1
Bus-to-Route and Route-to-Bus Approaches in
Hybrid Electric Buses Fleet for Li-ion Battery
Lifetime Extension
Enfoque Autob
´
us-a-ruta y Ruta-a-autob
´
us en flotas de
autobuses h
´
ıbridos-el
´
ectricos para extensi
´
on de la vida
´
util de
bater
´
ıas de litio-ion
Abstract— This paper aims to propose a methodology
for managing Li-ion battery life for a whole fleet, focusing
on improving the total operating costs for hybrid electric
buses. This problem is approached in two different ways:
bus-to-route and route-to-bus. Bus-to-route optimization
focuses on developing an energy management strategy
for each bus in the fleet. A techno-economic, energy and
battery aging analysis of the fleet was carried out. As
per the outcomes of this analysis, buses were grouped
according to the state of health of the battery of each
bus. Based on this analysis and classification, the route-
to-bus approach was applied. This technique involves a
re-evaluation of the energy management system and a re-
organization of the buses according to the state of health
of the battery of each bus. Increases in battery (BT) life of
up to 10.7% were obtained using the proposed approach.
Keywords— Fleet energy management, battery state
of health, dynamic programming, hybrid electric bus, Li-
ion battery.
Resumen— El objetivo de este art
´
ıculo es proponer
una metodolog
´
ıa para la gesti
´
on de la vida
´
util de las
bater
´
ıas de Litio-ion de una flota completa de autobuses
el
´
ectrico-h
´
ıbridos orientado a mejorar el costo total de
operaci
´
on. Esta propuesta se ha abordado desde dos puntos
de vista: autob
´
us-a-ruta y ruta-a-autob
´
us. La optimizaci
´
on
autob
´
us-a-ruta se enfoca en la generaci
´
on de la gesti
´
on
energ
´
etica de cada autob
´
us perteneciente a la ruta. Para es-
te efecto, se ha llevado a cabo un an
´
alisis tecno-econ
´
omico,
energ
´
etico y de esperanza de vida de las bater
´
ıas de la flota
de veh
´
ıculos. Partiendo del resultado de este an
´
alisis, los
veh
´
ıculos han sido agrupados de acuerdo al par
´
ametro del
estado de salud de la bater
´
ıa de cada autob
´
us. Bas
´
andose en
dicho an
´
alisis y clasificaci
´
on, se ha realizado la aplicaci
´
on
del enfoque ruta-a-autob
´
us. Esta t
´
ecnica considera tanto
una re-evaluaci
´
on de la gesti
´
on energ
´
etica y/o una re-
organizaci
´
on de los autobuses de acuerdo al estado de salud
de sus bater
´
ıas. Se ha conseguido un incremento de hasta
10.7 % de la vida
´
util de las bater
´
ıas haciendo uso de los
enfoques propuestos.
Palabras Clave— Gesti
´
on energ
´
etica de flota, estado
de salud de bater
´
ıa, programaci
´
on din
´
amica, autob
´
us
h
´
ıbrido-el
´
ectrico, bater
´
ıa de Litio-ion.
I. INTRODUCTION
All sectors in Europe have recorded a reduction in
greenhouse gas (GHG) emissions since 1990, except for
the transport sector [1]. In 2015 transport was respon-
sible for 25.8% of total emissions in Europe, with road
transport being the most pollutant sub-sector, accounting
for 72.8% in 2015 [1]. The most polluted areas are
those with high population density. In these areas, public
transport is a significant contributor to GHG emissions.
Buses typically run in urban areas for approximately 16
hours per day. Furthermore, buses are the most used
mean of public transport [2]. Thus, buses are a critical
candidate for mass electrification.
Hybrid and electric bus integration is a challenging
process. Several studies have pointed out the high initial
Jon Ander López-Ibarra
1,3,♦
, Aitor Milo
1,3
, Haizea Gaztañaga
1
, Victor Isaac Herrera
2,♣
, Haritza
Camblong
3,4,♠
1
IKERLAN Technology Research Centre Energy Storage and Management Area Gipuzkoa, Spain
2
Escuela Superior Politécnica de Chimborazo, Facultad de Informática y Electrónica, Riobamba, Ecuador
3
University of the Basque Country Gipuzkoa, Spain
4
ESTIA Research, France
Email:
♦
jonander.lopez@ikerlan.es,
♣
isaac.herrera@espoch.edu.ec,
♠
aritza.camblong@ehu.eus
Fecha de Recepción: 16 – May – 2019 Fecha de Aceptación: 13 – Jun – 2019
Revista Técnico - Cientíca PERSPECTIVAS
Volumen 1, Número 2. (Julio - Dicimbre 2019)
e -ISSN: 2661-6688
2
investment cost in comparison to conventional buses [3],
[4]. To overcome this problem and to create a more
attractive option for investors, total cost of ownership
(TCO) is a critical point to focus on. Breaking down
and analyzing TCO factors, it can be seen that despite
the recent decrease in lithium-based battery (BT) prices
(79% since 2010 [5]), BTs still have a significant influ-
ence on TCO, as their cost accounts for approximately
a quarter of the total bus price [5], [6]. Moreover,
BTs have a shorter life than electronic power systems,
needing to be replaced and thus further increasing the
TCO. Therefore, new techniques are needed to maximize
BT life and minimize TCO. A proposed technique is
continuous monitoring of the state of health (SOH) of
the BTs, employing new opportunities for digitalization.
Such monitoring would enable processing, analysis, and
decision making which would extend BT life.
Digitalization is the process of equipping vehicles
with sensors to collect data, storing this information
in cloud data to be later analyzed [7]. This new trend
enables the energetic operation of each vehicle to be
monitored. Indeed, monitoring the operation of each
vehicle in a fleet makes it possible to widen the scope of
the current EMS (Energy Management Strategy) from a
localized EMS to a fleet level EMS. This new level will
extend flexibility for energy management and improve
energy efficiency across the entire fleet. However, the
main challenge for fleet level EMS and digitalization
is the large volume of data to be managed by fleet
managers. Consequently, new automated and advanced
tools are needed for data analysis and decision making,
a growing area of research [8], [9].
In this regard, a methodology for analyzing, process-
ing, and making decisions based on this processed data
was proposed in [10], to improve the techno-economic
performance of the whole fleet. Focusing on the decision
making element of the methodology as mentioned above,
the contribution of this paper lies in an approach for
managing a BT life system for an entire fleet. This ap-
proach has been addressed in two ways: bus-to-route and
route-to-bus. In the technical literature, a methodology
for managing BT aging at the fleet level has yet to be
presented.
Bus-to-route optimization is focused on the energy
management strategy (EMS) of each bus in the fleet. The
acquired data from the fleet is analyzed in terms of route
energy, operation and BT aging for each urban route.
Based on these analyses, each bus has been classified to
facilitate decision-making. Based on this classification,
the route-to-bus approach was implemented. This tech-
nique is focused on extending the fleet’s BT life.This
Table I: Scenario approach [11].
Parameters Ser Par
Driving Cycle Profiles/Bus Configuration 10 10
Electric Motor Power [kW] 196.5 196.5
Internal Combustion Engine Power [kW] - 160
Genset Power [kW] 160 -
Battery Packs C-rate/Energy [-/kWh] 7/24 7/24
Opportunity Charging Point [kW] 150 150
stage consists of a re-evaluation of the EMS and the re-
organization of buses with the best SOH buses being
allocated to the most energy demanding routes and vice
versa. For ensuring the optimal operation of the re-
organized HEBs (Hybrid Electric Bus) in the fleet, a bus-
to-route re-optimization was carried out, with the updated
BT capacity and urban route. Increases in BT life of up
to 10.7% were obtained with the proposed approach
II. S
CENARIO OVERVIEW
The analysis in this paper was based on the fleet
described in Table I. The fleet is composed of the
following two power-trains, with the respective models
presented in detail in [10], [11].
- Parallel HEB with Battery: a power train pulled with an
internal combustion engine (ICE) and an electric motor,
operated by a BT pack.
- Series HEB with Battery: a power train pulled by an
electric motor powered by a BT pack and a genset (GS).
The urban routes have been generated from a database
of standardized driving cycles. Each generated cycle has
been applied to an HEB, as explained in Table I, creating
a simulation of the fleet.
III. F
LEET MANAGEMENT BASED ON FLEET
LEARNING METHODOLOGY
The proposed fleet learning methodology is shown
in Figure 1, aiming to improve the techno-economic
performance of the whole fleet, reducing the operation
and maintenance costs [10]. This methodology is split
into four stages:
1. Design Stage:In this first stage, the bus EMS is
developed, based on off-line dynamic programming (DP)
optimization of the corresponding route for each line
[12]. The optimization problem is based on the following
cost function (J):
J =
N−1
i=0
m
f
(U(i)) · T
s
(1)
Revista Técnico - Cientíca PERSPECTIVAS
Volumen 1, Número 2. (Julio - Dicimbre 2019)
e -ISSN: 2661-6688
3
Figure 1: Proposed Fleet Learning methodology [10].
where
m
f
· T
s
is the fuel mass consumption at each
time step (
T
s
=1 s), determined by the torque (parallel
configuration) or power (series configuration) split factor
U, within the urban route length (N).
2. Real Operational Behavior:Once the EMSs are imple-
mented for each bus, at this stage the data is simulated
or extracted from a real fleet. For this study, the HEBs
mentioned above have been modeled and simulated in
MATLAB.
A daily trip consists of approximately 16 hours, with
a yearly operation of 300 days. To simulate real driving
behavior, some disruptions were randomly introduced to
these routes. For this scenario, road, auxiliary power, and
passenger disruptions were all considered [10].
3. Intelligent Fleet Manager: The third stage is aimed at
managing data. The acquired data is collected, processed
and analyzed in order to make decisions based on this
analysis and thus update the EMSs or re-organize the
fleet.
4. Fleet Learning Period: In this last stage, the decisions
taken are evaluated in terms of overall fleet efficiency.
I V. B
US-TO-ROUTE AND ROUTE-TO-BUS
Based on the proposed methodology, the contribution
of this paper lies in an approach for extending the BT
life system for a whole fleet. This approach, as shown in
Figure 2, has been addressed from two points of views,
the bus-to-route and the route-to-bus approaches. In the
following lines, the applied methodology is explained.
Stage 1: Design Stages
In the first stage the fleet bus-to-route optimization
was implemented. This optimization is focused on the
EMS of each bus in the fleet (as shown in Table I). The
EMS is based on that proposed in [10].
From the aforementioned paper [10], it was concluded
that global optimization based on DP does not harness
BT utilization. The reason for this lies in the cost function
(Eq. 1). This function is designed to minimize global fuel
consumption, with the constraint of starting and finishing
at the same state of charge (SOC) [13]. Consequently,
the optimization tends to under-utilize the BT, to avoid
extra-fuel consumption of the GS or ICE recharging the
BT.
The maximum and minimum SOCs are extracted
based on the optimal operation of each line. These
limits are used in the EMS to maintain the SOC within
limits. To harness BT consumption, these limits are set
according to the following equations:
SOC
max
= SOC
DP max
SOC
min
= SOC
DP min
· γ
(2)
where
γ is a constant defined for the depth of discharge
(DOD) management, which will enable BT utilization to
increase.
Stage 2: Real Operational Behavior
Based on the optimized routes, at this stage, a MAT-
LAB simulation of the fleet was carried out, as noted
in Section III Stage 2. Figure 3-A shows an example
of a simulated 7 day observation period. A full day
operation profile (24 hours) has been simulated, with a
depot charge at the end of the day. The available time to
charge the bus at the depot varies according to the daily
operation time, and the power value has been set as the
minimum to recharge until the SOC is 85%.
Daily operation is shown in 3-B , in which an
opportunity fast charging point of 150 kW is placed at
the terminal station for each round trip line, charging the
bus for 2.5 minutes. These charging points allow buses
to start and finish the day at the same SOC.
Stage 3: Intelligent Fleet Manager
This stage is where the novelty of the paper lies.
The obtained data is processed for subsequent fleet
data analysis. Based on this analysis, EMS updating
and/or the route-to-bus approach is implemented. A more
detailed description is given below.
Stage 3.1: Fleet Data Analysis
Revista Técnico - Cientíca PERSPECTIVAS
Volumen 1, Número 2. (Julio - Dicimbre 2019)
e -ISSN: 2661-6688
4
Figure 2: Bus-to-Route and Route-to-Bus battery lifetime methodology.
At this stage, an analysis of the whole fleet is carried
out, to facilitate the resulting decisions. The different bus
lines are analyzed based on the routes and fleet operation
analysis. Consequently, the BT aging evaluation plan is
designed, based on the BT aging analysis.
Routes and Fleet Operation Analysis
First, to extract information from the bus lines, an
energy evaluation is performed. The evaluated terms
are related to bus dynamics, intending to evaluate the
routes objectively, without taking into account vehicle
configuration. Consequently, the analyzed parameters are
mean speed, energy consumption per kilometer, and
aggressiveness (A), which is calculated as follows [14]:
A =
(a · v)dt
(v)
a>0
(3)
where
a is acceleration [m/s
2
] and v is speed [m/s].
In order to evaluate the operation of the fleet and the
Figure 3: A- Evaluated weekly driving SOC operation.
B- Daily SOC operation.
EMS behavior on each line, daily operation costs are
calculated. The evaluated operation parameters are mean
fuel consumption, BT aging, and mean recharging costs.
Battery Aging Analysis
For the fleet BT aging analysis, first, the optimized
operation based on DP is evaluated in terms of BT
aging. From this evaluation, the fleet BT aging prediction
(
max
Ψ
) aand the fleet evaluation point (P ) are deter-
mined, calculated as follows:
P =
min(Ψ
1
, Ψ
2
, ...Ψ
n
)
2
(4)
max
Ψ
=
n
k=1
(Ψ
1
, Ψ
2
, ...Ψ
n
)
n
(5)
where
Ψ represents the BT aging lifetime and n the
number of lines.
Once these points are defined, the data obtained from
the real driving simulation is processed until point
P .
From this evaluation, the SOH of each bus is calculated.
BBased on the analysis outlined above, three groups
are classified. This classification groups the routes with
the best SOH, the worst SOH and the routes with a
similar SOH.
Stage 3.2: Decision Maker
Route-to-bus
At this stage, the route-to-bus decision maker ap-
proach is implemented. This stage aims to extend BT life
for the whole fleet. Therefore, at this point a decision is
made, whether this is a re-evaluation of the EMS (for
those buses with a similar SOH) or a re-organization of
the buses (for those buses with the best and worst SOHs),
based on the classification of the SOH of buses. Firstly,
Revista Técnico - Cientíca PERSPECTIVAS
Volumen 1, Número 2. (Julio - Dicimbre 2019)
e -ISSN: 2661-6688
5
Figure 4: Bus routes evaluation results.
buses with the best SOH are swapped with the ones with
the worst SOH. Secondly, routes with similar SOH are
re- optimized.
Stage 4: Fleet Learning Period
Re-optimization
According to the decision taken for each bus, in this
stage, the EMS is updated. The EMS update for buses
with a similar SOH is based on the re-adjustment. For
the buses that are swapped to other lines, the EMS is
re-optimized with the updated BT capacity and urban
route. Finally, the obtained BT end-of-life points, and
the operation are analyzed to evaluate the proposed
approach.
V. R
ESULTS AND ANALYSIS
In order to validate the proposed bus-to-route and
route-to-bus approaches, the fleet presented in Table I
has been simulated as described in IV Stage 1.
The obtained results have been presented in four
subsections, following the battery extension approach
explained in Fig. 2.
A. Routes and fleet operation analysis
In this subsection, the energetic evaluation of the
routes and the fleet operation results are presented.
The bus routes energetic evaluation results are shown
in Fig. 4. The energetic evaluation has been performed
analyzing the correlation between the variables as men-
tioned earlier (explained in Section IV Stage 3.1). It is
noteworthy that the energy demand versus aggressive-
ness and versus mean speed has a positive correlation.
Therefore, these two correlations are worth mentioning
as indicators for the energetic classification of the routes.
Evaluating all the different lines energetically provides an
overview of the scenario, helping to define the EMS for
each line.
Fig. 5 depicts the operation costs of each line with the
EMS developed from the fleet bus-to-route optimization.
This analysis helps to evaluate each line economically,
providing a clear picture of the case scenario. The most
significant factor impacting on HEB operation costs is
fuel consumption, with lines 2 and 12 reflecting the
highest operating costs. However, the goal is to improve
TCO for the whole fleet, so all factors have to be con-
sidered in order to improve the variables to be managed.
The information from the battery and recharged energy
indicators are noteworthy in this study. These indicators
help to identify the lines that are most demanding in
terms of battery. Lines 11, 15, and 20 have the highest
BT costs.
B. Battery Aging Analysis
The BT aging analysis was performed from different
points of view. The observation period for data collection
lasted for seven days (Fig. 3) and evaluated up until the
fleet evaluation point with this operation.
Figure 6 shows the obtained BT aging results from the
DP optimization and the ”real” driving (RD) behavior.
From the obtained DP optimal operation of each line,
point P and
max
Ψ
have been determined based on Eqs.
4 and 5 respectively. Point
P has been used as the base-
line for the BT lifetime estimation reference.
max
Ψ
has
been determined a 131% greater than
P reference point.
It is noteworthy that the DP results predict shorter
BT lifetimes in the case of the parallel configuration
Figure 5: Bus lines operation costs.
Revista Técnico - Cientíca PERSPECTIVAS
Volumen 1, Número 2. (Julio - Dicimbre 2019)
e -ISSN: 2661-6688
6
(decrease of 20.3%) and similar lifetimes for the se-
ries configuration (an increase of 1.4%). Analyzing the
obtained results, it has been concluded that the parallel
configuration has higher BT use in the optimal operation
than in the real driving behavior. These lifetime estima-
tions have been calculated with the obtained results from
the one week simulation, replicating this operation until
the BT end-of-life.
A correlation of variables analysis was carried out.
From this analysis, the correlation of the BT consump-
tion, daily driven distance and BT aging has been found,
as shown in Fig. 7. TThis analysis demonstrates the
correlation of the three mentioned factors, BT aging
decreasing with BT consumption increase and daily dis-
tance increase. This correlation is a useful indicator for
the BT aging pre-analysis, evaluating BT degradability
for the different routes.
The SOH evaluation was the final evaluation. The ob-
tained results are depicted in Fig. 8, with the driven daily
distance and BT consumption of each route. Analyzing
the obtained results, three groups are proposed: the routes
with the best SOH (lines 5, 8 and 10), the routes with the
worst SOH (11, 15 and 20) and the routes with similar
SOH (remaining routes). It can be seen that the SOH is
higher for all series configuration as compared to parallel
configurations.
Further analysis of the obtained SOH results and
comparison of these results with the routes previously
identified as most BT demanding routes in Subsection
V-A, demonstrate that the buses with the worst SOH
match with these.
Figure 6: Battery aging variation.
Figure 7: BT aging correlations.
C. Route-to-bus decision maker
Based on the BT classification from the SOH results,
the route-to-bus decision maker is implemented. OFirstly,
buses in the group with the best SOH are swapped
to higher BT degrading routes and vice-versa. Table II
shows the new fleet outline. EMSs for HEBs are re-
optimized for the corresponding new lines.
Table II: Bus lines re-organization.
Group Bus number Current line Swap to line SOH [%]
Best SOH
5 5 11 90.65
8 8 20 90.51
10 10 15 90.04
Worst SOH
11 11 5 86.51
20 20 8 86.61
15 15 10 87.08
Secondly, lines within the similar SOH group are re-
optimized, re-determining the
γ (see Section IV Stage 1)
constant.
Figure 8: HEBs battery SOH at the evaluation point.
Revista Técnico - Cientíca PERSPECTIVAS
Volumen 1, Número 2. (Julio - Dicimbre 2019)
e -ISSN: 2661-6688
7
D. Re-optimized buses results and analysis
Finally, the re-organized scenario is evaluated, in
terms of BT aging and fuel consumption.
Fig. 9 depicts the obtained new BT aging results
from swapping the bus lines, as described in Table II.
The series configuration buses have shown a decrease in
BT life of up to 8.51%, as they were swapped to the
most demanding lines. On the contrary, the tendency for
parallel configuration buses has been to increase BT life
by up to 6.13%.
The new fleet BT aging layout is presented in Table
III. For most of the lines, after applying the proposed
approach, BT life has tended to increase. Buses without
a remarkable aging increase have not been updated.
The only buses with a BT life decrease have been the
aforementioned series configuration buses (buses 5, 8 and
10), as they were swapped to more demanding routes.
From the overall results, it is noteworthy the highest
increase for BT life for the series configuration was an
increase of 10.7%. The reason for this is that there is
greater flexibility for charging the BT than in the parallel
configuration (where the highest increase reached is up
to 6.13%).
Finally, to evaluate the change in fuel consumption
after the route-to-bus approach, the obtained variations
are presented in Fig. 10. There is a slight 1.1% increase
in fuel utilization from the pre-update scenario (before
the route-to-bus approach application) to the post-update
scenario (after the route-to-bus approach was applied).
This is considered normal behavior, since to manage BT
aging, fuel utilization increases. However, it is possible
reach a trade-off between fuel and BT utilization.
Figure 9: Bus lines re-organization lifetime estimations.
Table III: Bus lines new layout regarding BT aging.
Aging Prediction
Evaluations
Aging Prediction
Evaluations
Bus
1. 2.
Diff.
[%]
Bus
1. 2.
Diff.
[%]
1 2.43 2.69 +10.7 11 1.63 1.68 +3.07
2 2.39 2.5 +4.6
12 2.02 2.02 0
3 2.56 2.8 +9.38
13 2.02 2.03 +0.5
4 2.37 2.47 +4.22
14 1.98 1.98 0
5 2.91 2.89 -0.69
15 1.71 1.79 +4.68
6 2.53 2.58 +2
16 1.77 1.77 0
7 2.19 2.2 +0.46
17 1.87 1.88 +0.53
8 2.82 2.58 -8.51
18 1.87 1.87 0
9 2.54 2.54 0
19 1.77 1.77 0
10 2.62 2.53 -3.44
20 1.63 1.73 +6.13
VI. CONCLUSIONS
In this paper, a bus-to-route and route-to-bus approach
was presented, oriented to extending BT life. In order to
validate the proposed approach, a simulation of the fleet
was carried out, as presented in Table I.
The BT lifetime increase with the proposed method-
ology is up to 10.7%. It was possible to obtain higher BT
lifetimes for the series configuration than for the parallel
configuration. The main reason for this lies in the higher
flexibility for recharging BT with the GS for the series
HEBs than for the parallel HEBs.
The performed route-to-bus re-organization approach
compensates for BT aging imbalances. Therefore, the
buses with the best SOH have been swapped to the
most demanding lines and the buses with the worst SOH
have been swapped to the least demanding lines. As a
result, the HEBs swapped to the most demanding lines
show a decrease in BT life (up to an 8.51%), and those
moved to the least demanding lines show an increase in
BT life (up to 6.13%). For correct evaluation, the mean
Figure 10: Fleet fuel consumption evolution.
Revista Técnico - Cientíca PERSPECTIVAS
Volumen 1, Número 2. (Julio - Dicimbre 2019)
e -ISSN: 2661-6688
8
fuel consumption increase for the fleet was evaluated,
obtaining an increase of 1.1% after the route- to-bus
application.
Furthermore, the correlations for the bus lines evalua-
tion and the BT aging have been analyzed, to facilitate re-
organization of the buses. Firstly, for route correlations,
the energy demand versus aggressiveness and versus
mean speed is noteworthy. Secondly, in terms of BT
aging correlations, BT consumption and the driven daily
distance was identified.
Future research will focus on developing a clustering
classification for route-to-bus automation.
R
EFERENCES
[1] B. Eickhout. Heavy-duty vehicles co2 emissions and fuel ef-
ficiency. Available: http://www.europarl.europa.eu/legislative-
train/theme-resilient-energy-union-with-a-climate-change-
policy/file-heavy-duty-vehicles-co2-emissions-and-fuel-
efficiency, 20-04-2018 [Accessed: 11-09-2018].
[2] M. Steriu, “Local public transport in the european union,”
UITP. Advancing Public Transpot, Tech. Rep., 2016.
[3] L. Nurhadi, S. Bor
´
en, and H. Ny, “A sensitivity analysis of
total cost of ownership for electric public bus transport systems
in swedish medium sized cities,” Transportation Research
Procedia, vol. 3, no. July, pp. 818–827, 2014. [Online].
Available: http://dx.doi.org/10.1016/j.trpro.2014.10.058
[4] A. Lajunen, “Energy consumption and cost-benefit analysis of
hybrid and electric city buses,” Transportation Research Part
C: Emerging Technologies, vol. 38, pp. 1–15, 2014.
[5] J. F. C. M. Aleksandra O’Donovan, “Electric buses in cities.
driving towards cleaner air and lower co2,” Bloomberg New
Energy Finance, Tech. Rep., 2018.
[6] Z. Bi, L. Song, R. De Kleine, C. C. Mi, and G. A. Keoleian,
“Plug-in vs. wireless charging: Life cycle energy and
greenhouse gas emissions for an electric bus system,” Applied
Energy, vol. 146, no. February, pp. 11–19, 2015. [Online].
Available: http://dx.doi.org/10.1016/j.apenergy.2015.02.031
[7] M. Mesgarpour, D. Landa-Silva, and I. Dickinson, “Overview
of Telematics-Based Prognostics and Health Management Sys-
tems for Commercial Vehicles,” Communications in Computer
and Information Science, vol. 395, pp. 123–130, 2013.
[8] C. Marina Martinez, X. Hu, D. Cao, E. Velenis, B. Gao,
and M. Wellers, “Energy Management in Plug-in Hybrid
Electric Vehicles: Recent Progress and a Connected Vehicles
Perspective,” IEEE Transactions on Vehicular Technology,
vol. 66, no. 6, pp. 4534–4549, 2017. [Online]. Available:
http://ieeexplore.ieee.org/document/7496906/
[9] S. Mehar, S. Zeadally, G. Remy, and S. M. Senouci, “Sustain-
able Transportation Management System for a Fleet of Electric
Vehicles,” IEEE Transactions on Intelligent Transportation
Systems, vol. 16, no. 3, pp. 1401–1414, 2015. [Online].
Available: http://ieeexplore.ieee.org/document/6975190/
[10] J. A. Lopez, V. I. Herrera, A. Milo, and H. Gazta
˜
naga,
“Energy Management improvement based on Fleet
Learning for Hybrid Electric Buses,” in IEEE Vehicle
Power and Propulsion Conference (VPPC). Chicago, IL,
USA, USA: IEEE, 2018, pp. 2–7. [Online]. Available:
https://ieeexplore.ieee.org/document/8605025
[11] V. I. Herrera, A. Milo, H. Gazta
˜
naga, J. Ramos, and H. Camb-
long, “Adaptive and non-adaptive strategies for optimal energy
management and sizing of a dual storage system in a hybrid
electric bus,” IEEE Transactions on Intelligent Transportation
Systems, pp. 1–13, 2018.
[12] O. Sundstr
¨
om and L. Guzzella, “A Generic Dynamic Program-
ming Matlab Function,” 18th IEEE International Conference
on Control Applications, Saint Petersburg, Russia, no. 7, pp.
1625–1630, 2009.
[13] A. Sciarretta and L. Guzzella, “Control of Hybrid Electric
Vehicles,” IEEE Control Systems Magazine, vol. 27, no. April,
pp. 60–70, 2007.
[14] A. Lajunen, K. Kivek
¨
as, F. Baldi, J. Veps
¨
al
¨
ainen, and
K. Tammi, “Different approaches to improve energy consump-
tion of battery electric buses,” Vehicle Power and Propulsion
Conference 2018, 2018.
Revista Técnico - Cientíca PERSPECTIVAS
Volumen 1, Número 2. (Julio - Dicimbre 2019)
e -ISSN: 2661-6688
