Traffic Modeling of A Cooperative Charge While Driving System in A Freight Transport Scenario

Available online at www.sciencedirect.
com
ScienceDirect
Transportation Research Procedia 6 (2015) 325 – 350
4th International Symposium of Transport Simulation-ISTS’14, 1-4 June 2014, Corsica, France
Traffic Modeling of a Cooperative Charge While Driving System in

a Freight Transport Scenario
Francesco Deflorioa *, Luca Castelloa
a
DIATI, Politecnico di Torino, corso Duca degli Abruzzi 24, 10129 Torino, Italy
Abstract
The aim of this paper is to present a research study on a traffic model developed for analysing the performance of the wireless
inductive systems for charging while driving (CWD) fully electric vehicles (FEVs) from both traffic and energy points of view.
The design assumptions of the developed traffic model are aimed to simulate in particular a freight distribution service in a fully
cooperative traffic environment. In this case, the CWD service could be used to guarantee the minimum state of charge (SOC) of
the batteries at the arrival to the depot that allows the vehicles to shortly start with further activities. In this way, the fleet
manager could avoid wasting time for the stationary recharge, thus increasing the level of service of the freight distribution.
The CWD system is applied to a multilane ring road with several intermediate on-ramp entrances, where the slowest lane is
reserved for the dynamic charging activities, when authorized vehicles are present. A specific traffic model has been developed
and implemented adopting a mesoscopic approach, where vehicle energy needs and charging opportunities affect drivers’
behavior. Overtaking maneuvers, as well as new entries in the CWD lane of vehicles that need to charge, have been modeled by
taking into account a fully cooperative driving system among vehicles which manages adequate gaps between consecutive
vehicles. Finally, a speed control strategy in which vehicles can be delayed to create an empty time-space slot in the CWD lane,
is simulated at a defined node. This type of control, though is simulated to allow extraordinary maintenance operations, which
may require a free charging zone for a given time slot, could also be applied to support merging maneuvers for on ramp vehicles.
© 2015
© 2015TheTheAuthors.
Authors. Published
Published by Elsevier
by Elsevier B.V.is an open access article under the CC BY-NC-ND license
B.V. This
Selection and/or peer-review under responsibility of the Organizing Committee of ISTS’14.
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Selection and/or peer-review under responsibility of the Organizing Committee of ISTS’14
Keywords: Charging While Driving; traffic simulation; multi-lane road modeling; electric freight vehicles
* Corresponding author. Tel.: +39-011-0905601; fax: +39-011-0905699.

E-mail address: [email protected]
2352-1465 © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Selection and/or peer-review under responsibility of the Organizing Committee of ISTS’14
doi:10.1016/j.trpro.2015.03.025
326 Francesco Deflorio and Luca Castello / Transportation Research Procedia 6 (2015) 325 – 350
1. Introduction
The majority of fully electric vehicles (FEVs) currently satisfies the electric energy needs for motion with an on-
board battery. The extensive literature on FEVs includes discussions of the following: battery problems particularly
concerning limitations in size and power, their weight, life and recharge time and the lack of charging points. These
problems are even more relevant for freight distribution services, where the masses and the distances travelled are
relevant and where the stationary recharge would require many charging stations. However, electric vehicles could
represent one of the possible solutions to low air pollutants emissions in the city centers, where a freight distribution
service often has to deliver. For this reason, the charge while driving (CWD) system could represent a technology to
contain the batteries sizes and the recharging infrastructures costs without impacting on the vehicles autonomy.
In particular, Boulanger at al. (2011) analysed the problems related to battery charging management, the
uncertainty surrounding the monitoring of the state of charge (SOC), the limited availability of charging
infrastructure and the long time required to recharge; problems that have generated range anxiety. The use of
intelligent transport systems (ITS), in particular vehicle to vehicle or to infrastructure (V2X) communications, was
evaluated to allow drivers to accurately and confidently locate charging stations where they could recharge the
battery in the shortest amount of time (Ezell, 2010). Johnson at al. (2013) evaluated how connected vehicle
technologies can facilitate the rapid charging of FEVs at charging stations throughout the road network. The market
acceptance of FEVs, travel needs and consumer choices, particularly for the first car in the household, were also
analysed by Kirsch (2000). Moreover, extensive research has claimed that the challenges of battery inefficiency and
the large and wasted space in the FEVs can be overcome by the wireless power transfer (WPT) technology. This
technology electrically conducts energy from a source to an electric device without any interconnecting mediums
(Palakon et al, 2011). Finally, another important element that supports a battery charging modality with frequent and
low energy transfer while driving is that the SOC must be managed carefully and the batteries should never be fully
discharged to avoid an excessive shortening of the battery life cycle.
The aim of this research study is to provide a method to support preliminary studies about one of the possible
future technology that could contribute to use FEVs in freight transport. One of the goals of the European white
paper is to achieve essentially CO2-free city logistics in major urban centers by 2030 and CWD technology could
represent one of the possible ways to reach this target. The model can simulate the performance of the wireless
inductive charge of the electric vehicles while driving, from both traffic and energy points of view. Beginning with
an electric vehicle supply equipment (EVSE) layout defined and analysed in a previous study (Deflorio et al, 2013),
a model for the traffic flow simulation is implemented to quantify and describe traffic performance, useful for
drivers and operators, but also the electric power that should be provided by an energy supplier for a proper
management of the charging system.
2. The CWD service for freight vehicles and the technological scenario
The following paragraphs are aimed at providing a brief overview on how a CWD service based on assumptions
and system requirements defined in the eCo-FEV project (2013) operates.
A driver who wants to use the CWD service should send a request, even automatically, to the charging station
operator through an on-board unit (OBU). After the verification of operational requirements, the charging station
operator returns the confirmation message to the user and updates the list of authorized vehicles, while the driver
receives the authentication on the OBU. The CWD system should include some enforcement functions to prevent
unauthorized vehicles from using the reserved lane. The position of each authorized vehicle is monitored along the
CWD lane to switch on only the coils under the vehicles, thus avoiding energy wasting. The correct position
monitoring is important also for vehicles outside the CWD lane because they affect overtaking maneuvers. The
service should consider different classes of vehicles, according to their energy needs. Different speeds should
therefore be admitted in the CWD lane and overtaking maneuvers should be managed according to a cooperative
system among vehicles: advanced driver assistance systems (ADAS) guide drivers during their maneuvers and the
cooperative system eventually intervenes on vehicle speeds to maintain the required gaps in the vehicle flow. This
technology could serve both passenger vehicles – private or public – and freight vehicles for distribution services, on
which this paper is focused.
Francesco Deflorio and Luca Castello / Transportation Research Procedia 6 (2015) 325 – 350 327
The chosen example of freight distribution service starts from a depot quite close to the city center and connected
to the motorway network by the ring road. The primary actors involved in the delivery use case (Fig. 1) can be
described as:
x FEV-Private: it is an electric powered vehicle which provides to the ITS system its level of autonomy as well as
other in-vehicle information;
x Charging station operator: it is the operator that provides charging services to FEVs and manages the EVSE.
This operator provides also traffic information in the inductive charging lanes. If necessary, it offers booking and
payment facilities;
x Fleet operator drivers: they drive FEVs of the fleet and are usually provided with an established delivery
schedule;
x Info operator: this operator monitors and manages the road traffic and/or road side infrastructure systems. It
provides information on traffic conditions and events (accidents, road works, road diversions, traffic restrictions,
etc.). It may refer to an urban traffic management center or a highway operator.
Fig. 1. Example of a use case diagram for a freight distribution service with FEV.
The CWD is here supposed to be installed only along the right-hand lane of the ring road because that lane is
generally used by slower vehicles (Fig. 2). The EVSE includes inductive coils placed under the pavement surface,
forming different charging zones (CZs) at a relative distance (I). An authorized FEV that crosses the CZ couples to
its high frequency alternating magnetic field and the power is transferred to charge the battery.
Fig. 2. CWD system layout.

3. The choice of the traffic model
The presented methodology refers to a freight distribution service in a near future mobility scenario, where
vehicles motion could be less influenced by driver decisions. In particular, the rules that manage the vehicle
behavior are here developed according to a fully cooperative system where ADAS operate both on longitudinal
vehicle behavior - by speed control functions - and lateral vehicle behavior - by lane selection functions -.
The choice of the traffic modeling is derived from two primary requirements of the CWD system (eCo-FEV,
2013):
x CWD is supposed to be installed only along the right-hand lane of the ring road because that lane is generally
used by slower vehicles: the traffic model must consider the disaggregation of traffic data per lane;
x CWD can be used by vehicles with different charging needs and speeds: the traffic model must consider different
classes of vehicles, related to their electric charge level.
An extensive review of traffic modeling approaches can be found in Hoogendoorn and Bovy (2001).
Macroscopic models are not adequate to describe this type of problem because they lose the single vehicle
information and describe the mean behavior of the traffic flow on a road section. One possible approach to
effectively model multilane and multiclass problems could be microsimulation, where single vehicle trajectories and
interactions are modeled with a small time step resolution. A microsimulation model application example is reported
by Barceló et al. (2005). Although the microsimulation approach meets the principal requirements of the traffic
model for CWD, it does not model vehicle behavior according to energy needs. The current SOC of the vehicles
influence drivers’ decisions concerning lane changing behavior, i.e., vehicles try to enter or exit the CWD lane
according to their needs. Therefore, specific rules must be defined to obtain useful results from the traffic model.
In addition, the detailed rules implemented in a microsimulation model usually require an accurate calibration
process, aimed at replicating the actual driver behavior in traffic. However, currently the CWD system has been
installed only in small test sites and, unfortunately, there are no opportunities to observe driver behavior in large-
scale systems. Furthermore, even fully cooperative driving systems are not completely deployed. The most similar
case in an actual traffic scenario can be observed in long road tunnels in which vehicle spacing or headway greater
than a predefined threshold should be maintained and all vehicles travel in a predefined speed range for safety
reasons (e.g. the Mont Blanc tunnel). In such systems, the vehicle behavior is controlled by safety constraints, as in
the CWD model, although there are no interactions between vehicles, such as overtaking maneuvers and new entries
along the lane. Furthermore, the CWD technological environment may expand only in the future, involving another
generation of vehicles in which V2V will be used and many cooperative functions will be activated to facilitate the
drive. An example of a cooperative scenario in highway automation system is reported in Fig. 3.
Fig. 3. Cooperative lane change assistance use case [source: ETSI, 2011].
In such a system, the observation of the current drivers’ behavior is not relevant to model the traffic because
vehicle motions and interactions depend more on the activated functions of the ADAS system than on drivers’
decisions. A calibration process based on empirical observations of the current traffic would be compromised
whenever ADAS and full cooperative systems were considered, because they affect driving and traffic behaviors.
Consequently, a mesoscopic approach would be more appropriate to model the problem because it represents a
good compromise between the detailed resolution of the microscopic simulation and the current preliminary stage of
development of the CWD technology. A framework of mesoscopic traffic models can be found in Cascetta (2001),
whereas a recent application is proposed by Ben-Akiva et al. (2012).
4. The structure of the CWD simulation model
The developed model represents the road infrastructure as a sequence of road segments - in the following
indicated as “sections” - delimitated by “nodes” or “detection points”. Detailed traffic information is updated only at
nodes based on traffic information determined at upstream nodes. Therefore, the iterative algorithm assumes that
traffic conditions along the infrastructure can be described knowing only the data related to consecutive points
whose spacing, typically hundreds of meters, can be set according to the specific requirements of the analysis or of
the infrastructure. The vehicle time information is defined only at nodes, also with respect to new entries of vehicles
in the infrastructure. The model reproduces single vehicle trajectories without introducing a detailed time-space
resolution of the driving behavior and it determines aggregated traffic information, such as vehicle counting,
average headways, delays and the number of overtaking maneuvers along the CWD lane for any road section. The
logic scheme adopted for two consecutive nodes is reported in Fig. 4.
Fig. 4. Modeling of vehicle trajectories and possible interactions for two nodes.
Initially, the model estimates the arrival time of each vehicle at node (i) based on its arrival time at node (i-1) and
its speed that depends on the vehicle class. Then, it manages interactions between vehicles caused by new entries in
the CWD lane or by overtaking maneuvers.
Overtaking maneuvers that occur on the node are identified because the faster vehicle is not physically in the
CWD lane and an erroneous vehicle count would lead to an overestimation of the energy required by the CZ on the
samenode. Overtaking maneuvers are managed according to a cooperative driving model at constant speed and
vehicles do not recharge while they are outside the dynamic charging lane.
New vehicles in the CWD lane are caused by new entries in the infrastructure, as represented in Fig. 4, or by
“out” vehicles that move into the CWD lane. New entries can cause erroneous relative positioning between vehicles:
if the headway between two vehicles is less than the minimum admissible headway (headway min), the algorithm
corrects the arrival time of the following vehicle by slowing it down. In fact, due to safety and maybe technical
reasons, headways less than a threshold value between two vehicles in the CWD lane may not be allowed. Also the
entries in the CWD lane are managed according to a cooperative behavior: each vehicle that needs to recharge its
battery is moved into the CWD lane at node, creating the eventual necessary gap in the vehicles flow by slowing
down the following vehicles. The headway verification and correction is therefore performed only at discrete space
steps, according to the mesoscopic modeling of vehicle behavior. In an actual scenario, this process can be managed
by drivers or by the cooperative system adapting the vehicle speed along the entire section before the node where
the headway adjustment is performed.
The battery SOC, monitored along the road at each node, plays a crucial role because it influences drivers’
decisions whether to use the CWD service or not according to their destinations. For the analyzed freight
distribution service, this process can be simplified because all the vehicles have the identical destination and the
decision about charging does not depend on drivers, but on the fleet operator. Indeed, to restart the delivery
operations in the second part of the day, all the vehicles of the fleet may require an energy level adequate for their
operations. Vehicle SOC is also the parameter used to divide vehicles into different speed classes, according to their
recharging needs.
Finally, the algorithm allows the implementation of a speed control strategy, in which vehicles can be slowed
down to create an empty time-space slot in the CWD lane at a defined node. An example of this application is
described by Sun and Chen (2013). This type of control could be applied to support operations that require a free CZ
for a given time slot, as extraordinary maintenance activities of the coils for example. This strategy could be applied
even to support merging maneuvers for vehicles entering from the secondary accesses.
4.1. The model structure
The traffic simulator has been implemented using Visual Basic for Applications (VBA). It is composed of several
sub routines that will be described in detail in the following paragraphs. The algorithm operates according to the
functional scheme reported in Fig. 5: it generates an initial traffic state, modeled by a set of vehicles, and then it
iteratively analyses travel times and energy parameters of all the simulated vehicles node by node. The routine
“Strategy detection” runs only if the strategy is activated. Table 1 and Table 2 collect all input parameters required
by the algorithm with the values used in simulations.
4.1.1. Initial traffic state

The “Initial traffic state” routine operates according to two different approaches: the first one refers to the initial
node (node 0) whereas the other to the secondary accesses.
At node 0, it calculates the maximum speed that the considered FEV can reach: it is the speed at which the
resistance to motion is equal to the maximum power that the FEV engine can provide. The maximum speed is
estimated by an iterative process which progressively reduces the speed by steps, starting from the road speed limit.
The maximum speed of FEVs is then assumed as FEVs free flow speed (v0) and it is used to estimate the average
speed (v) in the unequipped lanes (“out” vehicles), according to the Northwestern speed-density model (Drake et al.
1967):
ͳ ݇ ʹ
‫ ݒ‬ൌ ‫ Ͳݒ‬ή ݁‫ ݌ݔ‬ቈെ ൬ ൰ ቉
ʹ ݇Ͳ (1)
In this simple and reliable traffic model - within the density range used - k and k0 are respectively average and
optimum densities. The relationship between the traffic flow parameters allows to estimate the average entering flow
and consequently the average headway between vehicles. Subsequently, the routine progressively generates the
initial headway of each couple of vehicles using a random algorithm according to the mean value and the standard
deviation of the random distribution (Daganzo, 1995). Headways are then used to calculate vehicle entering times.
Using a similar random algorithm, the routine assigns each vehicle its SOC; the SOC random distribution is limited
at the lower end by positive values and at the upper end by the battery size. According to the generated SOC and to
the introduced SOC thresholds, the routine defines position (“out” or “in”), status (“no charge”, “charge” or “emer”)
and speed of each vehicle.
Fig. 5. Algorithm structure.
Concerning the on ramps along the ring road, the routine receives the node in which each access is located, the
number of entering vehicles and, for sake of simplicity, the time window in which the entries are included as inputs.
Through a uniform distribution, the routine then assigns to each vehicle an entering time included in the time
window, while it sets the vehicle entering SOC according to the random algorithm used for vehicles at node 0.
4.1.2. Strategy detection

The “Strategy detection” routine operates if the strategy is active and only at the iteration corresponding to the
node in which the strategy acts (briefly indicated as strategy node). It sorts the vehicle arrival times in ascending
order and it selects the first vehicle whose arrival time is included in the strategy time window. The selected vehicle
is then slowed down till its arrival time is equal to the strategy end time. Its average speed in the section approaching
the strategy node is then evaluated to estimate the minimum headway that prevents two following vehicles from
overlapping (overlap headway). Following vehicles will be subsequently managed according to “Headway
correction” and “Time estimation next” routines, as explained in the next paragraphs.
4.1.3. Headway correction

The “Headway correction” routine calculates the headways between vehicles in the CWD lane and it compares
them to the minimum technical headway. Unfeasible headways for CWD can be generated by entering maneuvers:
if the headway between two vehicles is less than the acceptable limit, the routine corrects the arrival time of the
following vehicle at the detection point by slowing it down.
If the strategy is active, its duration considerably affects vehicle speed reductions. For low speed values, the
minimum technical headway in the CWD lane could be no longer adequate to guarantee the safety gap between two
following vehicles. The time required to a vehicle to cover its length plus the safety spacing could be higher than the
technical headway defined for operative speeds in the CWD lane. For this reason, the routine compares the technical
headway to the overlap headway previously defined: the minimum headway allowed on the strategy node is the
bigger between those two values. This headway regulation is applied to the time window comprised between the
initial time of the strategy and the end queue time, that is the time in which the propagation of the strategy effects, in
terms of vehicle platooning, are supposed to be ended, according to the calculations developed for the considered
node by the “Time estimation next” routine.
For each simulated vehicle, the “Headway correction” routine also evaluates other parameters which can be used
for internal checks, such as IDs of the previous and the following vehicles and the headway with the following
vehicle.
4.1.4. Overtaking at node

The “Overtaking at node” routine identifies vehicles that are overtaking at detection points. The routine analyses
couples of vehicles with different speeds. If the headway between these vehicles is less than half of the overtaking
maneuver duration, the faster vehicle is identified as overtaking. This function is useful to detect vehicles only if
they actually cross a detection point in the CWD lane: in this way, more realistic time profiles of the energy that
should be provided by the electric line on CZs are obtained. The implemented overtaking model refers to an
overtaking maneuver at constant speed developed in three phases: 1) lane changing from CWD lane, 2) overtaking
of the slower vehicle, 3) lane changing to CWD lane. The durations of the first and third phases are assumed equal
to 4 s, according to the prescriptions of the Italian technical standards for the design of road infrastructures (DM.
6792/2001). This value is also supported by Farah researches (2010, 2013). The second phase duration is the time
required to cover the length of the two vehicles at a speed equal to the difference between average speeds of vehicles.
4.1.5. SOC estimation

The “SOC estimation” routine firstly calculates the actual average speed of each vehicle along the last section
after slowdowns generated by headway corrections. The actual average speed is used to update the SOC level of
each vehicle by estimating the energy balance between consumption related to motion and the energy provided by
the coils in the CWD lane.
The implemented model for estimating the vehicle energy consumption is based on the resistance to motion. The
total resistance (Rtot) is given by the following relationship:
ܴ‫ ݐ݋ݐ‬ൌ ܴ݀‫ ݃ܽݎ‬൅ ܴ‫ ݈݈݃݊݅݋ݎ‬൅ ܴ݈ܽܿܿ݁݁‫ ݊݋݅ݐܽݎ‬൅ ܴ‫݁݌݋݈ݏ‬ (2)
Where:
ͳ
ܴ݀‫ ݃ܽݎ‬ൌ ή ߩ ή ܿ‫ ݔ‬ή ‫ ܣ‬ή ܵ ʹ
ʹ (3)
ܴ‫ ݈݈݃݊݅݋ݎ‬ൌ ݉ ή ሺ݂Ͳ ൅ ݂ʹ ή ‫ ʹ ݒ‬ሻ (4)
ܴ݈ܽܿܿ݁݁‫ ݊݋݅ݐܽݎ‬ൌ ݉ ή ܽ (5)
ܴ‫ ݁݌݋݈ݏ‬ൌ ݉ ή ݃ ή ‫݌‬ (6)
Therefore, the total resistance depends on the following parameters: air density (U) [kg/m3], drag coefficient of
the vehicle (cx), cross sectional area of the vehicle (A) [m2], vehicle speed relative to the air (S) [m/s], vehicle mass
(m) [kg], rolling coefficients (f0, f2) [m/s2, 1/m], vehicle average speed (v) [m/s], vehicle acceleration (a) [m/s2] and
average slope of the road (p). For sake of simplicity, the speed relative to the air will be taken as equal the vehicle
average speed whereas the average slope of the road will be assumed as negligible. Finally, the energy consumed by
the vehicle along a section is obtained by multiplying the power required to the engine because of the resistance to
motion to the time necessary to cross the section:
ܴ‫ ݐ݋ݐ‬ή ‫ݒ‬ ‫݊݋݅ݐܿ݁ݏܮ‬

‫ ݀݁݉ݑݏ݊݋ܿܧ‬ൌ ݈ܲ݁݁ܿ‫ ܿ݅ݎݐ‬ή ‫ ݐ‬ൌ ቈ ൅ ܲܽ‫ ݔݑ‬቉ ή
Ʉ݀ ܸ
(7)
Where Kd is the driveline efficiency, Paux [W] is the auxiliary power, that is the power that includes all
consumption not related to the vehicle motion - e.g. lights, air conditioning -, Lsection [km] is the length of the road
section and V [km/h] is the vehicle average speed.
The energy received from coils for vehicles in the CWD lane is strictly related to the system element dimensions
(EVSE layout and on-board devices), the power provided by coils (PCZ) [kW/m] and the occupancy time of the CZ
(tCZ), according to the following relationship:
‫݊݋݅ݐܿ݁ݏܮ‬ ‫݂݂ܼ݁ܥܮ‬
‫ ݀݁ݒ݅݁ܿ݁ݎܧ‬ൌ ܲ ή ݊‫ ܼܥ‬ή ‫ ܼܥݐ‬ൌ ൫ܲܿ‫ ݖ‬ή ‫ ܦܥܮ‬ή Ʉ‫ ݏ‬൯ ή ൬ ൰ή൬ ൰
‫ ܼܥܮ‬൅ ‫ܫ‬ ܸ (8)
Where LCD [m] is the on-board device length , Ks is the system efficiency that depends on the distance between
the coil(s) of the on-board device and the coil(s) of the CZ installed in the road pavement, LCZ [km] is the length of
CZs, I [km] is the inter-distance between CZs and LCZeff [km] is the CZ length in which vehicles effectively
recharge. LCZeff is calculated according to the following relationship:
‫ ݂݂ܼ݁ܥܮ‬ൌ ‫ ܼܥܮ‬െ ܶ‫ ݇ݎ‬ή ‫ܦܥܮ‬ (9)
The coefficient Trk is introduced to take into account the initial and final partial overlaps between coils on the
vehicle and in the pavement that reduce vehicle electric recharges.
4.1.6. SOC update for overtaking

For each section, the “SOC update for overtaking” routine calculates the time in which vehicles cannot recharge
because they are overtaking. Vehicle SOCs at the node are then updated according to the overtaking durations. In
the case of a single vehicle overtaking, the time is estimated according to the overtaking model previously described.
Instead, if the faster vehicle performs an overtaking maneuver of more than one vehicle, the overtaking time is
estimated by taking into account the first and the third lane-changing phases previously described and by calculating
the second phase duration as the ratio between the relative distance of the overtaken vehicles and the difference
between the average speeds of the fast and the slow vehicles. The scheme of both overtaking maneuver types is
reported in Fig. 6.
Fig. 6. Schematization of overtaking maneuvers.

4.1.7. Speed test

“Speed test” is a checking routine for traffic conditions next to saturation. It verifies the consistency between
vehicle actual average speeds, headways and the minimum spacing that avoids two vehicles overlap. The routine
returns an error message if an overlap case is detected. During the experimented simulations this event has never
occurred.
4.1.8. Status estimation next

The “Status estimation next” routine returns the lane position (“in” or “out” the CWD lane) and the status
(“emer”, “charge”, “no charge”) of each vehicle in the downstream section, according to the updated SOC on the
node. By defining its status, the routine assigns to each vehicle its free flow speed, according to recharging speeds
set in the CWD lane.
If the strategy is active, during its activation period, the routine excludes that “out” vehicles may enter in the
CWD lane at the considered node.
4.1.9. Time estimation next

The “Time estimation next” routine projects vehicles at the following detection point, estimating their arrival
times based on free flow speeds related to their status.
If the strategy is active, its effects on traffic are not only limited to the lane closure duration, but extend for a
wider period (recovery period). The propagation end time (end queue time) can be estimated according to the
scheme reported in Fig. 7.
Fig. 7. End queue estimation during the strategy.
The rate of arrivals and departures is estimated based on respectively the number of vehicles arriving on the node
and the minimum headway adopted for the vehicle flow before the node involved in the lane closure. For those
vehicles whose projected time is included between the initial time of the strategy and the end queue time, the routine
can operate according to two different approaches. The first one does not perform any further intervention: vehicle
relative positions at the strategy node are defined according to projections based on two different speed classes. The
second approach performs a standardization of vehicle speeds: all vehicles are projected according to the faster
speed class. This implies that overtaking maneuvers are not performed along the section approaching the strategy
node. Long strategy durations lead to very low average speeds in the approaching section. Therefore, overtaking
maneuvers would involve very slow vehicles, causing safety problems because of extreme interactions between
slow and fast vehicles in “out” lanes. So, the first approach is suggested only for very short time windows. For the
slow vehicle class, the second approach may have another impact: slow vehicles arriving at the strategy node shortly
after the initial time of the strategy can anticipate the lane closure because their speed is set as the one of the fast
vehicle class.
5. Verification and validation processes
An extensive verification process has been performed by analyzing, testing and reviewing activities, according to
the concepts defined in the ECSS (2009) standards. In particular, a technical verification of the model response is
performed based on the following four consecutive test case approach, each one aimed at verifying different aspects:
x Single vehicle: this first stage is devoted to ascertain if the single vehicle motion is correctly simulated, as well as
the relationship between its behavior and its energy needs. First, a verification of the correspondence between the
estimated SOC and the vehicle position, status and speed is performed. Then the verification of the accuracy of
the travel time based on the vehicle speeds is assessed. Finally, a verification of the coherence between the
implemented energy model and the vehicle SOC trend at each node is conducted.
x Uniform vehicle flow without overtakes: this second stage of the consistency verification of the model is
developed to assess if the model is able to correctly manage the headways between vehicles, even in the case of
new entries.
x Complex traffic interaction with overtaking maneuvers: the third stage aims to assess the global interaction
between vehicles, introducing overtaking maneuvers. The model has then been tested in a scenario in which
overtaking maneuvers are feasible. The number of overtaking maneuvers per section, the identification of the
vehicles that are overtaking on the nodes, the time required for the maneuvers and its influence on the SOC trend
because of the missed recharge have been verified.
x Strategy activation: the last stage checks the consistency of the strategy simulation results in case of incident. A
general case of traffic is simulated and vehicle trajectories in the space-time diagram are verified in the
neighborhood of the strategy node.
At this stage of the CWD development, the presented model has been validated by checking the satisfaction of
the established technical requirements, based on the System Engineering approach (INCOSE, 2011). The primary
functional requirements used for the validation of the model are the following:
x the model shall estimate the number of vehicles in the CWD lane for any detection point;
x the model shall consider possible random effects of input flows;
x the model shall represent the traffic flow at any detection point and reveal if concentration of traffic and
congestion occur along the lane;
x the model shall take into account different values of the minimum headway allowed in the CWD to estimate
possible effects on traffic and energy for the various CZs over time;
x the model shall consider also on ramp input flows along the CWD lane to assess possible interactions over time;
x the model shall consider the effects of different scheduling plans on the SOC of the vehicle fleet at the arrival at
the depot;
x the model shall involve the possibility of a temporary closure of the CWD lane, without interrupting the service
to FEVs.
In the following chapters, the model testing results are reported in an “ideal case”, in which all of the subsystems
and applications involved, such as the CWD booking and authorization functions, or the cooperative ADAS, which
enables the vehicle cruise control or the cooperative overtaking, work properly. In this scenario, all related system
information, such as the vehicle position and its SOC, is accurately known. This validation approach could be
considered as a “best-case” testing and it is consistent with the test-case-design methods applied to test software,
such as boundary value analysis (Myers et al., 2004), or distributed real time systems (Gutiérrez et al., 1998).
6. The CWD scenario and experimental results
The model is applied to the case of a freight distribution service in the city of Turin. After completing their
routes, vehicles have to come back to the depot to start a new service in the same day. For this reason, the CWD
system is here implemented on the ring road of Turin (Fig. 8), reserving the slower lane to charging activities. The
first section in the road model, between nodes 0 and 1, divides the traffic flow between “in” and “out” vehicles and
it is not equipped with coils.
Fig. 8. Turin ring road with main secondary accesses.
The considered part of the ring road is characterised by the primary stream and four secondary on ramps, for an
overall distance of 18 km. In the road model, all the road sections between nodes have identical lengths, set as 1 km.
For each on ramp, an entering traffic flow will be simulated thus increasing the number of vehicles that use the
CWD service toward the depot. In detail, all the algorithm input parameters are listed in Table 1 whereas on ramp
input flows are reported in Table 2.
Table 1. Inputs to the algorithm.
Type Input Value Unit

Simulation Number of replications 50
Strategy no/yes
Speed control application yes
Strategy node 16
Strategy application initial time 1080 s
Strategy application duration 120 s
Traffic Average density of the input traffic flow 20 veh/km
Critical density corresponding to maximum capacity 30 veh/km
Coefficient of variation of the headway distribution 0.3
Maximum traffic headway 1.5 s
Infrastructure Total length of the road 18 km
Section length 1 km
Length of the CZs 20 m

Inter-distance between CZs 30 m
System efficiency 0.85
Power provided per unit of length 50 kW/m
Minimum technical headway in CWD lane 2 s
Speed limit 110 km/h
Vehicle Average initial SOC 10 kWh
Standard deviation of the SOC distribution 5
Length of the on-board charging device 1 m
SOC upper threshold for “charge” vehicles 12 kWh
SOC upper threshold for “emer” vehicles 6 kWh
Free flow speed for “charge” vehicles 60 km/h
Free flow speed for “emer” vehicles 30 km/h
Length 6 m
Mass 2500 kg
Cross sectional area 4.9 m2
Cx 0.38
f0 0.12 m/s2
f2 0.000005 1/m
Driveline efficiency 0.75
Maximum power provided by the engine 60 kW
Auxiliary power 0.8 kW
Table 2. Input FEVs at each access
Access node Number of simulated FEVs

0 50
8 50
10 50
13 50
14 50
In particular, this paper describes a traffic model in which FEVs energy requirements have been modeled to take
into account the fleet operator need to minimize the recharging time at the depot. At their arrival at the depot, all
vehicles must have a minimum SOC level to restart further activities without wasting time for the stationary
recharge. In this way, the fleet operator could maintain its level of service, saving the resources needed to install
additional structures for vehicle stationary recharges. Deliveries are supposed to be handled by only one type of
FEV, a light-van, and the SOC requirement is modeled by taking into account two different classes of speed based
on the entrance SOC levels.
Therefore, for a given CWD system and considering one vehicle type, the only parameter that can affect the
CWD performance is the vehicle speed. Two operational speed values are implemented, according to previous
experiments (Deflorio et al., 2013). The first one allows vehicles to maintain their SOC: it is the speed adopted by
vehicles with an adequate level of charge with regards to the SOC requirement of the fleet operator. The range of
SOC that satisfies the fleet operator requirement is the one comprised between “charge” and “emer” thresholds.
These vehicles are in “charge” status and their speed is set to 60 km/h. The second speed (30 km/h) guarantees a
proper recharge - 1 kWh after about 2.7 km - and it is adopted by vehicles in “emer” status, whose SOC is low and
minor than the “emer” threshold”.
The CWD system model has been tested in several traffic scenarios to reveal its capability in simulating relevant
effects interesting for CWD operations. The traffic along the CWD lane depends also from the demand structure and
therefore from the input flows entering at the various on ramps of the ring road. For this reason, we have assumed
two reference scheduling plans of the freight distribution service, according to different possible fleet management
strategies.
The first one refers to the case where all vehicles distributing or picking parcels in the various zones of the city
center complete their services almost contemporary. In this case, interactions along the ring road will be less
relevant and arrivals to the depot will be distributed over a wider time period because vehicles start to enter into the
ring road almost simultaneously, but in different nodes. Following this strategy, vehicle delivery missions should
have almost identical durations for the various zones of the city.
In the second scenario, the service management strategy is oriented to concentrate all return trips in a smaller
time window. In this case, vehicles coming from farther zones will enter into the ring road earlier than closer
vehicles. Adopting this strategy, vehicle delivery missions are on average longer for the closer zones of the city,
because FEVs can come back to the depot in a shorter time.
As usual in traffic analyses, results depend on many random factors affecting the model rules at different levels.
In the simulated cases, the focus on random phenomena is related to the features of input flows at various entries.
6.1. Scenario1 - Simultaneous end of service in the city center
The scenario variability is illustrated by Fig. 9 in which the number of vehicles (# veh) using the CWD for the
whole simulation period, detected node by node is reported for 50 replications.
250
# veh
200
150
100
50
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Node
Fig. 9. Variability of the traffic flows in the CWD lane for 50 replications for scenario 1.
The following analysis will focus on one sample replication to better check traffic and energy behaviors in the
implemented simulation model. The replication #16 is selected for its significance because, with respect to the
indicator “# veh”, it is in the range [mean-std, mean+std] for all the nodes as represented in Fig. 10.
250
# veh
200
150
mean
mean-std
mean+std
100 Rep #16
50
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Node
Fig. 10. Sample replication with respect to the replication distribution of Scenario 1.
The first result analyzed of the selected replication is the chart illustrating space-time relationship for all vehicles
using the CWD service, node by node (Fig. 11). This chart describes the use of the various zones of the CWD
system during the period of analysis. Different slopes refer to different speed classes for “emer” and “charge”
vehicles and slope reductions represent speed variations when “emer” vehicles become “charge”. As expected, the
most part of the interactions takes place in the final part of the ring road, during the time window between 300s and
1200s.
2500
Time [s]
2000
1500
1000
500
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Node
Fig. 11. Space-time relationship for "in" vehicles in scenario 1.
To provide a more “traffic oriented” idea on simulation results, the counting and average speeds of the vehicles in
the CWD lane have been reported in the following charts (Fig. 12 and Fig. 13) for a time resolution of 1 minute at
each node of the ring road modeled. The average speed data refers only to “charge” vehicles because the speed of
“emer” vehicles is always 30 km/h and it is not relevant to show.
# vehicles Time [min]
node 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
0 17 15 32
1 10 17 5 32
2 1 11 10 7 5 34
3 2 11 9 1 7 5 35
4 2 11 9 1 7 5 35
5 2 12 9 1 8 4 36
6 4 12 9 1 3 6 3 38
7 1 4 12 9 1 3 2 4 3 39
8 21 12 1 4 12 9 1 3 2 4 3 72
9 15 14 4 2 1 4 12 9 1 3 2 5 2 74
10 15 17 18 8 6 4 4 2 4 12 9 1 3 2 1 5 1 112
11 10 18 24 8 6 4 4 4 4 12 9 1 3 2 1 1 4 1 116
12 11 13 23 14 6 8 4 4 12 9 1 3 2 1 1 1 3 1 117
13 22 12 1 11 13 18 13 6 11 7 4 12 9 1 3 2 1 1 1 1 2 1 152
14 19 25 23 2 11 14 18 8 5 7 6 9 7 12 9 1 3 2 1 1 1 1 2 1 188
15 11 25 19 16 12 14 18 8 7 6 6 5 8 15 9 1 3 2 1 1 1 1 2 1 192
16 1 13 17 19 14 24 15 18 8 2 7 10 5 5 16 11 1 3 2 1 1 1 1 2 1 198
17 1 15 18 13 16 19 24 20 8 2 4 10 8 5 14 11 2 1 3 2 1 1 1 1 2 1 203
18 2 15 19 13 9 22 19 28 9 2 4 7 8 8 14 9 2 3 3 2 1 1 1 1 2 1 205
Total 94 129 154 143 158 140 139 115 122 97 92 63 59 58 58 53 48 42 36 20 14 11 11 9 8 5 6 5 5 5 4 3 3 1 1910
Fig. 12. Count of “in” vehicles, node by node, with a 1 min time resolution in scenario 1.
AvSpeed [km/h] Time [min]

Node 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
1 60 59 60
2 86 62 60 62
3 73 60 60 61
4 60 60 60 30 30 57
5 60 62 60 60 40 30 58
6 73 60 60 60 60 45 61
7 86 60 60 60 60 60 60 61
8 60 60 60 60 60 60 60 60 60 30 59
9 60 59 86 60 60 60 60 60 60 60 60 30 60
10 60 62 64 60 73 73 60 60 60 60 60 60 60 60 62
11 60 64 58 60 60 73 60 60 60 60 60 60 60 60 30 60
12 62 60 59 60 54 60 60 60 60 60 60 60 60 60 60 30 60
13 60 60 86 60 60 60 60 55 60 60 60 60 60 60 60 60 60 60 60 60
14 60 57 56 60 60 61 60 60 86 60 54 60 60 60 60 60 60 60 60 60 60 30 30 59
15 60 58 57 78 60 60 60 60 73 30 60 60 54 58 60 60 60 60 60 60 60 60 60 60 59
16 86 64 60 59 59 58 59 60 60 60 45 55 60 60 58 60 60 60 60 60 60 60 60 60 60 59
17 60 63 61 63 57 56 56 58 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60
18 73 60 61 60 60 60 59 58 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60
Total 60 60 62 60 61 61 60 60 60 60 59 58 57 59 59 59 59 59 60 58 60 60 56 60 55 60 60 60 48 54 60 60 60 60 60
Fig. 13. Average speeds [km/h] of "in" and "charge" vehicles, node by node, with a 1 min time resolution in scenario 1.
The two charts show where and when relevant congestion events are detected by the model - contemporary high
counting values and low average speeds -. In this scenario, only few slowdowns occur and they are quite limited and
observable only in the final part of the ring road - after the node 12 -.
6.2. Scenario2 - Simultaneous arrival at depot
This scenario represents a planning case where the delivery plan has been set with a balanced use of vehicles for
the whole morning period, with almost the identical trip duration for all vehicles. This plan requires shorter delivery
durations for farther zones than for closer ones. According to the approach of scenario 1, the variability of scenario 2
is represented by the number of vehicles in the CWD lane for the whole simulation period, detected for 50
replications node by node. The trend, reported in Fig. 14, is quite similar of that one in Fig. 9, considering that the
total number of simulated vehicles is identical and the only difference is related to the time when vehicles entry in
the CWD lane.
250
# veh
200
150
100
50
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Node
Fig. 14. Variability of the traffic flows in the CWD lane for 50 replications for scenario 2.
The replication selected for its significance is the #2, because it is in the range [mean-std, mean+std] for all the
nodes, as depicted in Fig. 15.
250
# veh
200
150
mean
mean-std
mean+std
100 Rep #2
50
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Node
Fig. 15. Sample replication with respect to the replications distribution of Scenario 2.
The identifying characteristic of this scenario is clearly illustrated in Fig. 16: flows entering from the intermediate
on ramps use the CWD lane when the mainstream flow arrives at the considered access nodes. The effect of the
vehicle concentration in a smaller time period is evident in Fig. 17 reporting vehicle counting for each node for a
time interval of 1 minute.
2500
Time [s]
2000
1500
1000
500
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Node
Fig. 16. Space-time relationship for "in" vehicles in scenario 2.
Count Nv time [min]

Node 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
0 19 13 32
1 12 18 2 32
2 14 11 7 2 34
3 1 14 11 7 2 35
4 1 14 11 1 6 2 35
5 1 1 14 11 1 6 2 36
6 2 1 14 11 1 1 5 2 37
7 2 3 14 11 1 1 6 1 39
8 1 2 3 29 26 1 1 1 5 1 70
9 1 3 3 26 26 3 1 1 1 1 4 1 71
10 1 4 4 30 30 27 4 1 1 1 4 1 108
11 2 4 4 25 30 26 7 3 1 1 1 4 1 109
12 1 2 4 6 26 25 28 4 7 3 1 1 4 1 113
13 1 2 5 8 28 30 30 22 5 7 1 1 1 2 2 1 146
14 1 3 6 10 29 30 30 30 24 5 5 2 1 2 2 1 181
15 3 3 8 11 26 27 29 30 24 12 4 5 2 2 2 1 189
16 3 3 8 12 27 24 26 30 28 6 11 4 5 1 2 2 1 193
17 3 3 8 16 29 25 22 27 28 9 6 10 5 3 1 2 2 1 200
18 3 5 11 18 29 26 22 23 25 10 9 4 11 3 3 1 2 2 1 208
Total 19 25 33 29 35 31 38 35 57 66 74 79 110 110 126 121 147 141 147 111 100 65 55 23 28 13 18 7 7 6 5 3 3 1 1868
Fig. 17. Count of “in” vehicles, node by node, with a 1 min time resolution in scenario 2.
The effect on congestion is confirmed by average speeds, reported in Fig. 18 for the “charge” vehicle class. The
lower values – even lower than 40 km/h - are detected at nodes 13 and 14 in the time interval [13,18]. Values of 30
km/h refer to vehicles that change their status from “emer” to “charge” and not to traffic interactions.
AvSpeed [km/h] time [min]

Node 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
1 60 59 60
2 64 60 62
3 86 60 60 30 60
4 60 60 60 60 60
5 86 60 60 60 60 30 60
6 73 60 60 60 60 60 30 60
7 60 78 60 60 60 60 60 62
8 86 60 60 57 55 60 60 60 30 57
9 60 69 60 60 59 60 60 60 60 60
10 60 66 67 55 42 36 60 60 60 60 48
11 73 60 60 60 58 57 60 30 60 60 60 59
12 86 60 60 69 61 60 58 59 45 45 60 60 30 59
13 60 60 64 66 57 49 40 36 60 60 60 60 60 50
14 60 69 64 65 56 49 41 34 35 60 30 60 60 60 30 30 47
15 78 60 66 62 62 58 57 54 51 60 60 60 60 60 60 60 58
16 60 60 60 62 61 61 58 57 57 60 45 60 45 60 60 60 59
17 60 60 60 67 61 59 60 59 57 50 50 60 60 60 60 60 60
18 60 71 66 63 59 59 60 60 59 48 60 45 60 30 60 60 60 60
Total 60 63 61 61 61 61 61 60 59 57 55 55 59 56 54 54 54 55 57 58 59 57 51 49 55 60 60 45 50 60 60 60 60 57
Fig. 18. Average speeds [km/h] of "in" and "charge" vehicles, node by node, with a 1 min time resolution in scenario 2.
In this scenario, the congestion effect involves also “emer” vehicles, as shown in Fig. 19: at node 14, during the
minute 17, their speed decreases to 22 km/h.
AvSpeed [km/h] time [min]
Node 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Total
1 30 30 30
2 30 30 30
3 30 30 30
4 30 30 30
5 30 30 30
6 30 30 30
7 30 30 30
8 30 30 30 30 30
9 29 30 30 30 30
10 30 30 29 30 30 30 30
11 29 30 30 30 30 30 30
12 30 30 30 30 30 30 30
13 30 30 30 28 30 30 30 30 30 30
14 30 29 24 22 24 30 30 30 27
15 29 29 28 28 30 30 30 30 29
16 30 29 29 30 30 30 30 30 30
17 30 29 30 30 30 30 30 30 30
18 29 30 30 30 30 30 30 30
Total 30 30 30 30 30 30 30 30 30 30 29 30 30 30 28 28 28 29 30 30 30 30 30 30 30 30 30 29
Fig. 19. Average speeds [km/h] of "in" and "emer" vehicles, node by node, with a 1 min time resolution in scenario 2.
6.3. Comparison of energy and traffic performance in the two scenarios
To show the possible differences between the two scenarios also from the energy point of view, the average SOC
of the entire fleet is monitored node by node every minute. This analysis is not focused on a single replication but it
is averaged on 50 replications to obtain a more stable result.
Average SoC [kWh] Time [min]
Node 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
0 7.1 7.2 9.0 7.1
1 8.8 6.8 3.5 6.9
2 11.4 9.0 8.9 3.9 3.8 7.2
3 11.4 9.2 8.9 6.4 4.2 4.2 7.5
4 11.5 11.4 9.3 9.0 6.7 6.5 4.5 4.5 7.8
5 11.4 11.4 9.4 9.0 6.7 6.6 6.6 4.9 4.6 8.0
6 11.4 11.4 9.4 9.1 6.7 6.6 6.6 6.6 5.0 4.8 8.2
7 11.4 11.4 11.4 9.4 9.1 6.7 6.6 6.6 6.6 6.6 5.1 4.9 8.4
8 7.0 7.2 11.4 11.4 11.4 9.4 9.1 6.7 6.6 6.6 6.6 6.6 6.6 5.3 5.1 7.9
9 9.2 6.6 3.5 11.4 11.4 11.5 9.4 9.1 6.8 6.6 6.6 6.6 6.6 6.6 6.5 5.4 5.3 8.0
10 7.1 7.1 9.4 9.2 3.7 3.9 11.4 11.4 11.4 11.5 9.4 9.1 6.8 6.6 6.6 6.6 6.6 6.6 6.6 6.6 5.7 5.5 7.9
11 9.2 7.1 7.9 9.2 6.6 4.0 4.3 11.4 11.4 11.5 9.4 9.1 6.8 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.6 5.9 5.8 8.1
12 11.4 9.3 9.4 7.8 7.4 6.6 6.5 4.4 6.9 11.5 11.5 9.4 9.1 6.8 6.6 6.7 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.1 6.0 8.3
13 7.0 6.9 11.4 9.5 9.6 9.4 7.3 4.3 6.5 9.3 6.4 8.3 11.5 9.5 9.1 6.8 6.6 6.7 6.7 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.3 6.3 8.1
14 7.0 7.9 6.6 5.7 9.6 9.6 9.5 9.0 4.3 4.9 9.8 10.2 7.5 8.4 9.5 9.1 6.8 6.7 6.7 6.7 6.6 6.7 6.6 6.6 6.6 6.6 6.6 6.5 6.6 6.6 8.0
15 9.1 7.9 7.3 5.9 8.4 9.7 9.5 9.0 6.6 5.1 6.6 10.2 10.6 7.6 8.6 9.2 6.8 6.7 6.7 6.7 6.7 6.7 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.6 8.2
16 11.4 9.4 9.3 7.5 6.7 8.1 8.2 9.5 9.0 6.6 7.8 6.5 7.5 10.6 10.7 8.3 8.3 6.8 6.7 6.7 6.7 6.7 6.7 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.6 8.3
17 11.4 9.6 9.4 9.4 7.2 8.4 8.1 8.2 9.0 6.7 8.1 9.4 7.3 8.3 10.7 9.3 8.0 4.8 6.7 6.7 6.7 6.7 6.7 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.6 8.5
18 11.4 11.4 9.7 9.4 9.5 10.3 8.5 8.2 8.1 7.7 7.1 8.4 9.4 9.5 8.1 8.5 9.3 9.1 4.8 5.1 6.7 6.7 6.7 6.7 6.6 6.7 6.6 6.6 6.6 6.6 6.6 6.6 8.6
Total 7.0 8.1 8.0 8.5 8.4 8.6 8.4 8.5 8.3 8.3 8.2 8.2 8.2 8.5 8.3 8.5 8.2 8.2 8.0 7.4 5.9 6.0 6.4 6.4 6.5 6.4 6.5 6.5 6.6 6.6 6.6 6.6 6.6 6.6 8.1
Fig. 20. Average SOC [kWh] of "in" vehicles, node by node, with a 1 min time resolution, for scenario 1.

Node 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
0 7.1 6.9 7.0
1 8.8 6.7 3.4 6.8
2 11.4 9.0 8.8 3.9 3.7 7.1
3 11.4 9.3 8.8 6.2 4.2 4.1 7.5
4 11.4 11.4 9.4 9.0 6.6 6.5 4.6 4.4 7.7
5 11.4 11.4 9.4 9.0 6.6 6.6 6.6 4.8 4.5 8.0
6 11.4 11.4 9.4 9.0 6.6
4.9 6.6 6.6 6.6 4.7 8.2
7 11.3 11.4 11.4 9.4 9.0
6.6 6.6 6.6 6.6 6.6 5.0 4.9 8.4
8 11.4 11.4 11.4 7.9
6.6 7.8 7.0 6.6 6.6 6.6 6.6 5.2 5.2 7.8
9 11.4 11.4 11.5
6.0 9.3 7.9 4.1 6.6 6.6 6.6 6.6 6.6 5.4 5.3 8.0
10 11.4 11.4 11.4
7.5 11.5 8.3 8.3 4.1 6.2 6.6 6.6 6.6 6.6 6.6 5.7 5.5 7.9
11 11.4 11.4
8.5 11.4 11.5 9.4 7.9 4.4 4.2 6.2 6.6 6.6 6.6 6.6 6.6 5.8 5.8 8.1
12 11.4
9.5 11.4 11.4 11.5 9.4 8.1 4.9 4.5 4.4 6.2 6.6 6.6 6.6 6.6 6.5 6.1 6.0 8.3
13 11.5 11.4 11.4 11.4 8.9 8.4 8.6 7.9 4.7 4.6 4.6 6.2 6.6 6.6 6.6 6.6 6.6 6.3 6.3 8.2
14 11.4 11.4 11.4 11.4 11.5 8.7 8.5 8.4 8.9 7.3 4.9 4.7 4.6 6.2 6.6 6.6 6.6 6.6 6.6 6.6 6.6 8.3
15 11.4 11.4 11.4 11.5 11.5 9.6 8.9 8.6 8.7 7.9 4.8 5.0 4.8 4.6 6.2 6.6 6.6 6.6 6.6 6.6 6.6 8.4
16 11.4 11.4 11.4 11.5 11.5 9.7 9.6 8.9 8.9 7.9 5.5 4.8 5.1 4.9 4.7 6.3 6.6 6.6 6.6 6.6 6.6 8.6
17 11.4 11.5 11.4 11.4 11.5 11.5 9.8 9.6 9.5 9.2 8.2 5.5 5.2 4.9 5.2 4.9 4.9 6.2 6.6 6.6 6.6 6.6 8.7
18 11.4 11.4 11.4 11.5 11.5 11.5 9.9 9.7 9.5 9.7 8.5 5.9 5.1 5.2 5.0 5.2 5.0 4.9 6.3 6.6 6.7 6.6 8.9
Total 7.1 7.8 7.8 8.2 8.4 8.6 8.8 8.9 8.3 8.6 8.4 8.6 8.8 8.8 8.6 8.7 8.8 8.9 8.6 8.5 8.0 7.7 6.7 5.5 5.2 5.4 5.4 5.5 5.8 5.9 6.6 6.6 6.6 6.6 8.3
Although in scenario 2 the final SOC of the fleet at depot - node 18 - is higher on average with respect to scenario
1 - 8.9 kWh instead of 8.6 kWh -, a detailed analysis reveals that in scenario 2 more cases of vehicles with low SOC
occur from minute 22 to minute 29. Indeed in Scenario 1 only in minutes 19 and 20 vehicles arrive with an average
SOC less than 6 kWh. Different fleet SOC at the depot can be managed by the distribution service planner by
assigning a proper task and maybe related route to various vehicles according to their SOC.
Concerning the average delay, calculated as the difference between the “free flow” conditions and the estimated
travel time, scenario 1 has very few delays, mostly in the final part of the CWD lane and in the first part of the
simulated period (Fig. 22). On the contrary, scenario 2 presents higher delays and they occur in the identical part of
the CWD lane but in the middle of the simulation period (Fig. 23).
Average Delay Time [min]

Node 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
1 0.0 0.7 0.0 0.3
2 0.0 0.1 0.0 0.0 0.0 0.0
3 0.0 0.0 0.1 0.0 0.0 0.0 0.0
4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0
9 0.0 0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1
10 1.5 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
11 0.0 0.8 0.8 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.3
12 0.0 0.1 0.1 0.9 1.4 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.4
13 0.1 0.2 0.1 0.0 0.4 0.1 0.0 0.0 0.2 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1
14 4.4 1.2 0.2 0.2 0.1 0.0 0.0 0.1 0.1 0.0 0.0 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.5
15 0.1 2.3 2.7 0.2 1.5 0.4 0.0 0.0 0.0 0.1 0.1 0.0 0.0 0.2 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.8
16 0.1 0.2 0.1 0.8 0.3 1.5 2.7 0.0 0.0 0.0 0.0 0.2 0.2 0.0 0.0 0.4 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.6
17 0.1 0.3 0.1 0.1 0.4 1.3 1.4 0.8 0.0 0.0 0.0 0.0 0.1 0.2 0.0 0.0 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.5
18 0.0 0.1 0.2 0.1 0.2 0.1 0.8 1.5 0.6 0.8 0.0 0.0 0.0 0.0 0.1 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.4
Total 0.9 0.8 0.5 0.3 0.4 0.4 0.7 0.4 0.5 0.2 0.2 0.1 0.0 0.1 0.1 0.2 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.4
Fig. 22. Delays [s] of "in" vehicles, node by node, with a 1 min time resolution, for scenario 1.
Average Delay Time [min]

Node 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
1 0.0 0.5 0.0 0.3
2 0.0 0.1 0.0 0.0 0.0 0.0
3 0.0 0.1 0.0 0.0 0.0 0.0 0.0
4 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0
5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
8 0.0 0.0 0.0 0.0 4.9 9.9 7.9 0.0 0.0 0.0 0.0 0.0 0.0 3.9
9 0.0 0.0 0.0 0.0 0.0 1.6 0.6 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.6
10 0.0 0.029.5 0.0 0.1 8.9 26.3 0.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 14.2
11 0.0 2.3 0.0 0.0 0.0 0.2 4.2 0.4 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.5
12 0.6 0.0 0.0 0.0 0.1 0.1 1.6 0.8 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.5
13 0.1 0.0 0.0 0.0 4.1 23.7 41.6 31.6 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 15.8
14 0.0 0.0 0.0 0.0 0.1 6.9 27.0 48.5 54.3 44.7 1.9 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 27.8
15 0.0 0.0 0.0 0.1 0.2 0.4 1.8 3.0 5.2 6.0 1.0 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.4
16 0.0 0.0 0.0 0.1 0.2 1.0 1.0 1.0 2.4 3.9 1.6 0.2 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 1.4
17 0.0 0.0 0.0 0.0 0.1 0.2 1.1 1.1 0.1 0.8 2.0 0.8 0.1 0.2 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.7
18 0.0 0.0 0.0 0.1 0.1 0.3 1.1 1.0 0.2 0.0 0.6 0.5 0.2 0.2 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.5
Total 0.0 0.3 0.0 0.0 0.0 0.0 0.0 1.4 2.0 3.0 6.1 6.2 1.8 6.0 12.0 12.8 12.1 8.3 2.1 1.3 0.9 0.5 0.2 0.1 0.1 0.1 0.0 0.1 0.0 0.0 0.0 0.0 0.0 5.0
In a congested scenario, as scenario 2, the average charging time should increase because of the lower average
speeds. However, overtaking maneuvers are relevant in congested cases and vehicles do not recharge during the
maneuver. To verify this consideration, the average number of overtaking maneuvers is detected and reported in Fig.
24 for scenario 1 and in Fig. 25 for scenario 2.
Number of overtakes
along the Section Time [min]
Node 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
0 0.0 0.0 0.0 0.0
1 3.0 3.7 0.0 2.8
2 0.0 0.0 2.4 0.0 0.0 0.7
3 0.0 0.2 0.3 0.1 0.0 0.0 0.2
4 0.0 0.0 0.0 0.1 0.0 1.2 0.1 0.0 0.1
5 0.0 0.0 0.0 0.0 0.0 0.0 2.3 0.5 0.0 0.1
6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.7 0.4 0.0 0.1
7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.7 0.4 0.0 0.1
8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.9 0.4 0.0 0.0
9 3.3 3.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.8 0.3 0.0 1.3
10 0.0 0.0 0.0 2.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.3 0.3 0.0 0.2
11 3.1 3.8 1.2 0.0 3.1 0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.9 0.3 0.0 1.2
12 0.0 0.0 2.1 4.3 1.6 0.0 2.0 0.3 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.9 0.3 0.0 1.2
13 0.0 0.0 0.0 0.0 0.0 2.5 3.9 0.1 0.0 1.8 1.5 0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.7 0.3 0.0 0.8
14 0.0 1.6 3.9 0.2 0.0 0.0 0.0 1.7 0.8 0.3 0.0 1.9 1.9 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.6 0.3 0.0 0.8
15 3.1 4.8 3.4 1.4 1.0 0.0 0.0 0.0 2.1 1.3 0.4 0.0 2.3 1.9 0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.6
16 0.0 0.0 2.8 4.0 1.8 4.4 1.7 0.0 0.0 0.0 1.4 1.7 0.6 0.0 1.5 2.5 0.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.8
17 0.0 0.0 0.0 2.0 2.6 3.7 4.3 1.4 0.0 0.0 0.0 1.4 2.0 0.7 0.0 1.5 2.4 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.6
18 0.0 0.0 0.0 0.0 0.0 2.1 4.1 3.3 3.9 1.6 0.0 0.0 0.0 1.4 2.0 0.7 0.0 1.1 0.4 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.5
Total 0.0 1.5 2.5 1.4 1.1 0.9 1.4 1.3 1.6 1.1 1.2 0.6 0.6 0.5 0.6 0.6 1.2 0.8 0.9 0.6 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0 1.1
Fig. 24. Number of overtaking maneuvers detected in the CWD lane, node by node, with a 1 min time resolution, for scenario 1.
Number of overtakes
along the Section Time [min]
Node 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
0 0.0 0.0 0.0
1 2.8 4.0 0.0 2.9
2 0.0 0.0 2.4 0.0 0.0 0.7
3 0.0 0.3 0.5 0.0 0.0 0.0 0.2
4 0.0 0.0 0.0 0.2 0.1 1.2 0.2 0.0 0.1
5 0.0 0.0 0.0 0.0 0.0 0.0 2.7 0.5 0.0 0.1
6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 0.4 0.0 0.1
7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 0.3 0.0 0.1
8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.6 0.4 0.0 0.0
9 0.0 0.0 0.0 2.6 5.0 0.7 0.0 0.0 0.0 0.0 0.0 1.8 0.4 0.0 2.6
10 0.0 0.0 0.0 0.0 0.0 0.3 2.0 0.3 0.0 0.0 0.0 0.0 0.0 1.6 0.4 0.0 0.5
11 0.0 0.0 0.0 0.0 2.5 4.0 3.5 1.2 0.3 0.0 0.0 0.0 0.0 0.0 1.2 0.3 0.0 2.4
12 0.0 0.0 0.0 0.0 0.0 2.3 3.7 1.5 1.2 0.3 0.0 0.0 0.0 0.0 0.0 1.1 0.4 0.0 1.4
13 0.0 0.0 0.0 0.0 0.0 0.0 0.2 1.9 1.4 1.1 0.4 0.0 0.0 0.0 0.0 0.0 0.9 0.3 0.0 0.4
14 0.0 0.0 0.0 0.0 0.0 0.2 1.7 2.6 3.9 2.0 1.2 1.1 0.4 0.0 0.0 0.0 0.0 0.0 0.7 0.2 0.0 1.7
15 0.0 0.0 0.0 0.0 0.0 1.4 3.7 4.4 4.1 3.7 1.5 1.4 1.1 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.7
16 0.0 0.0 0.0 0.0 0.0 0.0 1.2 3.7 4.2 3.7 2.3 1.4 1.1 0.9 0.2 0.0 0.0 0.0 0.0 0.0 0.0 2.0
17 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.1 3.3 3.5 2.3 2.0 1.2 1.1 0.9 0.3 0.0 0.0 0.0 0.0 0.0 1.3
18 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.9 2.8 2.4 1.8 1.8 1.3 1.0 0.9 0.2 0.0 0.0 0.0 0.0 0.7
Total 0.0 1.2 2.2 0.9 0.1 0.0 0.0 0.0 0.0 0.9 1.7 0.9 1.5 1.2 0.8 0.8 1.6 2.0 1.9 2.0 2.1 1.9 2.0 1.4 1.1 0.9 0.8 0.5 0.5 0.1 0.0 0.0 0.0 0.0 1.3
Fig. 25. Number of overtaking maneuvers detected in the CWD lane, node by node, with a 1 min time resolution, for scenario 2.
In scenario 2 a higher number of overtakes concentrated after the node 14 and during the time interval [15, 25]
confirms the previous considerations. In scenario 1 the overtaking maneuvers are relevant in those sections, but in
another time interval [1, 10] and therefore they do not affect the mainstream flow.
6.4. Scenario 3: Simultaneous arrival with a CWD lane service interruption
The conditions of Scenario 2, which is the critical one, can also be monitored in the case of an incident involving
the CWD lane. For various reasons, such as maintenance, cleaning or inspection of the coils, the CWD lane could be
out of service in a particular node and it should be kept free of traffic for a certain period. This vehicle-free time
window can be obtained by reducing vehicle speeds in the CWD lane along the section approaching the node
involved in the incident - modeled by “Headway correction” and “Time estimation next” routines -. The strategy
will be applied by avoiding overtaking maneuvers in the approaching section.
In the experiments performed, the CWD lane closure lasts two minutes, starting from the 18th minute- 1080s
from the beginning of the simulation -. The data reported is averaged on 50 replications to show more reliable
results.
Node 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
0 7.1 7.2 9.0 7.1
1 8.8 6.8 3.5 6.9
2 11.4 9.0 8.9 3.9 3.8 7.2
3 11.4 9.2 8.9 6.4 4.2 4.2 7.5
4 11.5 11.4 9.3 9.0 6.7 6.5 4.5 4.5 7.8
5 11.4 11.4 9.4 9.0 6.7 6.6 6.6 4.9 4.6 8.0
6 11.4 11.4 9.4 9.1 5.0 6.7 6.6 6.6 6.6 4.8 8.2
7 11.4 11.4 11.4 9.4 6.6 9.1 6.7 6.6 6.6 6.6 5.1 4.9 8.4
8 11.4 11.4 11.4 6.6 7.9 8.0 7.9 6.6 6.6 6.6 6.6 5.3 5.1 7.9
9 11.4 11.4 6.611.5 9.3 7.9 4.6 6.6 6.6 6.6 6.6 6.5 5.4 5.3 8.1
10 11.4 11.4 7.511.4 11.5 8.4 8.3 4.1 6.6 6.6 6.6 6.6 6.6 6.6 5.7
5.5 8.0
11 11.4 8.511.4 11.4 11.5 9.4 7.9 4.3 4.2 6.6 6.6 6.6 6.6 6.6
6.6 5.9 5.8 8.1
12 9.611.4 11.4 11.4 11.5 9.4 8.2 4.8 4.2 4.6 6.6 6.6 6.6
6.6 6.6 6.6 6.1 6.0 8.3
13 11.5 11.4 11.4 11.5 8.9 8.4 8.6 8.2 4.8 4.6 4.6 6.6
6.6 6.6 6.6 6.6 6.6 6.3 6.3 8.2
14 11.5 11.4 11.4 11.4 11.5 8.7 8.4 8.2 8.9 7.7 5.2 4.7
4.6 6.6 6.6 6.6 6.6 6.6 6.5 6.6 6.6 8.3
15 11.4 11.4 11.4 11.5 11.5 9.6 8.8 8.6 8.5 8.2 5.2
5.0 4.8 4.7 6.6 6.6 6.6 6.6 6.6 6.6 6.6 8.4
16 11.4 11.4 11.4 11.5 11.5 9.7 8.8 9.2
9.2 8.0 5.3 4.7 6.6 6.6 6.6 6.6 6.6 6.6 6.6 8.8
17 11.4 11.4 11.5 11.4 11.5 11.5 9.8 9.5 3.4 10.2 9.4 8.5 5.8 4.6 4.4 6.6 6.6 6.6 6.6 6.6 6.6 9.0
18 11.5 11.4 11.5 11.4 11.5 11.5 9.9 9.5 10.4 3.7 10.2 10.2 8.8 6.2 4.9 4.6 4.7 6.6 6.6 6.6 6.6 6.6 9.1
Total 7.1 7.9 7.9 8.2 8.4 8.7 8.8 8.9 8.3 8.7 8.5 8.7 8.9 8.8 8.6 8.8 8.8 8.8 8.5 7.8 7.5 8.5 8.7 8.1 7.3 5.7 5.2 5.2 5.8 6.6 6.6 6.6 6.6 6.6 8.4
The impact of the incident on the SOC is shown in Fig. 26 in which the missing data for minutes 18 ad 19
confirm the CWD lane closure for that period at node 16. Data also show that vehicles arriving at the depot after the
temporary closure have a higher SOC than those in scenario 2 (see Fig. 21). The lower SOC values – 3.4 and 3.7
kWh – refers to “emer” vehicles that have avoided the lane closure because of the speed regulation. Indeed, their
charging time has been reduced because their speed has been set to 60 km/h. The propagation of the strategy effects
can be analyzed by average speeds. The average speed reductions, as a result of the simulation, are confirmed in Fig.
27 for the section approaching the node 16 during the time interval [20,23], in which the average speed drops off
even to 20 km/h at minute 20.
AvSpeed [km/h] Time [min]
Node 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
1 60 48 30 49
2 86 62 61 30 30 51
3 74 61 59 32 30 30 50
4 86 69 61 60 32 44 31 30 51
5 79 66 60 60 60 60 60 32 30 53
6 74 67 60 60 60 60 60 60 33 30 54
7 86 69 63 60 60 60 60 60 60 60 33 30 55
8 79 66 61 52 51 52 60 60 60 60 60 33 30 51
9 74 64 60 60 52 36 58 60 60 60 60 60 32 30 53
10 86 72 63 63 52 46 38 33 58 60 60 60 60 60 32 30 46
11 80 69 62 64 60 54 50 34 33 59 60 60 60 60 60 34 30 52
12 75 66 61 66 61 60 53 38 34 36 59 60 60 60 60 60 34 30 54
13 75 63 63 65 54 45 40 35 37 36 36 60 60 60 60 60 60 34 30 46
14 86 69 62 64 63 53 45 37 30 30 38 37 36 60 60 60 60 60 60 34 30 42
15 78 66 61 65 63 61 55 53 50 48 40 40 36 37 60 60 60 60 60 60 60 53
16 72 63 62 64 63 61 59 20 21 23 39 60 60 60 60 60 60 60 60 41
17 86 68 62 65 64 63 60 59 30 60 54 51 45 34 36 60 60 60 60 60 60 55
18 83 67 61 65 62 62 60 60 59 30 60 60 55 49 38 32 36 60 60 60 60 60 56
Total 61 55 56 55 56 56 56 53 55 53 52 53 55 51 50 49 50 49 49 29 39 44 50 51 44 40 38 43 52 60 60 60 60 50
Fig. 27. Average speeds [km/h] of "in" and "emer" vehicles, node by node, with a 1 min time resolution, for scenario 3.
The speed reduction has a sensible effect on the delay of the vehicles, estimated as the difference between the
simulated travel time and the predicted travel time on the base of the planned speed. As expected, the higher delays
occur in the section before the node 16, during the recovery period. Also negative values of delay are detected: they
are caused by the speed change for “emer” vehicles that are forced to proceed at the speed of the “charge” vehicles
during the “safety” control action. This case is represented in Fig. 29 in which the trajectories of all the charging
vehicles are traced for a sample replication (#2).The trajectory of the vehicle #237, represented by black dots, shows
that, if it travelled at 30 km/h, it would have been involved in the lane closure at node 16. However, because of the
speed control action, it increases its speed and it crosses the node 16 before the incident event. After the node 16, its
speed is set again at 30 km/h, as required by its charging needs, and it is then maintained until the final node.
Av Delay Time [min]
Node 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total
1 0.0 0.7 0.0 0.3
2 0.0 0.1 0.0 0.0 0.0 0.0
3 0.0 0.0 0.1 0.0 0.0 0.0 0.0
4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
6 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0
7 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0
8 0.0 0.0 0.0 1.3 3.4
0.0 8.9 0.0 0.0 0.0 0.0 0.1 0.0 1.8
9 0.0 0.0 0.0 0.0
0.0 2.1 1.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.9
10 0.0 0.0 0.0 0.0
23.6 3.6 13.2 0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 9.8
11 0.0 0.1 0.0
2.0 0.0 0.3 4.1 0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 1.4
12 0.0 0.0
0.7 0.0 0.1 0.2 1.2 0.2 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.5
13 0.0
0.2 0.0 0.1 3.2 14.0 26.8 37.7 1.4 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 13.8
14 0.0
0.1 0.0 0.0 0.1 3.8 15.0 33.6 55.6 50.1 8.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 24.0
15 0.0 0.0 0.0 0.1 0.2 0.3 1.8 2.7 5.5 7.4 1.7 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.6
16 0.0 0.1 0.0 0.2 0.2 0.8 -6.0 106.2 102.6 88.4 9.3 -40.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 44.2
17 0.1 0.0 0.0 0.1 0.1 0.2 1.0 0.6 0.0 0.0 1.6 4.2 1.7 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.2
18 0.0 0.0 0.0 0.0 0.1 0.2 0.7 0.3 1.0 0.0 0.1 0.0 1.1 1.5 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.4
Total 0.0 0.4 0.0 0.0 0.0 0.0 0.0 0.7 1.3 2.2 4.9 6.0 1.8 4.7 9.7 12.9 10.9 11.4 3.5 60.3 41.8 28.2 3.5 -2.1 0.8 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 9.4
2500
Time [s]
2000
1500
1000
500
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Node
Fig. 29. Space-time relationship for "in" vehicles in the case of application of the strategy.
7. Conclusions and further researches
This study presents a method for assessing the performance of the wireless power systems to dynamically charge
electric vehicles while driving, from both traffic and energy points of view. The model assumptions are aimed at
simulating a freight scenario for urban distribution in the city center with medium sized vans. The CWD lane has
been modeled in a realistic layout, as installed on a ring road with secondary on ramps, for supporting electric
charging operations during the return part of the vehicle trips. The set of speeds for CWD operations is relatively
low because, for the assumed P cz and the test vehicle considered, it is the most adequate to satisfy the fleet operator
need to guarantee a minimum SOC at the vehicle arrivals to the depot.
The implemented dynamic traffic simulator adopts a mesoscopic approach by updating traffic and energy data
only at defined nodes along the road, generally spaced in the order of hundreds of meters. The traffic simulator
operates according to a cooperative driving behavior among vehicles, both for the overtaking maneuvers and the
entries management and it is able to simulate different traffic conditions. Primary traffic parameters can be
estimated in the CWD lane, such as the vehicle counting, average speeds and delays, which are time dependent and
relevantly change along the road. The model also allows the implementation of a speed control strategy to manage
temporary incidents, for example due to extraordinary maintenance operations. This strategy could also be applied
in the case of high traffic volumes to facilitate the entries of the vehicles from the on ramps. The traffic model is
able to manage even platooning conditions and delays caused by the strategy when minimum headways in the CWD
lane are required and it is able to assess the effects of the strategy in terms of vehicle SOCs and speeds over time.
With respect to the traditional dynamic traffic models, in the proposal here presented, the current vehicle energy
needs affect drivers’ behavior. According to their SOC along the road, vehicles are simulated inside or outside the
charging lane and their speeds are set according to their charging needs.
The implemented dynamic traffic simulator has an approximation compatible with the stage of development of
the CWD technology and the deployment of the cooperative driving systems. For this reason, in our idea, the
presented model should be considered as a support methodology of a feasibility study for a promising future
technology, in ex-ante evaluations. When the CWD technology and the full cooperative system are available in large
scale applications, more investigations will be needed and then the model could be empirically calibrated, also
according to the observed driving behavior.
More developments can be carried out to improve the realism of the traffic simulator. For example, the speed in
the other lanes is estimated according to the initial traffic density. In a further enhancement, for very high FEV
traffic level, the model could also consider that, during the time period, the density in the other lanes may change for
both FEVs lane changing maneuvers and fluctuations of traditional “not electric” traffic. Even more complex
scenarios with different vehicle types and related traffic flows could be analyzed. Furthermore, the model could be
extended to flows of private cars, where the recharging needs should be related to the vehicle destinations, thus
requiring the estimated demand data of detailed Origin/Destination matrixes.
Acknowledgements
This study is partially supported by eCo-FEV project (Grant agreement no: 314411 - http://www.eco-
fev.eu/home.html) and the authors would like to thank all the project partners for their useful support and
suggestions on this subject.
References
Barceló J., Casas J. (2005), Dynamic network simulation with AIMSUN, in Simulation Approaches in Transportation Analysis, Vol.31, pp. 57-98,
Springer, http://dx.doi.org/10.1007/0-387-24109-4_3
Ben-Akiva M. , Gao S., Wei Z., Wen Y. (2012), A dynamic traffic assignment model for highly congested urban networks, in Transportation
Research, Part C: Emerging Technologies, Vol.24, pp. 62-82, http://dx.doi.org/10.1016/j.trc.2012.02.006
Boulanger A. G., Chu A. C., Maxx S., Waltz D.L. (2011), Vehicle electrification: status and issues,in Proceeding of the IEEE, Vol.99, Issue 6, pp.
1116-1138, http://dx.doi.org/ 10.1109/JPROC.2011.2112750
Cascetta E. (2001), Transportation systems engineering: theory and methods, Kluwer
Daganzo C.F. (1997), Fundamentals of Transportation and Traffic Operations, Pergamon
Deflorio F., Guglielmi P., Pinna I., Castello L., Marfull S. (2013), Modelling and Analysis of wireless “Charge While Driving” operations for
Fully Electric Vehicles, SIDT 2013 Scientific Seminar, Trieste (Italy), 18th October 2013. In press in Transportation Research Procedia
(10.1016/j.trpro.2015.01.008)
Drake J. S., Schofer J. L., May A. D. (1967), A Statistical Analysis of Speed Density Hypotheses, Third International Symposium on the Theory
of Traffic Flow Proceedings, Elsevier North Holland, Inc., New York
eCo-FEV (2013), Use cases and requirements for an efficient cooperative platform, Deliverable D200.1 of the eCo-FEV project, http://www.eco-
fev.eu/
ECSS (2009), Space engineering – Verification, ESA Requirements and Standards Division
ETSI TR 102 638 V1.1.1 (2009-06), Technical Report: Intelligent Transport Systems (ITS); Vehicular Communications; Basic Set of
Applications; Definitions
Ezell S. (2010), Explaining International IT Application leadership: Intelligent Transportation Systems, The Information Technology &
Innovation
Farah H., Toledo T. (2010), Passing behavior on two-lane highways, in Transportation Research Part F: Traffic Psychology and Behaviour,
Vol.13, Issue 6, pp. 355-364, http://dx.doi.org/10.1016/j.trf.2010.07.003
Farah H. (2013), Modeling drivers' passing duration and distance in a virtual environment, in IATSS Research, Vol.37, Issue 1, pp. 61-67,
http://dx.doi.org/10.1016/j.iatssr.2013.03.001
Gutiérrez, J., García, J., Harbour, M. (1998), Best-case analysis for improving the worst-case schedulability test for distributed hard real-time
systems, in Proceedings of the 10th Euromicro Workshop on Real-Time Systems, pp. 35-44, http://dx.doi.org/10.1109/EMWRTS.1998.684945
Hoogendoorn S. P., Bovy P. H. L. (2001), State-of-the-art of vehicular traffic flow modelling, in Special Issue on Road Traffic Modelling and
Control of the Journal of Systems and Control Engineering
INCOSE (2011), Systems Engineering Handbook v. 3.2.2., INCOSEϋTPϋ2003ϋ002ϋ03.2.2
Johnson J., Chowdhury M., He Y., Taiber J. (2013), Utilizing real time information transferring potentials to vehicles to improve the fast
charging process in electric vehicles, in Transportation research, Part C: Emerging Technologies, Vol.26, pp.352-366,
http://dx.doi.org/10.1016/j.trc.2012.10.009
Kirsch D.A. (2000), The electric vehicle and the burden of history, Rutgers University press
Myers G. J., Badgett T., Sandler C. (2004), The Art of Software Testing, John Wiley and Sons Inc.
Palakon K., Wang Y., Imura T., Fujimoto H., Hon, Yoichi (2011), Electric Vehicle Automatic Stop using Wireless Power Transfer Antennas, in
IECON 2011 - 37th Annual Conference on IEEE Industrial Electronics Society, pp. 3840-3845, http://dx.doi.org/10.1109/IECON.2011.6119936
Sun J., Chen S. (2013), Dynamic speed guidance for active highway signal coordination: roadside against in-car strategies, in Intelligent
Transport Systems, IET, Vol.7, Issue 4, pp. 432-444, http://dx.doi.org/ 10.1049/iet-its.2012.0084

Traffic Modeling of A Cooperative Charge While Driving System in A Freight Transport Scenario

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Traffic Modeling of A Cooperative Charge While Driving System in A Freight Transport Scenario

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Traffic Modeling of A Cooperative Charge While Driving System in A Freight Transport Scenario

Uploaded by

Copyright:

Available Formats

Available online at www.sciencedirect.

Traffic Modeling of a Cooperative Charge While Driving System in

* Corresponding author. Tel.: +39-011-0905601; fax: +39-011-0905699.

Fig. 2. CWD system layout.

3. The choice of the traffic model

4. The structure of the CWD simulation model

4.1. The model structure

4.1.1. Initial traffic state

Fig. 5. Algorithm structure.

4.1.2. Strategy detection

4.1.3. Headway correction

4.1.4. Overtaking at node

4.1.5. SOC estimation

ܴ‫ ݐ݋ݐ‬ൌ ܴ݀‫ ݃ܽݎ‬൅ ܴ‫ ݈݈݃݊݅݋ݎ‬൅ ܴ݈ܽܿܿ݁݁‫ ݊݋݅ݐܽݎ‬൅ ܴ‫݁݌݋݈ݏ‬ (2)

ܴ‫ ݈݈݃݊݅݋ݎ‬ൌ ݉ ή ሺ݂Ͳ ൅ ݂ʹ ή ‫ ʹ ݒ‬ሻ (4)

ܴ݈ܽܿܿ݁݁‫ ݊݋݅ݐܽݎ‬ൌ ݉ ή ܽ (5)

ܴ‫ ݁݌݋݈ݏ‬ൌ ݉ ή ݃ ή ‫݌‬ (6)

ܴ‫ ݐ݋ݐ‬ή ‫ݒ‬ ‫݊݋݅ݐܿ݁ݏܮ‬

‫ ݂݂ܼ݁ܥܮ‬ൌ ‫ ܼܥܮ‬െ ܶ‫ ݇ݎ‬ή ‫ܦܥܮ‬ (9)

4.1.6. SOC update for overtaking

Fig. 6. Schematization of overtaking maneuvers.

4.1.7. Speed test

4.1.8. Status estimation next

4.1.9. Time estimation next

Fig. 7. End queue estimation during the strategy.

5. Verification and validation processes

6. The CWD scenario and experimental results

Fig. 8. Turin ring road with main secondary accesses.

Table 1. Inputs to the algorithm.

Type Input Value Unit

Length of the CZs 20 m

Table 2. Input FEVs at each access

Access node Number of simulated FEVs

6.1. Scenario1 - Simultaneous end of service in the city center

Fig. 11. Space-time relationship for "in" vehicles in scenario 1.

AvSpeed [km/h] Time [min]

6.2. Scenario2 - Simultaneous arrival at depot

Fig. 16. Space-time relationship for "in" vehicles in scenario 2.

Count Nv time [min]

AvSpeed [km/h] time [min]

6.3. Comparison of energy and traffic performance in the two scenarios

Average SoC [kWh] Time [min]

Average Delay Time [min]

Average Delay Time [min]

6.4. Scenario 3: Simultaneous arrival with a CWD lane service interruption

7. Conclusions and further researches

Cascetta E. (2001), Transportation systems engineering: theory and methods, Kluwer

Daganzo C.F. (1997), Fundamentals of Transportation and Traffic Operations, Pergamon

INCOSE (2011), Systems Engineering Handbook v. 3.2.2., INCOSEϋTPϋ2003ϋ002ϋ03.2.2

You might also like