Plug in Hybrid Powertrain Configuration - autonomie.net - Presentations/HEVs and PHEVs... · Glider...
Transcript of Plug in Hybrid Powertrain Configuration - autonomie.net - Presentations/HEVs and PHEVs... · Glider...
Plug‐in Hybrid Powertrain Configuration C i U i Gl b l O i i iComparison Using Global Optimization
2008 ASME Dynamic Systems and Control Conference,
Ann Arbor,
October 21st 2008October 21st, 2008
Dominik A. Karbowski, Karl‐Felix Freiherr von Pechmann, Sylvain Pagerit Jason KwonPagerit, Jason Kwon
Argonne National Laboratory
Sponsored by Lee Slezak
This presentation does not contain any proprietary, confidential, or otherwise restricted information
Outline
Introduction
Vehicle Sizing
Global Optimization Algorithm
Simulation Results
Conclusion
2
PHEVs Come in Various PowertrainConfigurations: How to Compare Them?Configurations: How to Compare Them? Plug‐in Hybrid Electric Vehicles (PHEVs) combine:
– long range, high power density and easy energy refill
– low tailpipe emissions, high fuel displacement (on short daily trips)
Various hybrid powertrain configurations are possible: how do they compare in a PHEV context ?
Prototypes exist but built upon different requirements sizes electric ranges Prototypes exist, but built upon different requirements, sizes, electric ranges, etc.
Chevrolet (GM) Volt(Series)(Series)
Toyota Prius PHEV(Power‐Split)
Mercedes Sprinter PHEV(Parallel Pre‐Transmission)
3
Simulation tools must be used for a fair comparison !
Choosing the Most Suitable Simulation Tool
Argonne’s PSAT– Forward‐looking model
l d
Real‐world Oriented
R lt d d t l– Includes Transients
– Rule‐based control
Results depends on control design and tuning
Global OptimizationTheoretical tool
tSOCtP nm ,
0X
...
– Backward‐looking model
– Steady‐state model
– Control is an output
Theoretical tool
Each Vehicle is operated optimally tJtSOCtP '0,000 ,1',1 tJtSOCtP MMMM ',' ,1,1
tU0 tU M
Control is an output
TX
...
Global optimization allows a “fair” comparison
4
Global optimization allows a fair comparison
Study ProcessDefine Vehicle Requirements Industry standards
Size VehiclesImplement + Automated
Sizing
Optimally Simulate Vehicles
optimization code for various
configurationsDifferent cycles
Sizing
Vehicles Different cycles,Different distance
Analyze data
5
Outline
Introduction
Vehicle Sizing
Global Optimization Algorithm
Simulation Results
Conclusion
6
Requirements and Assumptions
Type Requirement Comments
All‐electric Range 10 mi (16 km) on the UDDS 90% to 30% SOC
Grade 6%@65 mph (105 km/h) @GVWR, w/o using the battery
0‐60 mph (97 km/h) time 9.3s @ 30% SOC
M i d 100 h (160 k /h)Maximum speed >100 mph (160 km/h)
Parameter Value
Glid 1142 kGlider mass 1142 kg
Drag coefficient 0.28
Frontal area 2.3 m2
Comparable to Toyota Camry,Rolling resistance 0.009
Wheel Radius 0.332 m
Comparable to Toyota Camry, Chevrolet Malibu
7
Automated Sizing Routine
Motor and battery power sized for the UDDSMotor Power for Cycle
Vehicle Assumptions
UDDS
Engine power sized for the grade
Engine (and motor in series config.) size increased if 0 60 mph requirements not met
Battery Power
y
increased if 0‐60 mph requirements not met
Battery energy sized to meet the AER requirements
M i f ti f t
Engine Power
Battery Energy
Mass is function of components characteristics
Based on PSAT simulations
Convergence
Yes
No
8
Sizing Results
Configuration Mass (kg)
Parallel 1782
Power‐Split 1824
Series 1793Series 1793
9
Outline
Introduction
Vehicle Sizing
Global Optimization Algorithm
Simulation Results
Conclusion
10
Global Optimization Uses a Backward Model, Based on Look-up TablesBased on Look-up Tables
ICE ESSICE GEN
ICE ACCESS
Thermal Elec. Mech.
ACC
MG
ACC ESSMG2MG1
TC
GEARBOX
MG
DRIVELINE
TCTC
PLANETARY
DRIVELINEDRIVELINE
Parallel Pre Transmission
Power‐Split SeriesPre‐Transmission
Engine (ICE), electric machines (MG) : input power (thermal or electric) is function of output torque and speed
11
Battery voltage and resistance is function of SOC
Optimization Problem State: Battery State‐of‐charge (SOC)
Command u: – gear and engine torque (Parallel)g g q ( )
– generator power (Series)
– engine power (Power‐split)
Link between state and command:
Initial and final conditions: initial SOC between 90 and 30% , final SOC=30%,
Cost function to minimize:
Constraints:
– Vehicle follows drive trace
– Keep ICE and EM speed/torque within limits
– Keep battery current within limits
S i & lit t d d h th t f i ICE t– Series & power‐split: torque and speed such that for a given power ICE operates optimally
Method: application of Bellman’s principle
12
Algorithm Outputs Results for Different SOCSOC
SOC.9
1 vehicle1 cycle1 distance
calculation.8
.6
1 run .4
.3
0.60 0.1 0.3 0.5
SOC
time
Plug‐to‐Wheel Wh/km
Outline
Introduction
Vehicle Sizing
Global Optimization Algorithm
Simulation Results
Conclusion
14
In Series, Engine Starts “Later” and Is Used More Often To Charge The BatteryOften To Charge The Battery
Power at the Wheels Above Which ICE Is ON 95% of the Time
Power‐split and parallel follows similar
4x 104
paralleli Power split and parallel follows similar
trends
Series “Engine ON” threshold is higher => ICE is ON less often and at higher 2
3
wer
(W)
seriessplit
loads1
Pow
0.15
Cha
rgin
g
parallelUDDS x1
UDDS x1
0 50 100 1500
Electric energy consumption (Wh/km)
0.1
ergy
use
d fo
r C seriessplit
UDDS x1
Ratio of ICE Energy Used to Charge
0
0.05
atio
of I
CE
Ene Ratio of ICE Energy Used to Charge
the Battery
Much more Charging for the SeriesLoad‐following
15
0 50 100 1500
Electric energy consumption (Wh/km)
Ra
In Series Config., ICE is Used at Higher Power for Higher Efficiency Higher Efficiency
Average Engine Power (when ON)
Different trends:1.5
2x 104
Different trends:– Series = independent of Wh/km
– Parallel = higher for higher Wh/km
Series: ICE power is much higher
1
Pow
er (W
)
parallelUDDS x1 p g
0 50 100 1500
0.5
Electric energy consumption (Wh/km)
seriessplit
0 37
0.38 parallel
UDDS x1
Electric energy consumption (Wh/km)
0.35
0.36
0.37
ffici
ency
seriessplit Overall Engine Efficiency
Similar trends to average power
Series = engine operates close to maximal
0.32
0.33
0.34Eff Series engine operates close to maximal
efficiency (36%)
Parallel = much lower efficiency in CS mode
Power Split = in betweenUDDS x1
16
0 50 100 1500.32
Electric energy consumption (Wh/km)
p
On The UDDS Power Split Seems to Be the Best CompromiseCompromise
Charge‐Sustaining: Series fi f l d
5
m)
parallel
config. uses most fuel, due to poor electric transmission efficiency
Parallel uses the least fuel3
4
n (L
/100
km
parallelseriessplit
Parallel uses the least fuel, as engine‐to‐wheel path is more efficient
In electric‐only, parallel is 2
3
nsum
ptio
n
penalized by the transmission efficiency
Power‐Split performs well i b th it ti0
1
Fuel
Con
is both situations0 50 100 150
0
Electric energy consumption (Wh/km)
17
Similar Hierarchy When Using Other CyclesI h t i i d th diff b t th i In charge‐sustaining mode, the difference between the series configuration and the two other ones is more important on more energy‐intensive cycles (HWY) or aggressive (LA92)
Th diff b t ll l d th th i l i ifi t
6
The difference between rarallel and the other ones is less significant on LA92 or HWY
LA92Parallel
4
5
(L/1
00km
) HWY
UDDS32 km
ParallelSeriesPower‐Split
120
km/h
2
3
Con
sum
ptio
n LA92
HWY
UDDS16 km
∞ km
(charge‐sustaining)
0
0
0
0
0
0120
60
UDDS LA92
HWY
0
1
Fuel
LA92UDDS
HWYUDDS
12 km24 km
0 200 400 600 800 1000 1200 14000 500 1000 15000 200 400 600 80000HWY
s s s
0 50 100 150 200Electrical Energy Consumption (Wh/km)
18
Outline
Introduction
Vehicle Sizing
Global Optimization Algorithm
Simulation Results
Conclusion
19
Conclusion
Comparing different powertrains requires a common set of requirements
An automated routine using PSAT allows to quickly size vehicles
Global optimization can be used to ensure “fair” comparison, because control is optimal
Series is potentiallymore efficient on the “electric‐only” side, parallel (pre‐tx) is potentiallymore efficient on the “charge‐Sustaining” side, power‐split is a good compromise
Global optimization is a theoretical tool, and in real‐world optimal control is harder to get because:– it requires the knowledge of the cycle beforehand
– real‐world must also take into account drivability, safety, etc.
However it provides the theoretical boundaries, as well as control patterns
Future work will use patterns obtained from global optimization to implement adaptive rule‐based controls.
20