p o ide
will be e y high. In he same way, he PEMFC s ack p o ides a powe ha is lowe han
he powe demand when he Li-ion ba e y is highly cha ged because in his case, he Li-ion
ba e y mus be solici ed o educe i o a medium cha ge in o de o ake ad an age o he
ene gy eco e y om b aking. Based on his gene al idea, he choice o hese ules emains
a bi a y acco ding o he desi ed unc ioning, while espec ing he esponse ime o he
wo sou ces. The SOC o he Li-ion ba e y should be main ained be ween 30% and 80%.
This p ocedu e p e en s deep discha ging and o e cha ging, which can educe he li espan
o a Li-ion ba e y. The uzzy ules su ace is shown in Figu e 5.
Senso s 2022,22, 5669 8 o 22
Table 4. Ene gy managemen uzzy logic ules.
P c
Pload
N VL L M H VH
SOC
LZE L M H VH VH
MZE ZE VL L M H
HZE ZE ZE VL VL L
Senso s2022,22,56698o 23


Figu e4.Membe ship unc ions:(a)inpu pa ame e (SOC);(b)inpu pa ame e (Pload);
(c)ou pu pa ame e (P c).
3.1.2.FuzzyIn e enceRules
The uzzylogic ulesco esponding o hisene gymanagemen a edesignedand
p esen edinTable4.Thechoiceo  he ulesisacco ding o hedesi edope a ionon he
PEMFCs ack.Fo example,when heSOCo  heLi‐ionba e yislowand hepowe de‐
mandedby he ehicleis e yhigh, hen hepowe  ha  hePEMFCs ackmus p o ide
willbe e yhigh.In hesameway, hePEMFCs ackp o idesapowe  ha islowe  han
he
powe demandwhen heLi‐ionba e yishighlycha gedbecausein
hiscase, heLi‐ionba e ymus besolici ed o educei  oamedium
cha ge
ino de  o akead an ageo  heene gy eco e y omb aking.Basedon hisgene al
idea, hechoiceo  hese ules emainsa bi a yacco ding o hedesi ed unc ioning,
while espec ing he esponse imeo  he wosou ces.TheSOCo  heLi‐ionba e y
shouldbemain ainedbe ween30%and80%.Thisp ocedu ep e en sdeepdischa ging
ando e cha ging,whichcan educe heli espano aLi‐ionba e y.The uzzy ulessu ‐
aceisshowninFigu e5.
Thep oposed uzzylogiccon olle uses heMamdaniin e encep ocedu e,wi h he
cen oidme hod o de uzzi ica ion[36].
Table4.Ene gymanagemen  uzzylogic ules.
P
c
P
load

NVLLMHVH
SOC
LZELMHVHVH
MZEZEVLLMH
HZEZEZEVLVLL

Figu e5.Fuzzylogiccon olsu ace.
3.2.DCBusVol ageRegula ionandPEMFCS ackCon e e Con ol
Theelec icalene gysou cesused osupply heEVmus bewellcon olled ia he
con e e sby egula ing hei cu en sand/o ou pu  ol ages.Thecon e e sconnec ed
o heene gysou cescon ol heou pu powe and ol age[39].The egula iono  heDC
bus ol ageand hecon olo  hePEMFCs ackpowe a ep esen edinFigu e6.PEMFC
s ackandLi‐ionba e ya e, espec i ely,connec ed o heDCbus iaunidi ec ionaland
bidi ec ionalDC‐DCcon e e s.Thepowe managemen blockgene a es he e e ence
cu en 𝐼
∗whichhas obelimi edinaslopeino de  o espec  hecons ain s ela ed
o hePEMFCs ackdynamics.APIcon olle isused ocon ol hePEMFCs ackpowe 
byadjus ing hecu en  oi s e e ence alue𝐼
∗.TheLi‐ionba e y egula es heDCbus
Figu e 5. Fuzzy logic con ol su ace.
The p oposed uzzy logic con olle uses he Mamdani in e ence p ocedu e, wi h he
cen oid me hod o de uzzi ica ion [36].
3.2. DC Bus Vol age Regula ion and PEMFC S ack Con e e Con ol
The elec ical ene gy sou ces used o supply he EV mus be well con olled ia he
con e e s by egula ing hei cu en s and/o ou pu ol ages. The con e e s connec ed
o he ene gy sou ces con ol he ou pu powe and ol age [
39
]. The egula ion o he DC
bus ol age and he con ol o he PEMFC s ack powe a e p esen ed in Figu e 6. PEMFC
s ack and Li-ion ba e y a e, espec i ely, connec ed o he DC bus ia unidi ec ional and
bidi ec ional DC-DC con e e s. The powe managemen block gene a es he e e ence
cu en
I∗
FC
which has o be limi ed in a slope in o de o espec he cons ain s ela ed o
he PEMFC s ack dynamics. A PI con olle is used o con ol he PEMFC s ack powe by
adjus ing he cu en o i s e e ence alue
I∗
FC
. The Li-ion ba e y egula es he DC bus
ol age by acking he e e ence ol age
VDC_ e
. A double PI egula ion loop is used o
main ain he DC bus ol age close o i s e e ence and o con ol he Li-ion ba e y powe .
Senso s 2022,22, 5669 9 o 22
Senso s2022,22,56699o 23


ol ageby acking he e e ence ol age𝑉_.AdoublePI egula ionloopisused o
main ain heDCbus ol ageclose oi s e e enceand ocon ol heLi‐ionba e ypowe .

Figu e6.DCbus ol age egula ion,Li‐ionBa e ycon e e con ol,andPEMFCs ackcon e e 
con ol.
3.3.ModelP edic i eDi ec To queCon ol
Themodelp edic i edi ec  o quecon ols a egybasicp incipleis op edic  he
sys em u u ebeha io o e  imeusing hePMSMmodel[33].Thiss a egyisused o
con olbo h he luxand o queo  hePMSMin heEVsys em.Thenume icalimplemen‐
a iono  heMPDTCalgo i hm o PMSMin heEVcanbedi idedin o wos eps.S ep1
is op edic  hecon olled a iables,and heseconds epis oselec  he ol age ec o  o
beappliedin henex sampling ime.Tode e mine hebes  ol age ec o  ousein he
nex sampling ime,acos  unc ioniscons uc ed.Theop imal ol age ec o ischosen
basedon heobjec i eswi h heminimume o , esul ingin educed ipples.InMPDTC,
hecon olled a iablesa ep edic edusingFo wa dEule app oxima ion[40].
Theblockdiag amand he lowcha o MPDTCa eillus a edinFigu es7and8.

Figu e7.Thep oposedMPDTCscheme.
Figu e 6.
DC bus ol age egula ion, Li-ion Ba e y con e e con ol, and PEMFC s ack con e e con ol.
3.3. Model P edic i e Di ec To que Con ol
The model p edic i e di ec o que con ol s a egy basic p inciple is o p edic he sys-
em u u e beha io o e ime using he PMSM model [
33
]. This s a egy is used o con ol
bo h he lux and o que o he PMSM in he EV sys em. The nume ical implemen a ion
o he MPDTC algo i hm o PMSM in he EV can be di ided in o wo s eps. S ep 1 is o
p edic he con olled a iables, and he second s ep is o selec he ol age ec o o be
applied in he nex sampling ime. To de e mine he bes ol age ec o o use in he nex
sampling ime, a cos unc ion is cons uc ed. The op imal ol age ec o is chosen based
on he objec i es wi h he minimum e o , esul ing in educed ipples. In MPDTC, he
con olled a iables a e p edic ed using Fo wa d Eule app oxima ion [40].
The block diag am and he lowcha o MPDTC a e illus a ed in Figu es 7and 8.
Senso s2022,22,56699o 23


ol ageby acking he e e ence ol age𝑉_.AdoublePI egula ionloopisused o
main ain heDCbus ol ageclose oi s e e enceand ocon ol heLi‐ionba e ypowe .

Figu e6.DCbus ol age egula ion,Li‐ionBa e ycon e e con ol,andPEMFCs ackcon e e 
con ol.
3.3.ModelP edic i eDi ec To queCon ol
Themodelp edic i edi ec  o quecon ols a egybasicp incipleis op edic  he
sys em u u ebeha io o e  imeusing hePMSMmodel[33].Thiss a egyisused o
con olbo h he luxand o queo  hePMSMin heEVsys em.Thenume icalimplemen‐
a iono  heMPDTCalgo i hm o PMSMin heEVcanbedi idedin o wos eps.S ep1
is op edic  hecon olled a iables,and heseconds epis oselec  he ol age ec o  o
beappliedin henex sampling ime.Tode e mine hebes  ol age ec o  ousein he
nex sampling ime,acos  unc ioniscons uc ed.Theop imal ol age ec o ischosen
basedon heobjec i eswi h heminimume o , esul ingin educed ipples.InMPDTC,
hecon olled a iablesa ep edic edusingFo wa dEule app oxima ion[40].
Theblockdiag amand he lowcha o MPDTCa eillus a edinFigu es7and8.

Figu e7.Thep oposedMPDTCscheme.
Figu e 7. The p oposed MPDTC scheme.
Senso s 2022,22, 5669 16 o 22
Figu es 15 and 16 clea ly show ha he EMS based on uzzy logic ensu es a con enien
powe low and gi es good pe o mance wi h he changes in d i ing condi ions, and
ba e y s a es o cha ge.
D i ing Cycle Tes
The New Yo k Ci y Cycle (NYCC) d i ing cycle is adop ed o analyze he pe o mance
o he p oposed EMS based on uzzy logic. Th ee di e en scena ios a e de ined o es he
pe o mance o EV unde di e en s a es o cha ge o he Li-ion ba e y. The h ee scena ios
a e as ollows:
Scena io 1: he ini ial SOC o he Li-ion ba e y is 60% when s a ing.
Scena io 2: he ini ial SOC o he Li-ion ba e y a s a up is 80%.
Scena io 3: he ini ial SOC o he Li-ion ba e y is 30%.
Figu e 17 shows he powe cu es o he PEMFC s ack, Li-ion ba e y, mo o , and
Li-ion ba e y SOC in scena io 1. I can be seen in Figu e 17a, ha he PEMFC s ack does no
ope a e in he low powe ange [0–5 kW] and becomes ac i e in he high powe demand
whe e he e iciency o he PEMFC sys em is ela i ely high. The Li-ion ba e y is con igu ed
o p o ide all he cha ging powe in he low powe demand ange, assis he uel cell in he
high powe demand ange, and abso b he b aking ene gy when he ehicle decele a es.
The inal SOC alue o he Li-ion ba e y in his scena io is 59.5%, as seen in Figu e 17a.
Senso s 2022, 22, 5669 17 o 23
(a)
(b)
Figu e 17. The pe o mance o he EV unde NYCC d i ing cycle in Scena io 1: (a) powe cu es;
(b) Li-ion ba e y SOC cu e.
(a)
(b)
Figu e 18. The pe o mance o he EV unde NYCC d i ing cycle in Scena io 2: (a) powe cu es;
(b) Li-ion ba e y SOC cu e.
(a)
(b)
Figu e 19. The pe o mance o he EV unde NYCC d i ing cycle in Scena io 3: (a) powe cu es;
(b) Li-ion ba e y SOC cu e.
Figu e 17.
The pe o mance o he EV unde NYCC d i ing cycle in Scena io 1: (
a
) powe cu es;
(b) Li-ion ba e y SOC cu e.
Figu e 18 shows he powe cu es and he SOC o he Li-ion ba e y in he second
scena io. As seen in Figu e 18a, he Li-ion ba e y is egula ed o p o ide a la ge amoun o
powe o p o ec i om o e cha ging, and he PEMFC s ack in e enes in he high powe
demand. The Li-ion ba e y SOC dec eases om 80% o 74% as shown in Figu e 18b.
Senso s 2022, 22, 5669 17 o 23
(a)
(b)
Figu e 17. The pe o mance o he EV unde NYCC d i ing cycle in Scena io 1: (a) powe cu es;
(b) Li-ion ba e y SOC cu e.
(a)
(b)
Figu e 18. The pe o mance o he EV unde NYCC d i ing cycle in Scena io 2: (a) powe cu es;
(b) Li-ion ba e y SOC cu e.
(a)
(b)
Figu e 19. The pe o mance o he EV unde NYCC d i ing cycle in Scena io 3: (a) powe cu es;
(b) Li-ion ba e y SOC cu e.
Figu e 18.
The pe o mance o he EV unde NYCC d i ing cycle in Scena io 2: (
a
) powe cu es;
(b) Li-ion ba e y SOC cu e.

Senso s 2022,22, 5669 17 o 22
Figu e 19a shows he powe cu es in he hi d scena io. The uel cell is ope a ed o
p o ide powe o he load and cha ge he Li-ion ba e y o p o ec i om deep discha ges.
The Li-ion ba e y SOC inc eases om 30% o 46.5%, as shown in Figu e 19b.
Senso s 2022, 22, 5669 17 o 23
(a)
(b)
Figu e 17. The pe o mance o he EV unde NYCC d i ing cycle in Scena io 1: (a) powe cu es;
(b) Li-ion ba e y SOC cu e.
(a)
(b)
Figu e 18. The pe o mance o he EV unde NYCC d i ing cycle in Scena io 2: (a) powe cu es;
(b) Li-ion ba e y SOC cu e.
(a)
(b)
Figu e 19. The pe o mance o he EV unde NYCC d i ing cycle in Scena io 3: (a) powe cu es;
(b) Li-ion ba e y SOC cu e.
Figu e 19.
The pe o mance o he EV unde NYCC d i ing cycle in Scena io 3: (
a
) powe cu es;
(b) Li-ion ba e y SOC cu e.
The esul s show ha he p oposed ene gy managemen s a egy based on uzzy
logic is obus and could imp o e he e iciency o he PEMFC s ack, mainly because his
p oposed EMS makes he PEMFC s ack ope a e in he high-e iciency egion.
5. Real-Time Pla o m Using RT-LAB
The eal- ime simula ion o he p oposed con ol is implemen ed in his sec ion using
he disc e e eal- ime simula o , he RT LAB pla o m. Figu e 20a depic s he eal- ime
simula ion bench se up in he LTII labo a o y o Bejaia, which consis s o he ollowing
componen s: (1) a hos PC, (2) an OP5700 eal- ime digi al simula o , (3) a HIL con olle
and da a acquisi ion in e ace OP8660, and (4) a digi al oscilloscope. As i is shown in
Figu e 20b, he i s s ep owa d eal- ime simula ion is he model sepa a ion. The EV
sys em is spli in o compu a ion and console blocks. Blocks ha con ain compu a ions
such as ene gy managemen s a egy, MPDTC, ehicle dynamics, mo o and powe sou ces
models a e placed on he compu a ion subsys em, which is cons i u ed o a mas e (SM)
bloc highligh ed wi h ed dashed lines and a sla e (SS) block. Scopes and cons an s a e
placed in he console block. In RT-LAB, each compu a ion subsys em is assigned o a
di e en co e. In o he wo ds, each subsys em is coded in C and buil o execu ion among
i s p ocesso s using Ma hwo ks code gene a o Real-Time-Wo kshop (RTW) [
43
]. A e
compila ion, he code is loaded in o a ge OP5700 ia TCP/IP p o ocol and execu ed ia
pa allel p ocessing. Finally, all obse a ions in display blocks a e moni o ed and displayed
on he digi al oscilloscope ia I/O channels.
A eal- ime simula ion was pe o med a e he EV simula ion sys em was decomposed
and adap ed o use in he RT LAB pla o m.
In o de o e alua e he wo echniques p oposed in his wo k, namely he MPDTC on
he mo o side, and he EMS based on uzzy logic on he sou ce side, a d i ing cycle unde
di e en ope a ing modes has been applied (Figu e 9). Figu es 21–24 p esen expe imen al
esul s o he wo echniques p oposed. The esul s ob ained using a disc e e eal- ime
simula o RT LAB a e e y close o he simula ion esul s wi h he same ema ks p e iously
men ioned in he simula ion esul s. Expe imen al esul s p o e he e ec i eness o he
p oposed con ol echniques.
Senso s 2022,22, 5669 18 o 22
Senso s2022,22,566918o 23


5.Real‐TimePla o mUsingRT‐LAB
The eal‐ imesimula iono  hep oposedcon olisimplemen edin hissec ionusing
hedisc e e eal‐ imesimula o , heRTLABpla o m.Figu e20adepic s he eal‐ ime
simula ionbenchse upin heLTIIlabo a o yo Bejaia,whichconsis so  he ollowing
componen s:(1)ahos PC,(2)anOP5700 eal‐ imedigi alsimula o ,(3)aHILcon olle 
andda aacquisi ionin e aceOP8660,and(4)adigi aloscilloscope.Asi isshownin
Figu e20b, he i s s ep owa d eal‐ imesimula ionis hemodelsepa a ion.TheEVsys‐
emisspli in ocompu a ionandconsoleblocks.Blocks ha con aincompu a ionssuch
asene gymanagemen s a egy,MPDTC, ehicledynamics,mo o andpowe sou ces
modelsa eplacedon hecompu a ionsubsys em,whichiscons i u edo amas e (SM)
blochighligh edwi h eddashedlinesandasla e(SS)block.Scopesandcons an sa e
placedin heconsoleblock.InRT‐LAB,eachcompu a ionsubsys emisassigned oadi ‐
e en co e.Ino he wo ds,eachsubsys emiscodedinCandbuil  o execu ionamong
i sp ocesso susingMa hwo kscodegene a o Real‐Time‐Wo kshop(RTW)[43].A e 
compila ion, hecodeisloadedin o a ge OP5700 iaTCP/IPp o ocolandexecu ed ia
pa allelp ocessing.Finally,allobse a ionsindisplayblocksa emoni o edanddis‐
playedon hedigi aloscilloscope iaI/Ochannels.
A eal‐ imesimula ionwaspe o meda e  heEVsimula ionsys emwas
decomposedandadap ed o usein heRTLABpla o m.


(a)(b)
Figu e20.(a)Expe imen alse upo RT‐labpla o ma LTIIlabo a o y;(b)RT‐labsys em
a chi ec u e.
Ino de  oe alua e he wo echniquesp oposedin hiswo k,namely heMPDTC
on hemo o side,and heEMSbasedon uzzylogicon hesou ceside,ad i ingcycle
unde di e en ope a ingmodeshasbeenapplied(Figu e9).Figu es21–24p esen ex‐
pe imen al esul so  he wo echniquesp oposed.The esul sob ainedusingadisc e e
eal‐ imesimula o RTLABa e e yclose o hesimula ion esul swi h hesame ema ks
p e iouslymen ionedin hesimula ion esul s.Expe imen al esul sp o e hee ec i e‐
nesso  hep oposedcon ol echniques.
Figu e 20.
(
a
) Expe imen al se up o RT-lab pla o m a LTII labo a o y; (
b
) RT-lab sys em a chi ec u e.
Senso s 2022, 22, 5669 19 o 23
(a)
(b)
(c)
(d)
Figu e 21. Expe imen al esul s o he EV ac ion chain unde di e en d i ing modes
using he MPDTC s a egy (a) S a o cu en ; (b) Vehicle speed esponse; (c) Elec omag-
ne ic o que; (d) S a o lux.
(a)
(b)
Figu e 22. Expe imen al wa e o m unde di e en d i ing modes (a) Ba e y s a e o cha ge; (b) DC
bus ol age.
60 (A)
1.6 (s)
8.8 (Km/h)
1.6 (s)
Ca speed
50 (N.m)
1.6 (s)
T
e
0.05 (Wb)
1.6 (s)
Flux
20 (%)
1.6 (s)
SOC
200 (V)
1.6 (s)
DC bus ol age
Figu e 21.
Expe imen al esul s o he EV ac ion chain unde di e en d i ing modes using he MPDTC
s a egy (a) S a o cu en ; (b) Vehicle speed esponse; (c) Elec omagne ic o que; (d) S a o lux.
Senso s 2022,22, 5669 19 o 22
Senso s 2022, 22, 5669 19 o 23
(a)
(b)
(c)
(d)
Figu e 21. Expe imen al esul s o he EV ac ion chain unde di e en d i ing modes
using he MPDTC s a egy (a) S a o cu en ; (b) Vehicle speed esponse; (c) Elec omag-
ne ic o que; (d) S a o lux.
(a)
(b)
Figu e 22. Expe imen al wa e o m unde di e en d i ing modes (a) Ba e y s a e o cha ge; (b) DC
bus ol age.
60 (A)
1.6 (s)
8.8 (Km/h)
1.6 (s)
Ca speed
50 (N.m)
1.6 (s)
T
e
0.05 (Wb)
1.6 (s)
Flux
20 (%)
1.6 (s)
SOC
200 (V)
1.6 (s)
DC bus ol age
Figu e 22.
Expe imen al wa e o m unde di e en d i ing modes (
a
) Ba e y s a e o cha ge; (
b
) DC
bus ol age.
Senso s 2022, 22, 5669 20 o 23
Figu e 23. Expe imen al wa e o m o he powe managemen o he elec ic ehicle unde di e en
d i ing modes.
(a)
(b)
Figu e 24. Expe imen al wa e o m o he pe o mance o he elec ic ehicle unde di e en s a e
o cha ge (a) SOC = 80%; (b) SOC = 30%.
6. Conclusions
In his pape , he au ho s ocused on imp o ing he pe o mance o an EV by in o-
ducing a Fuzzy-MPDTC-based con ol. This con ol is di ided in o wo pa s: The i s
pa is dedica ed o he con ol o he ac ion machine by in oducing a MPDTC echnique
o PMSM con ol. The second pa p oposes a uzzy logic-based EMS o he elec ic e-
hicle powe sys em. To e alua e hese echniques, a d i ing cycle unde di e en ope a -
ing modes has been p oposed. The main conclusions a e lis ed as ollows. On he mo o
side, se e al objec i es a e achie ed by applying a p ede ined cos unc ion, he elec o-
magne ic o que esponse ollows i s e e ence wi h small ipple alues, a 54.54% im-
p o emen compa ed o he classical DTC; he cu en ipple is educed and he e e ence
acking is ensu ed. On he sou ces side, he uzzy logic-based EMS p o ides obus pe -
o mance unde a ious ba e y s a es o cha ge and apid a ia ion in powe demand.
Real- ime simula ion esul s we e ob ained using a disc e e eal- ime simula o ; he RT
LAB pla o m con i med he e ec i eness and obus ness o he p oposed con ol ech-
niques.
Au ho Con ibu ions: Concep ualiza ion, K.K. and T.R.; me hodology, K.K., T.R., M.B. and S.M.
(Smail Mezani); so wa e, K.K.; alida ion, K.K., T.R., M.B. and A.O.; o mal analysis, K.K., T.R.,
S.M. (Smail Mezani), A.O. and D.R.; in es iga ion, K.K., T.R., S.M. (Smail Mezani), A.O. and D.R.;
esou ces, K.K. and T.R.; da a cu a ion, K.K., T.R., S.M. (Smail Mezani), A.O. and D.R.; w i ing—
o iginal d a p epa a ion, K.K., M.B.; w i ing— e iew and edi ing, K.K., T.R., S.M. (Smail Mezani),
A.O., L.P., V.B., M.B., S.M. (S anisla Misak), S.S.M.G. and D.R.; isualiza ion, K.K., T.R. and S.M.
16 (kW)
1.6 (s)
Mo o powe
Ba e y powe
Fuel Cell powe
16 (kW)
1.6 (s)
Mo o powe
Ba e y powe
Fuel Cell powe
16 (kW)
1.6 (s)
Mo o powe
Ba e y powe
Fuel Cell powe
Figu e 23.
Expe imen al wa e o m o he powe managemen o he elec ic ehicle unde di e en
d i ing modes.
Senso s 2022, 22, 5669 20 o 23
Figu e 23. Expe imen al wa e o m o he powe managemen o he elec ic ehicle unde di e en
d i ing modes.
(a)
(b)
Figu e 24. Expe imen al wa e o m o he pe o mance o he elec ic ehicle unde di e en s a e
o cha ge (a) SOC = 80%; (b) SOC = 30%.
6. Conclusions
In his pape , he au ho s ocused on imp o ing he pe o mance o an EV by in o-
ducing a Fuzzy-MPDTC-based con ol. This con ol is di ided in o wo pa s: The i s
pa is dedica ed o he con ol o he ac ion machine by in oducing a MPDTC echnique
o PMSM con ol. The second pa p oposes a uzzy logic-based EMS o he elec ic e-
hicle powe sys em. To e alua e hese echniques, a d i ing cycle unde di e en ope a -
ing modes has been p oposed. The main conclusions a e lis ed as ollows. On he mo o
side, se e al objec i es a e achie ed by applying a p ede ined cos unc ion, he elec o-
magne ic o que esponse ollows i s e e ence wi h small ipple alues, a 54.54% im-
p o emen compa ed o he classical DTC; he cu en ipple is educed and he e e ence
acking is ensu ed. On he sou ces side, he uzzy logic-based EMS p o ides obus pe -
o mance unde a ious ba e y s a es o cha ge and apid a ia ion in powe demand.
Real- ime simula ion esul s we e ob ained using a disc e e eal- ime simula o ; he RT
LAB pla o m con i med he e ec i eness and obus ness o he p oposed con ol ech-
niques.
Au ho Con ibu ions: Concep ualiza ion, K.K. and T.R.; me hodology, K.K., T.R., M.B. and S.M.
(Smail Mezani); so wa e, K.K.; alida ion, K.K., T.R., M.B. and A.O.; o mal analysis, K.K., T.R.,
S.M. (Smail Mezani), A.O. and D.R.; in es iga ion, K.K., T.R., S.M. (Smail Mezani), A.O. and D.R.;
esou ces, K.K. and T.R.; da a cu a ion, K.K., T.R., S.M. (Smail Mezani), A.O. and D.R.; w i ing—
o iginal d a p epa a ion, K.K., M.B.; w i ing— e iew and edi ing, K.K., T.R., S.M. (Smail Mezani),
A.O., L.P., V.B., M.B., S.M. (S anisla Misak), S.S.M.G. and D.R.; isualiza ion, K.K., T.R. and S.M.
16 (kW)
1.6 (s)
Mo o powe
Ba e y powe
Fuel Cell powe
16 (kW)
1.6 (s)
Mo o powe
Ba e y powe
Fuel Cell powe
16 (kW)
1.6 (s)
Mo o powe
Ba e y powe
Fuel Cell powe
Figu e 24.
Expe imen al wa e o m o he pe o mance o he elec ic ehicle unde di e en s a e o
cha ge (a) SOC = 80%; (b) SOC = 30%.
Senso s 2022,22, 5669 20 o 22
6. Conclusions
In his pape , he au ho s ocused on imp o ing he pe o mance o an EV by in oduc-
ing a Fuzzy-MPDTC-based con ol. This con ol is di ided in o wo pa s: The i s pa is
dedica ed o he con ol o he ac ion machine by in oducing a MPDTC echnique o
PMSM con ol. The second pa p oposes a uzzy logic-based EMS o he elec ic ehicle
powe sys em. To e alua e hese echniques, a d i ing cycle unde di e en ope a ing
modes has been p oposed. The main conclusions a e lis ed as ollows. On he mo o side,
se e al objec i es a e achie ed by applying a p ede ined cos unc ion, he elec omagne ic
o que esponse ollows i s e e ence wi h small ipple alues, a 54.54% imp o emen
compa ed o he classical DTC; he cu en ipple is educed and he e e ence acking
is ensu ed. On he sou ces side, he uzzy logic-based EMS p o ides obus pe o mance
unde a ious ba e y s a es o cha ge and apid a ia ion in powe demand. Real- ime
simula ion esul s we e ob ained using a disc e e eal- ime simula o ; he RT LAB pla o m
con i med he e ec i eness and obus ness o he p oposed con ol echniques.
Au ho Con ibu ions:
Concep ualiza ion, K.K. and T.R.; me hodology, K.K., T.R., M.B. and S.M.
(Smail Mezani); so wa e, K.K.; alida ion, K.K., T.R., M.B. and A.O.; o mal analysis, K.K., T.R., S.M.
(Smail Mezani), A.O. and D.R.; in es iga ion, K.K., T.R., S.M. (Smail Mezani), A.O. and D.R.; esou ces,
K.K. and T.R.; da a cu a ion, K.K., T.R., S.M. (Smail Mezani), A.O. and D.R.; w i ing—o iginal d a
p epa a ion, K.K., M.B.; w i ing— e iew and edi ing, K.K., T.R., S.M. (Smail Mezani), A.O., L.P., V.B.,
M.B., S.M. (S anisla Misak), S.S.M.G. and D.R.; isualiza ion, K.K., T.R. and S.M. (Smail Mezani);
supe ision, S.M. (Smail Mezani) and T.R.; p ojec adminis a ion, T.R.; unding acquisi ion, T.R. and
S.M. (Smail Mezani). All au ho s ha e ead and ag eed o he published e sion o he manusc ip .
Funding:
This pape was suppo ed by he ollowing p ojec s: This wo k was suppo ed by he
Doc o al g an compe i ion VSB—Technical Uni e si y o Os a a, eg. no. CZ.02.2.69/0.0/0.0/19
073/0016945 wi hin he Ope a ional P og amme Resea ch, De elopmen and Educa ion, unde
p ojec DGS/TEAM/2020-017 “Sma Con ol Sys em o Ene gy Flow Op imiza ion and Manage-
men in a Mic og id wi h V2H/V2G Technology”, FV40411 Op imiza ion o p ocess in elligence o
pa king sys em o Sma Ci y, p ojec TN01000007 Na ional Cen e o Ene gy and Tai Uni e si y
Resea che s Suppo ing P ojec TURSP 2020/34. Tai Uni e si y, Tai , Saudi A abia o suppo ing
his wo k.
Ins i u ional Re iew Boa d S a emen : No applicable.
In o med Consen S a emen : No applicable.
Da a A ailabili y S a emen : No applicable.
Acknowledgmen s:
The au ho s app ecia e he Doc o al g an compe i ion VSB—Technical Uni e -
si y o Os a a, eg. no. CZ.02.2.69/0.0/0.0/19 073/0016945 wi hin he Ope a ional P og amme
Resea ch, De elopmen and Educa ion, unde p ojec DGS/TEAM/2020-017 “Sma Con ol Sys-
em o Ene gy Flow Op imiza ion and Managemen in a Mic og id wi h V2H/V2G Technol-
ogy”, FV40411 Op imiza ion o p ocess in elligence o pa king sys em o Sma Ci y, p ojec
TN01000007 Na ional Cen e o Ene gy and Tai Uni e si y Resea che s Suppo ing P ojec TURSP
2020/34. Tai Uni e si y, Tai , Saudi A abia o suppo ing his wo k.
Con lic s o In e es : The au ho s decla e no con lic o in e es .
Re e ences
1.
Un-Noo , F.; Padmanaban, S.; Mihe -Popa, L.; Mollah, M.N.; Hossain, E. A comp ehensi e s udy o key elec ic ehicle (EV)
componen s, echnologies, challenges, impac s, and u u e di ec ion o de elopmen . Ene gies 2017,10, 1217. [C ossRe ]
2.
Mok ani, Z.; Rekioua, D.; Meba ki, N.; Rekioua, T.; Bacha, S. P oposed ene gy managemen s a egy in elec ic ehicle o
eco e ing powe excess p oduced by uel cells. In . J. Hyd ogen Ene gy 2017,42, 19556–19575. [C ossRe ]
3.
Kim, H. Pe sonaliza ion o Elec ic Vehicle Accele a ing Beha io Based on Mo o To que Adjus men o Imp o e Indi idual
D i ing Sa is ac ion. Senso s 2021,21, 3951. [C ossRe ] [PubMed]
4.
Chang, C.; Zhou, Y.; Lian, J.; Liang, J. Con ol Me hod o he Dual-Winding Mo o o Online High-F equency Resis ance
Measu emen in Fuel Cell Vehicle. Senso s 2022,22, 2051. [C ossRe ]
5.
Zhang, Y.; Zhang, C.; Huang, Z.; Xu, L.; Liu, Z.; Liu, M. Real-Time Ene gy Managemen S a egy o Fuel Cell Range Ex ende
Vehicles Based on Nonlinea Con ol. IEEE T ans. T ansp. Elec i . 2019,5, 1294–1305. [C ossRe ]
Senso s 2022,22, 5669 21 o 22
6.
Kamal, E.; Adouane, L. Op imized EMS and a Compa a i e S udy o Hyb id Hyd ogen Fuel Cell/Ba e y Vehicles. Ene gies
2022
,
15, 738. [C ossRe ]
7.
Wei, H.; Fan, L.; Ai, Q.; Zhao, W.; Huang, T.; Zhang, Y. Op imal ene gy alloca ion s a egy o elec ic ehicles based on he
eal- ime model p edic i e con ol echnology. Sus ainable Ene gy Technol. Assess. 2022,50, 101797. [C ossRe ]
8.
Dje ioui, A.; Houa i, A.; Zeghlache, S.; Saim, A.; Benkho is, M.F.; Mesbahi, T.; Machmoum, M. Ene gy managemen s a egy
o Supe capaci o / Fuel Cell ene gy s o age de ices o ehicle applica ions. In . J. Hyd ogen Ene gy
2019
,44, 23416–23428.
[C ossRe ]
9.
Abdelsha y, M.A.; Hassan, H.; Ju asz, J. Op imal design o a g id-connec ed desalina ion plan powe ed by enewable ene gy
esou ces using a hyb id PSO-GWO app oach. Ene gy Con e s. Manag. 2018,173, 331–347. [C ossRe ]
10.
Thoun hong, P.; Raël, S.; Da a , B. Ene gy managemen o uel cell/ba e y/supe capaci o hyb id powe sou ce o ehicle
applica ions. J. Powe Sou ces 2009,193, 376–385. [C ossRe ]
11.
Zhang, X.; Mi, C.C.; Mas u , A.; Daniszewski, D. Wa ele - ans o m- based powe managemen o hyb id ehicles wi h mul iple
on-boa d ene gy sou ces including uel cell, ba e y and ul acapaci o . J. Powe Sou ces 2008,185, 1533–1543. [C ossRe ]
12.
Depa u e, C.; Lhomme, W.; Sica d, P.; Bouscay ol, A.; Boulon, L. Real- ime backs epping con ol o uel cell ehicle using
supe capaci o s. IEEE T ans. Veh. Technol. 2017,67, 306–314. [C ossRe ]
13.
Benmouna, A.; Beche i , M. Combined passi i y based con ol and op imal con ol o ene gy managemen o uel cell/ba e y
hyb id sys em. Asian J. Con ol. 2019,21, 1857–1868. [C ossRe ]
14.
Aouzellag, H.; Ghedamsi, K.; Aouzellag, D. Ene gy managemen and aul ole an con ol s a egies o uel cell/ul a capaci o
hyb id elec ic ehicles o enhance au onomy, e iciency and li e ime o he uel cell sys em. In . J. Hyd ogen Ene gy
2015
,40,
7204–7213. [C ossRe ]
15.
Sanka , K.; Jana, A.K. Nonlinea mul i a iable sliding mode con ol o a e e sible PEM uel cell in eg a ed sys em. Ene gy
Con e s. Manag. 2018,171, 541–565. [C ossRe ]
16.
Qi, Z.; Fu, X.A. Neu al Ne wo k Fuzzy Ene gy Managemen S a egy o Hyb id Elec ic Vehicles Based on D i ing Cycle
Recogni ion. Appl. Sci. 2020,10, 696. [C ossRe ]
17.
Soumeu , M.A.; Gasbaoui, B.; Abdelkhalek, O.; Ghouili, J.; Toumi, T.; Chaka , A. Compa a i e s udy o ene gy managemen
s a egies o hyb id p o on exchange memb ane uel cell ou wheel d i e elec ic ehicle. J. Powe Sou ces
2020
,462, 228167.
[C ossRe ]
18.
Huang, Z.; Zhang, C.; Zeng, T.; L , C.; Chan, S.H. Modeling and ene gy managemen o a pho o ol aic- uel cell-ba e y hyb id
elec ic ehicle. Ene gy S o age 2019,1, e61. [C ossRe ]
19.
Odeim, F.; Roes, J.; Wülbeck, L.; Heinzel, A. Powe managemen op imiza ion o uel cell/ba e y hyb id ehicles wi h expe i-
men al alida ion. J. Powe Sou ces 2014,252, 333–343. [C ossRe ]
20.
Zhang, X.; Liu, L.; Dai, Y.; Lu, T. Expe imen al in es iga ion on he online uzzy ene gy managemen o hyb id uel cell/ba e y
powe sys em o UAVs. In . J. Hyd ogen Ene gy 2018,43, 10094–10103. [C ossRe ]
21.
Lin, C.; Luo, W.; Lan, H.; Hu, C. Resea ch on Mul i-Objec i e Compound Ene gy Managemen S a egy Based on Fuzzy Con ol
o FCHEV. Ene gies 2022,15, 1721. [C ossRe ]
22.
Valdez-Resendiz, J.E.; Rosas-Ca o, J.C.; Mayo-Maldonado, J.C.; Claudio-Sanchez, A.; Ruiz-Ma inez, O.; Sanchez, V.M. Imp o e-
men o ul acapaci o s-ene gy usage in uel cell-based hyb id elec ic ehicle. In . J. Hyd ogen Ene gy
2020
,45, 13746–13756.
[C ossRe ]
23.
U banski, K.; Janiszewski, D. Senso less Con ol o he Pe manen Magne Synch onous Mo o . Senso s
2019
,19, 3546. [C ossRe ]
24.
Oubelaid, A.; Taib, N.; Nikolo ski, S.; Alha bi, T.E.A.; Rekioua, T.; Flah, A.; Ghoneim, S.S.M. In elligen Speed Con ol and
Pe o mance In es iga ion o a Vec o Con olled Elec ic Vehicle Conside ing D i ing Cycles. Elec onics
2022
,11, 1925. [C ossRe ]
25.
Also yani, I.M.; Id is, N.R.N.; Lee, K.B. Dynamic Hys e esis To que Band o Imp o ing he Pe o mance o Lookup-Table-Based
DTC o Induc ion Machines. IEEE T ans. Powe Elec on. 2017,33, 7959–7970. [C ossRe ]
26.
Espa za Sola, T.; Chiu, H.J.; Liu, Y.C.; Rahman, A.N. Ex ending DC Bus U iliza ion o Induc ion Mo o s wi h S a o Flux O ien ed
Di ec To que Con ol. Ene gies 2022,15, 374. [C ossRe ]
27. Moa i, Y.; Kouzi, K.; Iqbal, A. Adap i e op imized DTC-SVM using me aheu is ic Ba algo i hm o DSIM ed by IMC based on
obus syne ge ic speed con olle . In . T ans. Elec . Ene gy Sys . 2021,31, e12697. [C ossRe ]
28.
Ta usan, S.A.A.; Jidin, A.; Jamil, M.L.M. The op imiza ion o o que ipple educ ion by using DTC-mul ile el in e e . ISA T ans.
2022,121, 365–379. [C ossRe ]
29.
Kakouche, K.; Guendouz, W.; Rekioua, T.; Mezani, S.; Lubin, T. Applica ion o uzzy con olle o minimize o que and lux
ipples o PMSM. In P oceedings o he In e na ional Con e ence on Ad anced Elec ical Enginee ing (ICAEE), Algie s, Alge ia,
19–21 No embe 2019. [C ossRe ]
30.
Ko kmaz, F.; Caki , M.F.; Topaloglu, I.; Gu buz, R. A i icial Neu al Ne wo k Based DTC D i e o PMSM. In . J. Ins um. Con ol
Sys . 2013,3, 1–7. [C ossRe ]
31.
Li, G.; Hu, J.; Li, Y.; Zhu, J. An Imp o ed Model P edic i e Di ec To que Con ol S a egy o Reducing Ha monic Cu en s and
To que Ripples o Fi e-Phase Pe manen Magne Synch onous Mo o s. IEEE T ans. Ind. Elec on.
2018
,66, 5820–5829. [C ossRe ]
32.
Elshe biny, H.; Ahmed, M.K.; Elwany, M. Compa a i e E alua ion o To que Con ol S a egies o In e io Pe manen Magne
Synch onous Mo o o Elec ic Vehicles. Pe iod. Poly ech. Elec . Eng. Compu . Sci. 2021,65, 244–261. [C ossRe ]

Senso s 2022,22, 5669 22 o 22
33.
Mesloub, H.; Benchouia, M.H.; Goléa, A.; Goléa, N.; Benbouzid, M.E.H. P edic i e DTC schemes wi h PI egula o and pa icle
swa m op imiza ion o PMSM d i e: Compa a i e simula ion and expe imen al s udy. In . J. Ad . Manu . Technol.
2016
,86,
3123–3134. [C ossRe ]
34.
Hassan, M.R.; Mossa, M.A.; Dousoky, G.M. E alua ion o Elec ic Dynamic Pe o mance o an Elec ic Vehicle Sys em Using
Di e en Con ol Techniques. Elec onics 2021,10, 2586. [C ossRe ]
35.
Mo apon, S.N.; T emblay, O.; Dessain , L.A. De elopmen o a gene ic uel cell model: Applica ion o a uel cell ehicle simula ion.
In . J. Powe Elec on. 2012,4, 505–522. [C ossRe ]
36.
Mo apon, S.N.; Dessain , L.A.; Al-Haddad, K. A compa a i e s udy o ene gy managemen schemes o a uel-cell hyb id
eme gency powe sys em o mo e-elec ic ai c a . IEEE T ans. Ind. Elec on. 2013,61, 1320–1334. [C ossRe ]
37.
Podde , A.K.; Chak abo y, O.; Islam, S.; Kuma , N.M.; Alhelou, H.H. Con ol S a egies o Di e en Hyb id Ene gy S o age
Sys ems o Elec ic Vehicles Applica ions. IEEE Access 2021,9, 51865–51895. [C ossRe ]
38.
T emblay, O.; Dessain , L.A.; Dekkiche, A.I. A Gene ic Ba e y Model o he Dynamic Simula ion o Hyb id Elec ic Vehicles.
In P oceedings o he IEEE Con e ence on Vehicle Powe and P opulsion (VPPC), A ling on, TX, USA, 9–12 Sep embe 2007.
[C ossRe ]
39.
Zandi, M.; Payman, A.; Ma in, J.P.; Pie ede ici, S.; Da a , B.; Meibody-Tabe , F. Ene gy Managemen o a Fuel
Cell/Supe capaci o /Ba e y Powe Sou ce o Elec ic Vehicula Applica ions. IEEE T ans. Veh. Technol.
2010
,60,
433–443. [C ossRe ]
40.
Ahmed, A.A.; Akl, M.M.; Rashad, M.E.E. A compa a i e dynamic analysis be ween model p edic i e o que con ol and
ield-o ien ed o que con ol o IM d i es o elec ic ehicles. In . T ans. Elec . Ene gy Sys . 2021,31, e13089. [C ossRe ]
41.
Ban, F.; Lian, G.; Zhang, J.; Chen, B.; Gu, G. S udy on a No el P edic i e To que Con ol S a egy Based on he Fini e Con ol Se
o PMSM. IEEE T ans. Appl. Supe cond. 2019,29, 3601206. [C ossRe ]
42.
Sand e-He nandez, O.; Rangel-Magdaleno, J.; Mo ales-Capo al, R. A Compa ison on Fini e-Se Model P edic i e To que Con ol
Schemes o PMSMs. IEEE T ans. Powe Elec on. 2018,33, 8838–8847. [C ossRe ]
43.
Becha , M.; Hazzab, A.; Habbab, M.; Sica d, P.; Slimi, M. RT-LAB Pla o m o Real-Time Implemen a ion o Luenbe ge Obse e
Based Speed Senso less Con ol o Induc ion Mo o . J. Au om. Mob. Robo . In ell. Sys . 2019,13, 65–72. [C ossRe ]