U.S. patent application number 13/207566 was filed with the patent office on 2013-02-14 for methods and apparatus for estimating power usage.
This patent application is currently assigned to FORD GLOBAL TECHNOLOGIES, LLC. The applicant listed for this patent is Oleg Yurievitch Gusikhin, Mark John Jennings, Perry Robinson MacNeille, Brian Petersen, Edward Andrew Pleet. Invention is credited to Oleg Yurievitch Gusikhin, Mark John Jennings, Perry Robinson MacNeille, Brian Petersen, Edward Andrew Pleet.
Application Number | 20130041552 13/207566 |
Document ID | / |
Family ID | 47595729 |
Filed Date | 2013-02-14 |
United States Patent
Application |
20130041552 |
Kind Code |
A1 |
MacNeille; Perry Robinson ;
et al. |
February 14, 2013 |
Methods and Apparatus for Estimating Power Usage
Abstract
A computer-implemented method includes establishing a road
network model on which a plurality of simulated vehicles may be
run. The method also includes setting up a plurality of scenarios
under which vehicle driving conditions vary to be run on the road
network model. The illustrative method includes receiving energy
usage related data for a plurality of simulated vehicles run in at
least one of the plurality of scenarios on the road network model.
The method further includes calculating a total energy consumption
for each of the vehicles. The method additionally includes
repeating the receiving and calculating steps to determine how
various elements of the road network model and scenarios effect
vehicle energy consumption.
Inventors: |
MacNeille; Perry Robinson;
(Lathrup Village, MI) ; Pleet; Edward Andrew;
(Livonia, MI) ; Jennings; Mark John; (Saline,
MI) ; Gusikhin; Oleg Yurievitch; (West Bloomfield,
MI) ; Petersen; Brian; (Ferndale, MI) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
MacNeille; Perry Robinson
Pleet; Edward Andrew
Jennings; Mark John
Gusikhin; Oleg Yurievitch
Petersen; Brian |
Lathrup Village
Livonia
Saline
West Bloomfield
Ferndale |
MI
MI
MI
MI
MI |
US
US
US
US
US |
|
|
Assignee: |
FORD GLOBAL TECHNOLOGIES,
LLC
Dearborn
MI
|
Family ID: |
47595729 |
Appl. No.: |
13/207566 |
Filed: |
August 11, 2011 |
Current U.S.
Class: |
701/32.9 ;
701/439; 703/2 |
Current CPC
Class: |
G01C 21/3469
20130101 |
Class at
Publication: |
701/32.9 ; 703/2;
701/439 |
International
Class: |
G01C 21/34 20060101
G01C021/34; G06F 7/60 20060101 G06F007/60; G06F 17/10 20060101
G06F017/10 |
Claims
1. A computer-implemented method comprising: establishing a road
network model on which a plurality of simulated vehicles may be
run; setting up a plurality of scenarios under which vehicle
driving conditions vary to be run on the road network model;
receiving energy usage related data for a plurality of simulated
vehicles run in at least one of the plurality of scenarios on the
road network model; and calculating a total energy consumption for
each of the vehicles.
2. The method of claim 1, wherein the road network model includes a
geometry of a road.
3. The method of claim 1, wherein the road network model includes a
length of a road.
4. The method of claim 1, wherein the road network model includes a
number of lanes of a road.
5. The method of claim 1, wherein the road network model includes a
vehicle flow for a road.
6. The method of claim 5, wherein the road network model includes a
vehicle composition of the flow.
7. The method of claim 1, wherein the road network model includes
desired speeds for a road.
8. The method of claim 1, wherein the road network model includes
traffic signal data for a road.
9. The method of claim 1, wherein the road network model includes
at least weather data affecting travel on the road.
10. The method of claim 9, wherein the weather data includes at
least visibility data.
11. The method of claim 9, wherein the weather data includes at
least temperature data.
12. The method of claim 1, wherein each of the plurality of
scenarios provides different variables to a road network model to
adjust elements of the road network model such that the impact of
changes in a particular element on energy consumption can be
observed.
13. The method of claim 1, further including repeating the
receiving and calculating steps to determine how various elements
of the road network model and scenarios effect vehicle energy
consumption.
14. A computer implemented method comprising: breaking a route into
a plurality of segments having the same or a similar energy use
affecting characteristic; based on one or more energy use affecting
characteristics of each segment, assigning a total energy usage
cost to the segment; adding the total energy usage cost of all
segments comprising a route to determine a total route energy cost;
repeating the adding step for a plurality of routes comprised of
varying segments between a current location and a destination; and
presenting a driver with the route having the lowest total route
energy cost.
15. The method of claim 14, wherein the energy use affecting
characteristics include a road type.
16. The method of claim 14, wherein the energy use affecting
characteristics include a speed limit.
17. The method of claim 14, wherein the energy use affecting
characteristics include a gradient.
18. A machine readable storage medium storing instructions that,
when executed, cause a processor to perform the method comprising:
establishing a road network model on which a plurality of simulated
vehicles may be run; setting up a plurality of scenarios under
which vehicle driving conditions vary to be run on the road network
model; receiving energy usage related data for a plurality of
simulated vehicles run in at least one of the plurality of
scenarios on the road network model; calculating a total energy
consumption for each of the vehicles; and repeating the receiving
and calculating steps to determine how various elements of the road
network model and scenarios effect vehicle energy consumption.
19. The machine readable storage medium of claim 18, wherein each
of the plurality of scenarios provides different variables to a
road network model to adjust elements of the road network model
such that the impact of changes in a particular element on energy
consumption can be observed.
Description
TECHNICAL FIELD
[0001] The illustrative embodiments generally relate to methods and
apparatus for estimating power usage.
BACKGROUND
[0002] Evaluating real world fuel consumption of a vehicle is
useful in developing algorithms for low cost routing and
distance-to-empty. This calculation enables vehicle features that
can result in significant fuel and travel time savings. With
advances in digital systems, there is an explosion of inputs
available to electronic vehicle features that can influence
emissions, energy consumption and travel time.
[0003] Traffic simulation tools help in replicating real life
traffic and driver behavior. Different scenarios can be analyzed to
understand vehicle behavior under varying conditions.
[0004] Marketing studies have revealed that range anxiety is the
number one concern for battery electric vehicle (BEV) owners. The
illustrative embodiments enable vehicle features that can eliminate
range anxiety by presenting real-world estimates of distance to
empty and also ensure the best fuel economy by presenting low
energy routes.
SUMMARY
[0005] In a first illustrative embodiment, a computer-implemented
method includes establishing a road network model on which a
plurality of simulated vehicles may be run. The illustrative method
also includes setting up a plurality of scenarios under which
vehicle driving conditions vary to be run on the road network
model.
[0006] Also, the illustrative method includes receiving energy
usage related data for a plurality of simulated vehicles run in at
least one of the plurality of scenarios on the road network model.
The illustrative method further includes calculating a total energy
consumption for each of the vehicles.
[0007] The illustrative method additionally includes repeating the
receiving and calculating steps to determine how various elements
of the road network model and scenarios effect vehicle energy
consumption.
[0008] In a second illustrative embodiment, a computer implemented
method includes breaking a route into a plurality of segments
having the same or a similar energy use affecting characteristic.
This illustrative method also includes assigning a total energy
usage cost to the segment based on one or more energy use affecting
characteristics of each segment.
[0009] Further, this illustrative method includes adding the total
energy usage cost of all segments comprising a route to determine a
total route energy cost and repeating the adding step for a
plurality of routes comprised of varying segments between a current
location and a destination. This illustrative method additionally
includes presenting a driver with the route having the lowest total
route energy cost.
[0010] In a third illustrative embodiment, a machine readable
storage medium stores instructions that, when executed, cause a
processor to perform the method including establishing a road
network model on which a plurality of simulated vehicles may be
run. The exemplary method also includes setting up a plurality of
scenarios under which vehicle driving conditions vary to be run on
the road network model.
[0011] Further, the illustrative method includes receiving energy
usage related data for a plurality of simulated vehicles run in at
least one of the plurality of scenarios on the road network model
and calculating a total energy consumption for each of the
vehicles. Also, the illustrative method includes repeating the
receiving and calculating steps to determine how various elements
of the road network model and scenarios effect vehicle energy
consumption.
BRIEF DESCRIPTION OF THE DRAWINGS
[0012] FIG. 1 shows an illustrative example of a fuel efficiency
testing process;
[0013] FIG. 2 shows an illustrative example of a road network
modeling process;
[0014] FIG. 3 shows an illustrative example of a scenario setup
process; and
[0015] FIG. 4 shows an illustrative example of an energy
consumption determination process.
DETAILED DESCRIPTION
[0016] As required, detailed embodiments of the present invention
are disclosed herein; however, it is to be understood that the
disclosed embodiments are merely exemplary of the invention that
may be embodied in various and alternative forms. The figures are
not necessarily to scale; some features may be exaggerated or
minimized to show details of particular components. Therefore,
specific structural and functional details disclosed herein are not
to be interpreted as limiting, but merely as a representative basis
for teaching one skilled in the art to variously employ the present
invention.
[0017] Evaluating the real world fuel consumption of a vehicle is
one method of developing algorithms for low-cost routing and
distance-to-empty. This calculation enables vehicle features that
can result in significant fuel and travel time savings. With
advances in digital systems, there is an explosion of inputs
available to electronic vehicle features that can influence
emissions, energy consumption and travel time. Making use of this
data will eventually result in cost savings for a consumer.
[0018] Traffic simulation tools help in replicating real life
traffic and driver behavior. Different scenarios can be analyzed to
understand the most important influencers on real world fuel
economy. The energy consumption under different traffic and road
conditions can also be evaluated using traffic simulation.
[0019] The illustrative embodiment reflect development of software
modules for embedded and cloud-based applications that receives
inputs such as driver characteristics, road topology, vehicle
characteristics, weather, traffic, etc. and output energy
consumption for route optimization and distance to empty
computations. This can also enable applications on embedded
platforms such as SYNC, mobile platforms such as smart phones and
in web-based applications in the cloud.
[0020] The illustrative embodiments enable vehicle features that
can eliminate range anxiety by presenting real-world estimates of
distance to empty and also ensure the best fuel economy by
presenting low energy routes.
[0021] The illustrative embodiments include a laboratory method of
computing real-world fuel consumption from external data available
in digital formats. The external data available in digital formats
is used as an input into a traffic simulator called VISSIM2, which
can generate realistic drive cycles. Drive cycles are then input
into a powertrain simulation to compute energy along a specific
route for a specific vehicle. The energies for the entire set of
vehicles are statistically analyzed for average energy consumption
and the expectation interval energy consumption along the
route.
[0022] A method of computing energy from drive cycles called
modeFrontier is introduced and makes the energy consumption
analysis much simpler and more suitable for embedded processors and
cloud-based applications. Four dimensional energy tables are
populated with energy values using CVSP, with the dimensions being
vehicle weight, speed, road gradient and accessory loads. Actual
road gradient and vehicle acceleration from the VISSIM drive cycle
are combined into a singled variable for use in the tables.
Throughout any simulated drive only the speed and vehicle
acceleration vary, so the table was divided into sub-tables for
each accessory load and vehicle weight. A cubic spline surface was
created in the speed and road gradient dimensions for each
sub-table to more accurately estimate the fuel consumption.
[0023] sCVSP was also used to compute the maximum acceleration the
vehicle is capable of at a specific weight, road gradient and
speed. This was passed into VISSIM manually as an
acceleration-velocity curve. If this curve was incorrect the
accelerations would be either too large for the vehicle to achieve
or never large enough to represent maximum vehicle
acceleration.
[0024] VISSIM is a simulation package that can analyze private and
public transport operations under constraints such as lane
configuration, traffic composition, traffic signals, public
transportation stops, etc., thus making it a useful tool for the
evaluation of various alternatives based on transportation
engineering and planning measures of effectiveness.
[0025] VISSIM can be applied as a useful tool in a variety of
transportation problem settings. The following list provides a
selective overview of previous applications of VISSIM:
[0026] Development, evaluation and fine-tuning of signal priority
logic
[0027] Evaluation and optimization of traffic operations in a
combined network of coordinated and actuated traffic signals.
[0028] Feasibility and traffic impact studies of integrating light
rail into urban street networks.
[0029] Analysis of slow speed weaving and merging areas.
[0030] Easy comparison of design alternatives including signalized
and stop sign controlled intersections, roundabouts and grade
separated interchanges.
[0031] Capacity and operations analysis of complex station layouts
for light rail and bus systems.
[0032] With its built-in Dynamic Assignment model, VISSIM can
answer route choice dependent questions such as the impacts of
variable message signs or the potential for traffic diversion into
neighborhoods for networks up to the size of medium sized
cities.
[0033] Modeling and simulating flows of pedestrians--in streets and
buildings--allow for a wide range of new applications. VISSIM can
also simulate and visualize the interactions between road traffic
and pedestrians.
[0034] The traffic simulator is a microscopic traffic flow
simulation model including the car following and lane change logic.
The signal state generator is a signal control software pooling
detector information from the traffic simulator on a discrete time
step basis (down to 1/10 of a second). It then determines the
signal status for the following time step and returns this
information to the traffic simulator.
[0035] The accuracy of a traffic simulation model is mainly
dependent on the quality of the vehicle modeling, e.g. the
methodology of moving vehicles through the network. In contrast to
less complex models using constant speeds and deterministic car
following logic, VISSIM uses a psycho-physical driver behavior
model. The basic concept of this model is that the driver of a
faster moving vehicle starts to decelerate as he reaches his
individual perception threshold when approaching a slower moving
vehicle. Since he cannot precisely determine the speed of the other
vehicle, his speed will fall below that vehicle's speed until he
starts to slightly accelerate again after reaching another
perception threshold. This results in an iterative process of
acceleration and deceleration.
[0036] Stochastic distributions of speed and spacing thresholds
replicate individual driver behavior characteristics. VISSIM's
traffic simulator not only allows drivers on multiple lane roadways
to react to preceding vehicles (4 by default), but also neighboring
vehicles on the adjacent travel lanes are taken into account. The
alertness of drivers approaching a traffic signal is increased
within 100 meters of a stop line.
[0037] Some of the VISSIM model inputs are listed below:
[0038] Behavior of the driver-vehicle-unit
[0039] Psycho-physical sensitivity thresholds
a. Ability to estimate distance b. Aggressiveness c. Memory of the
driver
[0040] Actual acceleration/deceleration based on current speed and
desired speed and
[0041] aggressiveness
[0042] Accessory usage based on weather, daylight and driver
preferences
[0043] Interdependence of driver-vehicle units
[0044] Rules to define relationships between leading and following
vehicles
[0045] Rules to define the relationship between vehicles in
adjacent travel lanes
[0046] Rules to define behavior at intersections and traffic
signals
[0047] Actual speed and acceleration
[0048] Behavior of driver-vehicle units with respect to the
road
[0049] Speed limits
[0050] Number of lanes
[0051] Behavior of driver-vehicle units with respect to traffic
[0052] Road model with speed limits, lanes, and gradient
[0053] Volume of vehicles in the road model
[0054] Distribution of vehicle lengths, top speeds and maximum
acceleration
[0055] Traffic control devices, timing etc.
[0056] The ability of the simulator to depict real life traffic
scenarios and driving behavior is extremely useful in understanding
the different road or traffic or driver characteristics that affect
the energy consumption of a battery electric vehicle.
[0057] sCVSP is the corporate standard tool for vehicle performance
and fuel economy modeling and simulation. Among its main features
are:
[0058] Used on Ford vehicle programs to set Performance & Fuel
Economy targets.
[0059] Model architecture and subsystem interfaces allow
interchange subsystem and component models based on vehicle
hardware. A global bus enables the communication between the
vehicle system control (VSC) and vehicle components.
[0060] Includes extensive set of component models that have been
developed over the years and are validated with test data.
[0061] Includes extensive vehicle and component parameter database.
These parameters can be calibrated and optimized to improve vehicle
performance.
[0062] Supported by company-wide processes to generate vehicle and
component parameter data for new programs.
[0063] Includes standard test management and report generating
capabilities that allow design engineers understand the behavior of
components, subsystems and the vehicle.
[0064] Features and capabilities can be extended by users
a. New models can be added to existing libraries b. New libraries
with new models can be added
[0065] In order to have a on-line capability to predict the
distance to empty in BEVs, sCVSP energy usage results are computed
in advance and recorded in a table as shown below. Each entry in
the table is the work needed for locomotion in Wh/mile for a given
speed, acceleration, ground grade, accessory load and vehicle
weight. The vehicle weight was simplified and parameterized by the
number of passengers in the vehicle assuming 150 lbs for a
passenger. The work is provided at the battery terminals as well as
at the wheels The former value includes parasitic losses in the
powertrain but not parasitic losses in the battery.
[0066] In the calculation the large table is reduced to separate
2-dimensional sub-tables for a specific accLoad and number of
passengers. The sub-tables have two variables remaining, % grade
and VehSpeed, that are the only variables that change during a
single drive cycle. The sub-tables are further reduced to a cubic
spline surface dimensioned by % grade and vehicle speed. The values
computed by sCVSP become the corner nodes for each value in the
table. These cubic-spline surfaces are then used to estimate the
power from the drive cycle, with vehicle acceleration and actual
road grade combined into the single % grade value.
TABLE-US-00001 acc.sub.-- grade.sub.-- speed.sub.-- batt.sub.--
batt.sub.-- batt.sub.-- batt.sub.-- whl.sub.-- whl.sub.--
whl.sub.-- whl.sub.-- load_Watt perc kph whr . . . whr . . . whr .
. . whr . . . whr . . . whr . . . whr . . . whr . . . 400.000 -6.00
10.0 -212.72 -225.22 -237.70 -250.20 -376.63 -392.66 -408.67
-424.70 400.000 -6.00 30.0 -288.43 -302.73 -317.01 -331.31 -354.00
-369.90 -385.78 -401.68 400.000 -6.00 50.0 -273.35 -287.92 -302.46
-317.01 -318.89 -334.68 -350.45 -366.23 400.000 -6.00 70.0 -234.25
-248.99 -263.60 -278.22 -270.65 -286.35 -302.01 -317.68 400.000
-6.00 90.0 -177.01 -191.77 -206.49 -221.23 -208.58 -224.18 -239.75
-255.34 400.000 -6.00 110.0 -104.60 -119.24 -133.85 -148.47 -133.27
-148.79 -164.27 -179.77 400.000 -6.00 130.0 -19.74 -34.20 -48.60
-63.02 -44.44 -59.90 -75.31 -90.74 400.000 -4.00 10.0 -96.92
-104.94 -112.94 -120.95 -229.18 -239.28 -249.35 -259.44 400.000
-4.00 30.0 -154.64 -163.76 -172.86 -181.97 -206.56 -216.53 -226.47
-236.44 400.000 -4.00 50.0 -136.31 -145.50 -154.66 -163.83 -171.46
-181.32 -191.15 -200.99 400.000 -4.00 70.0 -95.61 -104.78 -113.93
-123.09 -123.23 -132.99 -142.71 -152.46 400.000 -4.00 90.0 -37.60
-46.74 -55.85 -64.97 -61.17 -70.84 -80.47 -90.12 400.000 -4.00
110.0 36.10 25.94 16.41 7.38 14.14 4.55 -5.00 -14.56 400.000 -4.00
130.0 131.83 121.66 111.53 101.38 102.95 93.43 83.95 74.46 400.000
-2.00 10.0 20.48 17.19 13.92 10.63 -81.38 -85.52 -89.64 -93.78
400.000 -2.00 30.0 -19.47 -23.15 -26.80 -30.47 -58.77 -62.78 -66.78
-70.79
[0067] sCVSP was also used to compute the maximum acceleration
verses time as an input into VISSIM. sCVSP simulations of the three
EPA cycles demonstrate the extremes of maximum accelerations
imposed by the sCVSP model of the vehicle. It is necessary that
VISSIM and sCVSP have the same acceleration limits or VISSIM will
generate accelerations that can not be achieved by the vehicle. The
lower bound of acceleration (maximum deceleration) was -0.85 at
zero mph, varying linearly to -0.75 at 80 mph; theoretically it may
be possible to have greater decelerations.
[0068] FIG. 1 shows an illustrative example of a fuel efficiency
testing process. In this illustrative embodiment, a road network
model is first established 201. In at least one example, the model
is established in a VISSIM environment.
[0069] Next, a plurality of scenarios are setup under which testing
conditions can be performed on the road model 203. Multiple
replications of the scenarios are run to establish baseline results
204, providing aggregate data with a high degree of accuracy.
[0070] For each relevant virtual vehicle in a given scenario,
speed, acceleration, distance traveled, etc. are obtained 205, and
this data is fed into calculation software 207. Fuel consumption
for that vehicle is then calculated based on the inputs 208. From
this data, total energy consumption can then be determined 209.
[0071] After a given scenario is completed, the process can advance
to a next scenario 211.
[0072] FIG. 2 shows an illustrative example of a road network
modeling process. The road network under consideration is set-up in
the traffic simulator. The geometry 301 and length of the road 303,
number of lanes 305, vehicle flows 307, vehicle compositions 309,
desired speeds 311, traffic signal data 313, the driver model 315,
etc. are some of the inputs that may need to be set-up before
running the simulation. Additional inputs may be added to the
simulator as desired, and not all of the previously mentioned
inputs need to be used.
[0073] FIG. 3 shows an illustrative example of a scenario setup
process. In this illustrative example, a particular scenario is
selected for initialization 401. Road characteristics may be input
403 if desired. For example, without limitation, gradient 405
and/or number of lanes 407 may be adjusted.
[0074] Also, in this embodiment, traffic characteristics 409 may be
input for the scenario. This may include, but are not limited to, a
vehicle flow rate 411 and a vehicle mix 413.
[0075] Further, in this illustrative embodiment, driver
characteristics may be input 415 to represent certain driving
behaviors. These characteristics may include, but are not limited
to, driver speeds 417 and cruise control usage data 419.
[0076] Also, in this example, weather data may be input for the
scenario 421. This data may include, but is not limited to,
visibility adjustments 423 and temperature adjustments 425.
[0077] FIG. 4 shows an illustrative example of an energy
consumption determination process.
[0078] The parameters used to calculate the energy consumed by a
particular battery electric vehicle during a simulation run may
include: speed 503, acceleration 505, distance travelled 507,
accessory loads and number of passengers. VISSIM can output the
speed, acceleration and distance travelled by all the vehicles in
the simulation at every instant (every second in this case). The
number of passengers is fixed at one and the accessory loads are
assumed to remain fixed throughout the simulation run. It should be
noted that the accessory loads affect only the eventual energy
consumption and not the drive cycle.
[0079] Using the energy tables from sCVSP the outputs from VISSIM
are processed in MATLAB (or C-code on board the vehicle) 509 to get
the energy consumption by the battery electric vehicles in a given
scenario 511. The acceleration/deceleration values from VISSIM are
mapped into corresponding gradient values and thereby taking into
account the regenerative ability of a battery electric vehicle
during braking 513. A MATLAB/C code calculates the energy consumed
at each and every time instant for each electric vehicle and sums
them to give the total energy consumption 515. Regenerative
capability of a battery electric vehicle is considered in the
energy tables.
[0080] The models used by the traffic simulator are stochastic in
nature. Hence, for a given scenario and a particular simulation
run, each of the battery electric vehicles will have a different
drive cycle and therefore different energy usage. A characteristic
of a particular simulation is the statistical variance of the
energy utilization of large samples of vehicle. The sample size can
be determined by plotting the variance (or standard deviation)
verses sample size and observing the point at which the variance
approaches a steady state. Based this analysis 120 vehicles was
selected as a reasonable sample size.
[0081] It was observed that, in this example, variation in standard
deviation is negligible once the sample size is more than 120.
Hence, averaging the energy values of 120 vehicles for each
scenario would provide good statistics. In order to get a sample
size of at least 120 vehicles, the simulation needs to be run
multiple times depending on the scenario being tested. For example,
assuming the flow to be 2000 vehicle per hour with 2% battery
electric vehicles and a simulation time of one hour, at least 4
simulation runs need to be performed to get a good sample of 120
vehicles. It should be noted that in a particular run, only those
vehicles are chosen which traverse the whole road length. Vehicles
which are unable to complete the whole trip are excluded from the
energy calculations.
[0082] Presenting the mean energy consumed in a given scenario is
not enough. It is important to know the variation in the values.
Hence, the mean energy values are associated with a `confidence
interval`. A confidence interval is a range of numbers relevant to
the parameter of interest. For example, a 95% confidence interval
means that if we repeatedly draw samples of a given size N from a
certain population and we construct a confidence interval for each
sample, then 95% of these intervals on average will contain the
true value of the unknown parameter as an interior point. It is
incorrect to interpret a 95% confidence interval to mean that there
is a 95% chance that the interval contains the true value of the
unknown parameter as an interior point. This is because there is
one value of the unknown parameter, and the confidence interval
either contains this value or does not contain it.
[0083] Thus, confidence intervals should not be interpreted as
probabilities but should rather be interpreted in the context of
repeated sampling.
[0084] Through testing, it can be seen that gradient has a
prominent effect on the energy consumption of a battery electric
vehicle. There is a rapid increase in the energy values as we move
from a gradient of -4% to 4%. This is because, the vehicle needs
more energy to climb uphill (positive gradient) and it can gain
energy through regenerative braking while going downhill (negative
gradient). Relatively, congestion doesn't seem to have a big effect
on the energy usage.
[0085] There is a slight reduction in energy as the flow conditions
approach a congested scenario. This effect is directly related to
the decrease in the speeds for congested flows.
[0086] Also, speed of the vehicles has a big impact on the energy
consumption. A vehicle travelling at around 100 km/hr will consume
about 30% more energy than a vehicle travelling at around 80 km/hr.
Also, the number of lanes on the freeway doesn't seem to have an
effect on the overall energy consumption.
[0087] It can be seen that the vehicles travelling in cruise
control use significantly lower energy than the vehicles travelling
without cruise. The fluctuations in acceleration/deceleration and
hence the speed results in higher energy consumption for vehicles
which are not using cruise control. The drop in energy consumption
with increase in flow values is directly related to the drop in
speeds as the flow conditions become congested. It should also be
noted that the energy consumption and its variation remains fixed
across different flow values when the cruise control is set to 80
km/hr. But, in the case of cruise control at 100 km/hr, there is a
larger statistical variance in energy usage (dotted lines diverge)
as the flow values increase. This is because at high flows, the
vehicles are unable to maintain a cruise speed of 100 km/hr due to
the increase in flow density. But, the vehicles seem to maintain a
cruise speed of 80 km/hr even when the traffic flow increases.
[0088] The effect of accessory loads is comparable to the effect of
flow conditions. For example, a vehicle travelling in a congested
road (6000 vehicles per hour) and a hot weather (2000 W accessory
load) uses almost the same energy as that used by a vehicle
travelling in free flow (2000 vehicles per hour) and a cold weather
(800 W accessory load).
[0089] For a given accessory load the energy usage is the lowest
for a residential road and highest for a freeway, mainly because of
low speeds and stop-go nature of traffic on a residential road.
[0090] In fact, the energy usage per mile with 400 W accessory load
is more than halved from a freeway to a residential road. But, the
travel time on a residential road is almost four times that on a
freeway. This shows a trade-off between travel times and energy
usage. Increase in the accessory loads has very little impact on
the energy usage on a freeway where high speed is the primary
driver of energy consumed. On the other hand, the accessory loads
drastically affect the energy usage on a residential road to such
an extent that, the energy used per mile with 2000 W accessory load
is almost the same as that on an urban road.
[0091] The energy consumption results from various scenarios across
different road types have been analyzed to understand the various
factors that affect energy consumption of a battery electric
vehicle. The results show that road gradient has a very significant
effect on the energy usage across all the three road
types--freeway, urban roads and residential roads. Accessory loads
have a strong effect across different road types. The results show
that at very high accessory loads the energy usage on a residential
road is equal to the energy used on an urban road, although there
is a difference in the desired speeds on these road types.
[0092] The speed of the vehicles also has a prominent effect on the
energy usage. Cruise control on freeways helps in reducing the
energy cost. Also, significant energy gains are possible in stop-go
traffic scenarios because of regenerative braking in a battery
electric vehicle. Hence, urban roads with traffic signals and
residential roads are likely to be preferred over freeways to
achieve lower energy consumption. But, it should be noted that
there is a trade-off here, between energy consumed and travel
times.
[0093] The results can be used to develop cost functions that can
evaluate the total energy consumed along various possible routes
between an origin and destination and finally give the customer the
minimum energy route. One way of using the results in the cost
function is through the use of energy look-up tables and polynomial
curve fitting. For example, any route can be broken up into
segments which have the same characteristic, (either road type or
gradient or speed limits etc) and each segment can be assigned a
cost which is equal to the energy consumed by the vehicle to travel
that segment. Adding up the costs across all the segments will give
the overall cost to travel that particular route. This can be done
for all possible routes between two locations and the final output
could be the route which uses the lowest amount of energy. Energies
on different segments can be calculated by fitting a polynomial
curve to available data.
[0094] Accessory loads can be related to the weather and
temperature conditions. Hence, overall, accessory loads will have a
significant effect in deciding which road type to choose while
making a certain trip.
[0095] There is hardly any change in the energy consumption of a
battery electric vehicle when the % of heavy goods vehicles is
increase from 4% to 10%. A significant change in energy consumption
`might` occur when the % of heavy goods vehicles is increased to an
even higher value.
[0096] While exemplary embodiments are described above, it is not
intended that these embodiments describe all possible forms of the
invention. Rather, the words used in the specification are words of
description rather than limitation, and it is understood that
various changes may be made without departing from the spirit and
scope of the invention. Additionally, the features of various
implementing embodiments may be combined to form further
embodiments of the invention.
* * * * *