Received: 7 October 2019 | Accepted: 24 January 2020

DOI: 10.1002/cae.22203

RESEARCH ARTICLE

Light electric vehicle powertrain: Modeling, simulation,

and experimentation for engineering students using PSIM

Salvador Vidal‐Bravo1 | Javier De La Cruz‐Soto2 |

1
Mario Roberto Arrieta Paternina | Mónica Borunda2 |
Alejandro Zamora‐Mendez3

1
Department of Electrical Engineering,
National Autonomous University of Abstract
Mexico (UNAM), Mexico City, México Electric transport has been gaining more interest and importance in the last
2
Gerencia de Control, Electrónica y years. Moreover, the variety of transport systems driven by an electric motor
Comunicaciones, CONACYT‐ Instituto
Nacional de Electricidad y Energias has also been increased. Therefore, it is very important to provide engineering
Limpias, Cuernavaca, Morelos, México students with solid and comprehensible knowledge, as well as handly tools, for
3
Faculty of Electrical Engineering, modeling an electric traction system. So far, this has been a difficult task for
Universidad Michoacana de San Nicolas
undergraduate students due to a variety of subjects involved, such as electric
de Hidalgo, Morelia, Michoacán, México
machines, power converters, control systems, and load estimations. This work
Correspondence is intended to provide a detailed and comprehensible description of the
Javier De La Cruz‐Soto, Department of
Renewable Energies, CONACYT‐Instituto
analysis, modeling, and implementation of a powertrain of a light utility
Nacional de Electricidad y Energias vehicle using power electronics simulation (PSIM) commercial software in
Limpias, Cuernavaca 62490, Morelos, order that the student acquires the required knowledge for analyzing multiple
México.
Email: javier.delacruz@ineel.mx configurations in the powertrain.

KEYWORDS
engineering students, light electric vehicle, powertrain, PSIM

1 | INTRODUCTION simulation inputs, whereas the fuel consumption and the

state‐of‐charge (SOC) are the monitored variables for the
The considerable efforts to develop electric vehicles (EVs) as energy management. Also, the modeling of EV is accom-
a strategy to mitigate global warming and fossil fuel plished in a Matlab & Simulink‐based environment in [14],
consumption have become a challenging and interesting emphasizing on the role of the driving style in the optimal
topic in the automotive industry. One of the most important performance and energy management. Likewise, in [13]
differences between electric and internal combustion engine dynamical modeling of an electrified powertrain was
(ICE) vehicles, is the powertrain system [1,2,10,21]. In EV presented considering the two‐inertia moment model to
powertrains, the performance assessment is made by represent the electric motor and dynamic load, focusing the
modeling and analyzing the mechanical and electrical analysis on control strategies based in the determination of
subsystems that compose the powertrain architecture. “hybrid states”, such as backlash and half‐shaft torque.
Thereby, the powertrain models are useful to analyze, design The analysis and modeling consist of an iterating process
and optimize the energy consumption estimation and to validate mathematical models. Mechanical load require-
autonomous vehicle control. For instance, in [31] a hybrid ments should be defined from a driving cycle [4,21,25].
electric vehicle (HEV) powertrain model is implemented There are standardized drive cycles for some kind of vehicles
using the backward power flow method, in which the and routes. In the modeling process, authors divide power-
acceleration, velocity and slope information is employed as trains into mechanical and electrical systems [4]. Mechanical

Comput Appl Eng Educ. 2020;1–14. wileyonlinelibrary.com/journal/cae © 2020 Wiley Periodicals, Inc. | 1
2 | VIDAL‐BRAVO ET AL.

systems are composed of electric motors, transmissions, performance prediction of the vehicle under a wide variety
and wheels [4]. On the other hand, electrical systems are of conditions. In Section 2, the modeling of the electrical
composed of power sources, power converters, and electric and mechanical systems through the mathematical
motors. Electric motors are the electromechanical interface formulation is done. Section 3 describes the simulation of
among electrical and mechanical systems. Both, electrical the electrical and mechanical systems of the powertrain
and mechanical components, are modeled individually and using a digital commercial simulator called power electro-
then integrated into a powertrain model [4,8,11,21,25]. As a nics simulation (PSIM©). PSIM is a friendly simulation
first approach, simplified models are preferred due to the package for power electronic components with modules for
short simulation time, since they mainly involve digital control and motor drives, suitable for vehicular
converter and motors models with constant efficiency. models simulations. To validate the analysis, Section 4
Moreover, light electric vehicles (LEVs) represent an describes an experiment carried out with a real‐world route.
important alternative in transport development, particularly The last section shows the conclusions of the analysis as
in larger cities, since they require less energy and space [11]. well as recommendations to improve results.
Some LEVs, such as golf carts, have been used in resorts,
hotels, airports, and recently as public transportation in
neighborhoods of Mexico City. So far, experimental 2 | MODELING OF A LIGHT
research has been performed with golf carts LEVs [7,12,29], E L E C T R I C V E H I C L E PO W E R T R A I N
with the aim to analyze the incorporation of fuel cells.
However, mechanical and electrical components are not An electric golf cart has a light electric vehicle powertrain
modeled in detail and their performance prediction is still that possesses electrical and mechanical power systems,
limited. Thus, this investigation contributes with a complete as shown in Figure 1. The electrical power system
modeling and experimentation of an LEV, where every consists of batteries, a DC–DC chopper motor driver and
subsystem is described in detail, including DC–DC power a DC motor. The speed profile of the vehicle depends on
converter and mechanical load models which are not taken the commands resulting from the braking and accelera-
into consideration in [31]. Also, energy consumption and tion handled by the driver. These speed variations
the SOC are discussed unlike the work reported in [13]. determine the speed and torque to be supplied by the
Specifically, a lead‐acid battery model that includes the SOC electrical motor. This is done through a throttle sensor,
estimation and the effect of the battery discharge/recharge which receives the speed signal operated by the driver
in terminal voltage is developed. Regarding the DC–DC and establishes a voltage reference value in the pulse
power converter, a switched model is implemented to width modulation (PWM) control, to control the desired
illustrate the commutation effect in the voltage and current speed and torque of the motor [24]. The electrical
signals. Finally, this development takes into account the power is provided by a pack of six‐lead acid batteries
efficiency associated with every powertrain's component for connected in series, and it is transferred to the
attaining realistic simulation results. mechanical system through a drive axle gearbox, which
In this work, we present a detail model and simulation links the electrical and mechanical power systems in the
of most of the components of an LEV to obtain a good powertrain.

FIGURE 1 Components of the light

electric vehicle
VIDAL‐BRAVO ET AL. | 3

The mechanical system encompasses the drive axle The SOC and the depth of discharge (DOD),
gearbox, drive axle shaft, and two rear wheels for drawing determine the amount of available energy and used
the driving force. The DC motor is in charge of providing energy, respectively, from 0 up to 100% according to the
the tractive forces that generate the motion of the vehicle following equations:
through a direct coupling between the electrical and
mechanical systems which forms an electromechanical 0 ≤ SOC ≤ 1, (3)
system with coupled electrical and mechanical dynamics. (4)
DOD = 1 − SOC,
In the following, we consider an LEV powertrain with a
conventional configuration, described above, in which where the battery capacity Q(t), is a function of time and
the transmission system behaves as a fixed gearing and is employed during operation, and it depends on the
assesses the performance of the model along with a current profile IB(t) according to the following equation:
validation carried out with a simulation of the model and
a real driving test. t
Q (t ) = Q (t 0 ) + ∫IB (t ) dt, (5)
t0

2.1 | Electrical system modeling where Q(t0) is the initial capacity [28]. Likewise, the
SOC can be represented as a function of time denoted by
The electrical power to impulse the vehicle flows through z(t) and defined by
the powertrain, from the electrical to the mechanical
system, providing the required torque to generate traction t
1
at the drive axle. Power consumption depends on the z (t ) = z (t 0 ) +
Cx
∫IB (t ) dt, (6)
t0
vehicle speed and the applied force on the wheels.
However, if the conversion efficiencies in the converter where the initial state of charge, z(t0), is the percentage
ηC, motor ηM, and transmission ηT are taken into con- of charge at the beginning of the analysis and can be
sideration, the total electric power supplied by the source determined by the terminal voltage or the electrolyte
PB, is reduced. Therefore, it is necessary to analyze each density [23,30].
subsystem into the powertrain. The electrical system Considering the concepts described above, a simple
consists of three subsystems: batteries, DC‐DC converter, model of the electrical behavior of a battery is illustrated
and electric motor, and their models are described as in Figure 2, [19,23,30]. In this case, a controlled voltage
follows: source VOC, connected in series with an internal
resistance Rint, changes the value of the voltage and
current as a function of time.
2.1.1 | Lead‐acid battery model Thus, the battery voltage VB(t), depends on the state
of charge z(t), and the current IB(t), as described by the
This corresponds to the device where electric energy is following expression:
stored. The main parameters that define the battery's
performance are established by its rated capacity, the VB (t ) = VOC (z (t )) − IB (t ) Rint , (7)
charge/discharge rate, and the state of charge (SOC). The
rated capacity Ex, is considered as the total energy stored Then, the battery power PB is the amount of the
by the battery in x hours and is given by energy available during the operating time and is obtained

Ex = Cx × VB,nom, (1)

where Cx is the battery current capacity during x hours,

and VB,nom corresponds to the rated voltage of the battery.
The SOC is the total capacity of discharge [23], and it is
given by

Q (2)
SOC = 1 − ,
Cx

where Q is the total amount of charge removed during

discharge. FIGURE 2 Equivalent circuit of a basic battery model
4 | VIDAL‐BRAVO ET AL.

motor terminals, the voltage stays constant, and current

flows through the circuit. On the other hand, if the
switch is off, the battery is disconnected, there is no
voltage, and the current ioff, flows in the winding of the
motor, as shown in Figure 4. The commutation times, ton
and toff, are related by a duty cycle, D, as

FIGURE 3 Basic configuration of a chopper fed DC motor ton t (8)

D= = on .
drive [26] ton + toff T

as the product between the voltage VB, and the current IB, at for a constant commutation frequency f = 1/T. Once the
the terminals of the battery, that is PB = VB · IB, duty cycle is established, it is possible to determine
the relationship between the battery voltage VB, and
the voltage of the electric motor VM, as VM = D · VB.
2.1.2 | DC–DC converter model Likewise, the DC motor current is given by IM = D · VB/R.

The DC–DC converter, also known as a chopper, is

commonly employed as a motor driver in traction 2.1.3 | DC motor model
systems. A basic configuration of a chopper fed DC motor
drive consists of a voltage source V, switch S, and diode The DC motor is modeled as a wound field DC motor that
D, and is shown in Figure 3. In this work, the load for the consists of a field inductor LS, connected in series with an
DC–DC converter is represented by a series of wound armature EA, formed with the field and armature resistances,
field DC motor. RS and RA, respectively, as presented in Figure 5 [3,5].
The chopper operation states are shown in Figure 4. If By applying Kirchhoff's law in the circuit depicted in
the switch S is on, the battery is connected with the Figure 5, it is possible to obtain the relationship between

FIGURE 4 Chopper operation states

VIDAL‐BRAVO ET AL. | 5

where the aerodynamic drag force corresponds to the

wind force on the vehicle in motion and depends on
the air density Ρair, the wind speed vW, and direction of
the wind, the aerodynamic drag coefficient cD, the frontal
area Af of the vehicle, and the vehicle speed vT, as
follows [6]:

1 (11)
FD = ρ CD AF (v T + v W )2.
2 air
FIGURE 5 Equivalent circuit of a wound field DC motor
connected in series [3,5]
The rolling resistance force FR, is due to the vehicle
tires deformation caused by the road surface [27]. This is
voltages at the motor terminals, and at the field and usually defined as a function of the normal force FN, and
armature windings, as follows: the rolling coefficient cR, as follows:

dIS FR = FN ⋅ c R . (12)
VM = EA + IA ⋅ RA + LS + IS ⋅ RS , (9)
dt
The rolling parameter is determined by the vehicle
where the electromotive force (EMF) generated by the speed, as well as the road and the tire conditions. The
armature is defined by EA = ∅ · kA · ωM, and depends on normal force acts over the vehicle, even if it is in move-
the magnetic flux ∅, the armature constant kA, and a ment or not, from the diagram of forces it is equal to the
rotational speed ωM. Likewise, the mechanical torque vertical component of the weight of the vehicle, namely,
produced in the motor shaft is τM = ϕ · kS · IA, which also
depends on the magnetic flux, a field constant kS, and the FN = mT ⋅ g cos α, (13)
armature current IA. Hence, the mechanical power is
given by PM = IA · EA = τM · ωM, and the electrical power where mT is the vehicle's total mass, g is the gravitational
is defined by PE = VM · IM. It is noteworthy to remark that acceleration, and α is the road slope.
according to these statements the mechanical power The third force that acts over the vehicle is due to
given by a motor is less than the electrical power the climbing resistance, thus it is called the climbing
necessary to operate it. resistance force FG, and from Figure 6 it is given by

FG = mT ⋅ g sin α, (14)
2.2 | Mechanical system modeling
Note that when the vehicle goes on a plain surface,
Mechanical components of the powertrain system employ the climbing resistance is zero.
the electromechanical torque from the DC motor to provide If it is provided enough traction power, then the
the traction force to the drive axle. Then, the mechanical traction force drives the vehicle. The traction power PT is
power is generated and transmitted to the mass of the the product of the necessary force to move the vehicle
vehicle, leading to the rotational inertia JM, the global times the linear speed vT, [27], however, it turns out that
efficiency of the system η, and the forces that affect
the dynamics of the mechanical system [21], which are
described as follows. For modeling the behavior of the
mechanical system, a simplified model based on two drive
wheels is considered and illustrated in Figure 6. This model
simplifies the vehicle movement to a two‐dimension
analysis involving the corresponding forces of the move-
ment along the longitudinal and vertical axis [3,5,9,27].
The vehicle traction force FT, is given as the sum of
the aerodynamic drag force FD, the rolling resistance
force FR, and the climbing resistance force FG,

TT = FD + FR + FG, (10) FIGURE 6 Forces involved in the two‐wheel vehicle model

6 | VIDAL‐BRAVO ET AL.

τM ⋅ ωM = τT ⋅ ω T . (20)

In the next section, a PSIM©‐based LEV simulation

model is developed and its validation is done with
experimental results presented in Section 4.

FIGURE 7 Mechanical connection between the M and L loads

3 | MODELING AND
it is more convenient to express it in terms of the angular SIMULATION OF THE
speed ωT and torque τT as follows: POWERTRAIN IN PSIM SOFTWARE

PT = FT ⋅ v T = τT ⋅ ω T , (15) The powertrain model is built with elements of PSIM

specified with nominal parameters of the LEV. The main
On the other hand, the dynamics of the mechanical parameters for the LEV are shown in Table 1. The overall
system in the vehicle are described by means of the configuration, including the electrical and mechanical
rotational equilibrium's law in a two‐mass system as systems, of the PSIM©‐based LEV model, illustrated in
shown in Figure 7. The electric motor is mechanically Figure 9. The electrical system consists of the power
coupled with the drive axle through a fixed gearing supply, power conversion, speed control, and electro-
transmission, and its dynamics are described in the mechanical subsystems, whereas, the mechanical system
following expressions: is formed by the transmission and traction subsystems.
As illustrated in Figure 9 the power supply subsystem
τM = JMωM + τfM , (16) is composed of six lead‐acid batteries, which are
(17) represented with six identical subcircuits connected in
τL = JLωL + τfL,
series. The battery is modeled with a PSIM© library, with
specifications described in Table 2, as a controlled‐voltage
where ω refers to the angular acceleration, J is the source in series with a resistor. Likewise, the battery
moment of inertia, τf corresponds to the friction losses subcircuit, shown in Figure 10, emulates the recharge
and gravity effects and the subindexes, M and L, stand for and discharge process, depending on the current flow
the motor and the mechanical loads, respectively [20]. direction. The subcircuit is linked with the main
When the friction losses are neglected, the mechan- circuit through the positive (B+) and the negative (B−)
ical coupling between the DC motor and the drive axle bidirectional ports. The state of charge of the battery is
using a gearbox to reduce speed and increasing torque is computed by measurements provided from the voltage
shown in Figure 8. According to Figure 8, the rotational and current sensors. By employing math function blocks
speed in the motor shaft ωM, is n times larger than the
rotational speed ωT in the wheels of the drive axle.
Likewise, the drive axle torque τT, is n times larger than TABLE 1 Parameter of the light electric vehicles
the torque τM applied by the motor. These relationships Parameter Symbol Value Unit
are presented in the following expressions:
Vehicle mass mT 592 kg

ωM = n ⋅ ω T. (18) Frontal area of the vehicle Af 2.32 m2

τT Wheel diameter Dw 0.215 m

τM = . (19)
n Dynamo shaft diameter Dd 0.0213 m
Coefficient of aerodynamic drag CD 0.2 –
Finally, since the mechanical power developed by the
motor is equal to the traction power it follows that Air density ρair 1.225 kg/m3
Rolling resistance coefficient CR 0.02 –
Gearbox efficiency ηgb 85 %
Load inertia JL 9.08 kg·m2
Gearbox ratio n 12.5 –
Height of CG from the ground h 0.54 m
FIGURE 8 Mechanical coupling between the DC motor and
Inertia moment of the load JL 9.08 kg·m2
the drive axle
VIDAL‐BRAVO ET AL. | 7

FIGURE 9 Schematic model of the light electric vehicles in PSIM. PSIM, power electronics simulation

TABLE 2 Battery specifications of 1,000 Hz. The reference voltage is 21.398 V, which is
equivalent to the maximum theoretical vehicle speed
Parameter Symbol Value Unit
when the gearbox ratio is 12.5:1. The voltage reference
Battery type – Lead‐acid –
can be adjusted when the gearbox is changed to modify
Rated voltage VB 6 V the vehicle speed. The voltage control provides the
Rated capacity (@ 20 hr) Cx 225 Ah desired speed by the driver, and it is compared with the
reference voltage to compute the duty cycle.
Initial state‐of‐charge SOC 100 %
The speed control subsystem is composed of a lookup
−6
Internal resistance Rint 66 × 10 Ω table and a series of time elements connected to the W
port Figure 11. The lookup table emulates the vehicle
throttle and introduces the drive cycle data, vehicle speed
in PSIM, it is possible to estimate the SOC and include versus time, to control the duty cycle in the PWM.
the voltage equations during the charge/discharge Different drive cycles can be simulated changing the text
process. To compute the discharge and recharge voltage, file that contains the speed values in time. It is possible to
the number of cells (battery pack) and the open‐circuit manually set the value of voltage employing a variable
voltage for each cell are used. To compute the output resistance element connected to the L and G ports in
power of the battery pack, the voltage, and the current Figure 11.
values are recorded. A DC machine element from the PSIM library is used
The subcircuit of the motor driver is a DC–DC to model the DC motor with the armature and field
chopper made up of a DC–DC step‐down converter. connected in series, according to the parameters specified
Power MOSFETs are used and the circuit is controlled by in Table 3, as shown in Figure 9. Even if EMF resistance
the PWM method, as depicted in Figure 11. The power is not included in Figure 9, its value has been taking into
converter subcircuit is linked with the main external account. The voltage and current measurements are
circuit to allow the flow of current from the battery measured in tests to estimate the electrical power at the
subcircuit. motor terminals.
A low‐pass filter, with an inductive element in The transmission subsystem includes a gearbox
parallel with a capacitor, is assembled to smooth the connected with the shaft of the DC machine. The gearbox
current and voltage signals. According to the real wiring ratio can be modified to change the vehicle speed. The
diagram, two bidirectional ports link the DC–DC con- torque and angular speed are recorded by sensors so that
verter with the armature (A1) and the field (S2) terminals the mechanical power is obtained. The gearbox is
of the motor. The PWM circuit consists of a sawtooth‐ connected to the traction load through a bidirectional port.
wave voltage source emulating the oscillator that The traction subcircuit, shown in Figure 12, represents
generates the PWM signal with a commutation frequency the traction load using function blocks in PSIM. A total
8 | VIDAL‐BRAVO ET AL.

FIGURE 10 Power supply subsystem in PSIM©. PSIM, power electronics simulation

FIGURE 11 Motor driver, DC–DC subcircuit in PSIM. PSIM, power electronics simulation
VIDAL‐BRAVO ET AL. | 9

TABLE 3 DC motor parameters vehicle and wind conditions are taken into account to
compute the drag force. As a result, the traction load
Parameter Symbol Value Unit
model computes the required torque to move the
Nominal voltage VM 36 V
mechanical load. An externally‐controlled load element in
Nominal rotation speed RPMm 3300 rpm PSIM uses the torque value to apply a force in the opposite
Nominal current IM 52 A direction of the output gearbox force applied over the
mechanical load circuit.
Rated power PM 1.86 kW
Armature resistance RA 0.085 Ω
Field resistance RS 0.032 Ω 4 | S HO R T TES T I N A R EAL ‐
Armature inductance LA 33 mH
WORLD ROUTE
Field inductance LS 25 mH
With the aim of having a reference framework and
Motor efficiency (nominal) ηm 83.33 % evaluate the performance of the model developed in
Moment of inertia JM 0.0079 kg × m2 PSIM, a short test is carried out. Figure 13 shows the
driving route using a dotted line which is a typical route
of the security personnel at the neighborhood.
vehicle mass of 592 kg is considered and includes the To determine the performance of the moving vehicle
vehicle chassis and components, batteries, driver, and components, the vehicle speed is measured from the
copilot. The slope angle of the route and rolling resistance wheels. Figure 14 shows the distribution of acquisition
values could be modified regarding the surface angle and instruments connected to the powertrain in the vehicle.
roughness conditions. The physical parameters of the Likewise, voltages and currents in the electrical system are

FIGURE 12 Traction load in PSIM©. PSIM, power electronics simulation

10 | VIDAL‐BRAVO ET AL.

FIGURE 13 Driving route map

FIGURE 14 Connection of the measuring instruments

VIDAL‐BRAVO ET AL. | 11

recorded following the wiring connections of the scheme.

A digital Rohde & Schwarz oscilloscope (model RTH 1004)
is used for the data acquisition. Also, a dynamo is used for
recording the speed of the traction axis.
When the wheel rotates, the AC voltage Vd, is
measured at the dynamo terminals by the oscilloscope,
such that the linear speed vT, of the vehicle, is determined
as a function of the voltage described by

200πDd Vd (21)
vT = ,
60 FIGURE 16 Mechanical power in the DC motor

where Dd stands for the wheel diameter of the dynamo.

Likewise, the battery voltage is monitored through a
Fluke meter (model 233Si).
Figure 15 shows the vehicle speed in the tractive
direction resulting from the simulated model in PSIM(c) and
the speed profile measured during a segment of the driving
test. As it is seen, the simulation results are consistent with
the ones provides by the model and validating it. Note that
there is a small shift between both profiles that is due to the
effects of the uncertainty in variables, for example, the
inertia moment of the load. Here, it is important to
emphasize that the only one input for the PSIM‐based FIGURE 17 Mechanical power in drive axle
simulation is the drive cycle of the vehicle.
The mechanical power of the DC motor and drive
axle is shown in Figures 16 and 17, respectively. As
expected, there is a correspondence among the pattern of
simulated and measured results. However, regions with
high variability on power demand exhibit deviations.
Despite those deviations, the model provides an accurate
prediction of the behavior of the vehicle.
Finally, the motor's voltage behavior is shown in
Figure 18. The solid line corresponds to the measured
voltage at the motor terminals, whereas the dashed and
dotted lines represent the simulated voltage at the motor
and at the motor armature, respectively. Note that the FIGURE 18 Behavior of the motor's voltage
voltage drop in the field winding has been neglected.
Once again, the predicted values from the simulated
model agree with the measured ones. 5 | COMPARISON AGAINST
SI MULI NK

As a matter of completeness, and to illustrate the advantages

of the proposed modeling tool, we present a simulation
implemented in Simulink using Matlab & Simulink.
After reviewing Refs [15–18], the DC wound series
electric motor, that is employed in our modeling, simula-
tion, and experimentation, is not embedded in a toolbox in
Simulink for powertrain applications, however, it is used
as a starter in topologies such as separately excited and
series excited. Thus, for comparison purposes, we have
FIGURE 15 Performance of the light electric vehicles: validated a small piece of our implementation related to
Simulation (dotted line) versus real test (continuous line) the electric motor using Simulink against the PSIM model.
12 | VIDAL‐BRAVO ET AL.

TABLE 4 DC machine PSIM and Simulink models

Differential and algebraic modeling

PSIM Simulink
d
Motor voltage vt = Ea + ia⋅Ra + La dt ia –
d
Field voltage vf = i f ⋅Rf + L f dt i f –
Armature voltage Ea = k ⋅ ϕ ⋅ ω m E = KE ω
Electromechanical torque τem = k ⋅ ϕ ⋅ ia τe = KTIa
dω dω
Angular speed J dtm = τem − τL J dt
= τe − τL − Bm ω − τf
Armature voltage Ea = Laf ⋅ if ⋅ ωm –
Electromechanical torque τem = Laf ⋅ if ⋅ ia –
Vt − Ia ⋅ Ra
Armature‐field mutual Laf = If ⋅ ωm
–
inductance
Voltage constant – KE = LafIf
Torque constant – KT = KE

Abbreviation: PSIM, power electronics simulation.

electromechanical torque (τe) signals are obtained for

both PSIM and Simulink environments models, re-
presented by a continuous and dotted lines, respectively.
Finally, to quantify the matching among models,
the signal‐to‐noise ratio (SNR) in dB is used as in [22].
The SNR for ia, ωm, and τe are of 15.17, 18.72, and
21.61 dB, respectively. These values indicate that
the error among signals is up to two thousand times
FIGURE 19 DC wound series motor implemented in smaller than the armature current, angular speed and
Simulink electromechanical torque signals obtained with the
Simulink model.
The model of the DC wound series motor is
implemented in Simulink using the PSIM‐based
differential and algebraic equations in Table 4 and its 6 | C O N C L U S I O N S AN D
parameters in Table 3. This model is exhibited in RE C O M M E N D AT I O N S
Figure 19 for a condition without load, which is T_L = 0.
Once the model is implemented, comparisons are done This work shows in detail the process of analytically
between PSIM and Simulink, as shown in Figure 20, modeling the performance of the subsystems that make
where the armature current (ia), angular speed (ωm), and up the powertrain of a utility vehicle. Moreover, the

FIGURE 20 Compared models

between Simulink and PSIM. PSIM, power
electronics simulation
VIDAL‐BRAVO ET AL. | 13

complete procedure for the simulation of these sub- 5. D. W. Gao, C. Mi, and A. Emadi, Modeling and simulation of
systems using the commercial software PSIM is pre- electric and hybrid vehicles, Proc. IEEE 95 (2007), no. 4, 729–745.
sented. Finally, the validation of the theoretical 6. J. G. Hayes, and G. A. Goodarzi, Electric powertrain: Energy
systems, power electronics and drives for hybrid, electric and
formulation and the simulated model are validated with
fuel cell vehicles, John Wiley & Sons, Hoboken, NJ, 2017.
measured data from a short field test.
7. H. Inc. (2014) Oakland university builds a hybrid battery fuel
Even though the results exhibit small deviations, we cell golf cart, available at https://tech‐labs.com/sites/default/
conclude that the model successfully predicts the perfor- files/HC Case Study Oakland University 1405.pdf
mance of the powertrain of the LEV. To improve the accu- 8. K. Jaber, A. Fakhfakh, and R. Neji, Modeling and simulation of
racy of these models, it is required to work with all system high performance electrical vehicle powertrains in vhdl‐ams, in
parameters as accurately as possible. Therefore, we strongly electric vehicles‐modelling and simulations, IntechOpen, Lon-
recommend measuring all quantities of the subsystems, in don, UK, 2011.
9. R. N. Jazar, Vehicle dynamics: Theory and application,
particular, the internal resistance of the batteries, the
Springer, New York, 2017.
efficiency of the converter, the resistance and inductance in
10. A. Khaligh and Z. Li, Battery ultracapacitor fuel cell, and hybrid
the motor windings, the total mass of the vehicle the gearbox energy storage systems for electric, hybrid electric, fuel cell, and
ratio and the power converter architecture. Likewise, it is plug‐in hybrid electric vehicles: State of the art, IEEE Trans. Veh.
important to take into account the pressure of the wheels Technol. 59 (2010), no. 6, 2806–2814.
and calibrate them before the test since low pressure may 11. A. Lebkowski, Light electric vehicle powertrain analysis, Zes-
cause a significant drop in the vehicle's performance. zyty Naukowe. Transport/Politechnika Slaska, 2017.
This work provides a valuable tool for engineering 12. J. Liang and Q. Jian, Simulation and test of a fuel cell hybrid golf
cart, J. Electr. Comput. Eng. 2014, 17–18.
students for modeling, simulating, and testing a power-
13. C. Lv et al., Simultaneous observation of hybrid states for cyber‐
train of a light electric vehicle. It disclosures in great physical systems: A case study of electric vehicle powertrain,
detail the knowledge needed to achieve it and, moreover, IEEE Trans. Cybernet. 48 (2018), no. 8, 2357–2367.
makes it easy to extend it and apply it to multiple con- 14. C. Lv et al., Driving‐Style‐based codesign optimization of an
figurations of electric vehicles powertrains. automated electric vehicle: A cyber‐physical system approach,
As future works are important to include some DC IEEE Trans. Ind. Electron. 66 (2019), no. 4, 2965–2975.
motor tests that allow obtaining a speed‐torque curve, the 15. MATLAB and Powertrain Blockset R, The MathWorks, Inc,
Natick, MA, 2019.
inertia moment constant, and the efficiency curve. The
16. MATLAB and Powertrain Blockset Reference R, The Math-
estimation of the capacitance values of the batteries and
Works, Inc, Natick, MA, 2019.
the implementation of energy management systems are 17. MATLAB and Powertrain Blockset Release Notes R, The
needed to improve the performance of the vehicle, MathWorks, Inc, Natick, MA, 2019.
likewise, the inclusion of generative brakes in the 18. MATLAB and Powertrain Blockset User's Guide R, The
powertrain model. MathWorks, Inc, Natick, MA, 2019.
19. S. M. Mousavi and G. M. Nikdel, Various battery models for
ORCID various simulation studies and applications, Renew. Sustain.
Energy Rev. 32 (2014), 477–485.
Javier De La Cruz‐Soto http://orcid.org/0000-0002-
20. G. Onwubolu, Mechatronics: Principles and applications,
3300-4628 Elsevier, Journal Hill, Oxford, 2005.
Mario Roberto Arrieta Paternina http://orcid.org/0000- 21. G. Park et al., Integrated modeling and analysis of dynamics for
0002-2121-6420 electric vehicle powertrains, Expert Syst. Appl. 41 (2014), no. 5,
Mónica Borunda http://orcid.org/0000-0003-4414-3815 2595–2607.
Alejandro Zamora‐Mendez http://orcid.org/0000-0001- 22. M. R. A. Paternina et al., Dynamic equivalents by modal de-
5479-9856 composition of tie‐line active power flows, IEEE Trans. Power
Syst. 32 (2017), no. 2, 1304–1314.
23. G. L. Plett, Equivalent‐circuit models, Artech House, Norwood,
REFERENCES
MA, 2015, pp. 29–33.
1. C. Chan, Design and assessment of an electric vehicle powertrain 24. V. Porobic et al., Educational framework for a motor drive
model based on real‐world driving and charging cycles, Proc. control systems: Design and performance assessment, Comput.
IEEE 90 (2002), 247–275. Appl. Eng. Educ. 25 (2017), 264–276.
2. C. C. Chan, The state of the art of electric, hybrid, and fuel cell 25. M. Pourabdollah et al., Convex optimization methods for power-
vehicles, Proc. IEEE 95 (2007), no. 4, 704–718. train sizing of electrified vehicles by using different levels of modeling
3. S. J. Chapman, Electric machinery fundamentals, 5th., McGraw‐ details, IEEE Trans. Veh. Technol. 67 (2018), no. 3, 1881–1893.
Hill, Boston, 2012. 26. M. H. Rashid, Power electronics handbook: Devices, circuits
4. G. Du, W. Cao, and S. Hu, Assessment of an electric vehicle and applications, Academic Press, Cambridge, MA, 2010.
powertrain model based on real‐world driving and charging 27. K. Reif, Brake control and driver assistance systems, Springer,
cycles, IEEE Transact. Veh. Technol., In Press. Fachmedien Wiesbaden, 2014.
14 | VIDAL‐BRAVO ET AL.

28. M. G. Simoes, and F. A. Farret, Modeling power electronics and

interfacing energy conversion systems, John Wiley & Sons, Mónica Borunda holds a PhD in
Hoboken, New Jersey, 2016. Physics from the International Sch ool
29. I. Tolj et al., Fuel cell‐battery hybrid powered light electric vehicle for Advanced Studies, SISSA, in Italy.
(golf cart): Influence of fuel cell on the driving performance, Int.
She was graduated as a Physicist at the
J. Hydrogen Energy 38 (2013), no. 25, 10 630–10 639.
Faculty of Sciences of the UNAM with
30. K. Young et al., Electric vehicle battery technologies, Springer,
New York, NY, 2013, pp. 15–56. the Gabino Barreda medal. She conducted research
31. B. Zhao, C. Lv, and T. Hofman, Driving‐cycle‐aware energy stays at CINVESTAV, Mexico, Fermilab, USA,
management of hybrid electric vehicles using a three‐dimensional McGill University, Canada and Abdus Salam Inter-
markov chain model, Automot. Innov. 2 (2019), 146–156. national Center for Theoretical Physics, Italy.
She made postdoctoral stays at the University of
AUTHOR BIOGRAPHIES Neuchâtel in Switzerland and at the University of
Granada in Spain. Later, she worked in R&D in
Salvador Vidal Bravo holds a BS in renewable energy companies in Spain, Qatar and
Renewable Energy Engineering from In- Germany. In 2014, she won a CONACYT research
stitute for Renewable Energy (IER‐UNAM) position and joined INEEL to work in projects in the
in 2019. In 2019, he joined the program in energy sector. She is the author of several interna-
MSc Electrical Engineering Power Systems, tional scientific articles and has participated in
UNAM. His areas of interest include renewable energy, several international workshops and conferences.
energy storage and electric vehicles. Currently, she belongs to the Mexican National
System of Researchers.
Javier de la Cruz is an Electrical
Engineer from ITSON. He studied Master's Alejandro Zamora‐Mendez obtained
degree in Electrical Engineering at Institu- his BS and MSc in Electrical Engineer-
to Tecnológico de Morelia, and hold ing from Universidad Michoacana de
a PhD in Electrical Engineering from San Nicolas de Hidalgo (UMSNH),
CINVESTAV‐Guadalajara. Since 2014 is with CONA- Morelia, Mexico, in 2005 and 2008,
CYT in the Renewable Energies Department at the respectively. He joined the Electrical Engineering
Instituto Nacional de Electricidad y Energías Limpias Faculty, UMSNH, in 2008, where he is a full‐time
(INEEL) in Cuernavaca Morelos. His research interests Professor. He received the PhD degree from
are Energy Storage, electrical machines and Renewable CINVESTAV‐Guadalajara in 2016. His areas of
Energies. interest include oscillation monitoring, signal pro-
cessing, coherency, modal identification, and power
Mario R. Arrieta Paternina holds a BS system dynamics and control.
and a MSc in Electrical Engineering from
National University of Colombia, in 2007
and 2009, respectively. In 2017, he
How to cite this article: Vidal‐Bravo S,
obtained his PhD degree in electrical
De La Cruz‐Soto J, Arrieta Paternina MR,
engineering from CINVESTAV. He joined the Depart-
Borunda M, Zamora‐Mendez A. Light electric
ment of Electrical Engineering, UNAM, in 2017. He was
vehicle powertrain: Modeling, simulation, and
a visiting scholar at Rensselaer Polytechnic Institute in
experimentation for engineering students using
2016. His areas of interest include model reduction,
PSIM. Comput Appl Eng Educ. 2020;1–14.
power system dynamics, WAMS‐PMU, and modeling,
https://doi.org/10.1002/cae.22203
control and simulation of power systems.