Hybrid Solar Vehicles: Perspectives, Problems, Management Strategies

1
Hybrid Solar Vehicles Perspectives, Problems,
Management Strategies

• HYATT REGENCY
• NOVEMBER 13-14, 2008
• ISTANBUL

I.Arsie, G.Rizzo, M.SorrentinoDIMEC, University
of Salerno, Italy
2
Outline

• Introduction
• HSV models and results
• Optimization of Management Strategies
• The Prototype
• Conclusions

3
The background
4
From conferences to cartoons
5
Possible Solutions?

• Kyoto Protocol A possible solution to fossil
  fuel depletion and global warming is an increased
  recourse to Renewable Energy (RE).
• Possible application to cars
• Fuels/Energy from RE (Bio-Fuels, H2)
• Solar Cars

6
Solar Energy
7
Solar Energy vs. Energy Consumption
The solar energy striking the US in one day is
almost equivalent to the energy consumption for
one and a half year
8
PV Panels
Today's most common PV devices use a
single-junction with poli-crystalline silicon,
with efficiency of about 12 Use of
mono-crystalline silicon results in higher
efficiency (15 and more)
Multi-junction cell
Much of today's research in multi-junction cells
focuses on gallium arsenide as one of the
component cells. Such cells have reached
efficiencies of around 40 under concentrated
sunlight (Fresnel lens).
9
PV efficiency trends
10
Solar Panels Production and Prices
The production of photovoltaic panels has
remarkably increased since 90s in terms of
installed power.
Their cost, after a continuous decrease and an
inversion of the trend occurred in 2004, appears
now quite stable
11
Outline

• Introduction
• HSV models and results
• Optimization of Management Strategies
• The Prototype
• Conclusions

12
Solar Cars
Various propotypes of solar cars have been
developed, for racing and demonstrative use
13
Limits of Solar Cars

• Solar Cars do not represent realistic alternative
  to normal cars, due to
• Limited power and performance.
• Limited range.
• Discontinuous energy source.
• High cost.

14
Hybrid Electric Vehicles
F.Porsche, 1900
Buick Skylark, 1974
Toyota Prius
Ford Escape
Honda Insight
GM Precept
Mercedes S400 Hybrid-Diesel
Peugeot 308 Hybrid-Diesel
15
HEV and PV a possible marriage?
16
About the dowry
Conventional/Hybrid car PV panels
Energy 600 KWh ?50 kg gasoline tank lt50 KWh/day 6 m2 _at_ 8.5KWh/m2/day
Power ?100 KW lt 1 KW
17
Energy Balance in a Solar Car
Net solar energy available to propulsion KWh/day
esunaverage insolation (KWh/m2day) APVeffective
panel area APV,H0.5 APV,V ?PVpanel efficiency
(0.13) ? reduction factor due to
charge/discharge processes in battery (0.9) ?
insulation reduction during driving, due to
shadow (0.9)
18
Solar Fraction
6 m2_at_12 or 3 m2_at_24
Driving hours per day
Solar energy can represent a significant
contribution for intermittent use (h1-2) and for
limited average power.
For average power from 5 to 10 KW and driving
hours from 1 to 2, solar contribution ranges from
18 to 60.
Site San Antonio, Texas Yearly Averaged Data
19
Statistics on Car Drivers
Some recent studies of the UK government stated
that

• about 71 of UK users reaches their office by car
• 46 of them have trips shorter than 20 min
• mostly with only one person on board.

Source Labour Force Survey, http//www.statistic
s.gov.uk/CCI/nscl.asp?ID8027
20
Power Demand
Extra-urban
Urban
Power demand can be determined integrating the
longitudinal vehicle model over a mission cycle.
During urban drive, limited average power can be
required to drive a small car.
Mass1000 Kg - Length3.75 m
21
Effects of Position on Energy
Average Yearly Energy (KWh/year)
Almost a factor 2 between maximum and minimum
latitudes.
For fixed panels, there is not a relevant loss by
adopting horizontal position with respect to
optimal tilt, particularly at low latitudes.
Energy absorbed with vertical position is
significantly lower, mainly at low latitudes.
Latitude (deg)
22
Some HSV prototypes
Viking 23 Western Washington University
Tokyo University of Agriculture and Technology
Solar Toyota Prius By Steve Lapp
Ultra-Commuter The University of Queensland
23
Solar Prius
Prius with an aftermarket 215 W monocristalline
solar module with peak power tracking and a 95
efficiency DC-DC Converter
It is estimated that the PV Prius will consume
somewhere between 17 and 29 less gasoline than
the stock Prius (range per day 5-8 miles)
24
Well to Wheel
25
HSV vs HEV
HEV ? Conventional Car Electric Motor
HSV ? HEV PV
26
HSV vs HEV

• Mission profile (HSV should be optimized for
  urban driving)
• Different SOC management strategies.
• Different structure (vehicle dimension, hybrid
  architecture)

27
HSV vs HEV control
28
Potential advantages of Series HSV

• No mechanical link between generator and wheels
• Very effective vibration insulation can be
  achieved
• Less constraints for vehicle layout
• Possible use of in-wheel motors with advanced
  traction control techniques
• Engines optimized for steady operation can be
  used
• ICE designed and optimized for steady conditions
• D.I. Stratified charge engine (4 or 2 strokes)
• Micro gas turbine
• Series architecture acts as a bridge towards the
  introduction of fuel cell powertrains.
• More suitable for V2G applications

29
Vehicle to Grid (V2G)

• V2G concept to connect parked electric driven
  vehicles (electric, hybrid, hybrid solar,
  fuel-cell) to the grid by a two-way computer
  controlled hook up.
• The power capacity of the automotive fleet is
  about 10 times greater than the electrical
  generating plants (in US) and is idle over the
  95.

• Advantages
• Reduction of costs for peak power production.
• Toward the distributed generation, with reduction
  of Transmission and Distribution (TD) costs.
• Facilitate integration of intermittent renewable
  resources.
• The value of the utility exceeds the costs for
  the two-way hook up and for the reduced vehicle
  battery life.

30
V2G Additional advantages for HSV
31
Engine control in a series HSV
In a series HSV, the Internal Combustion Engine
could operate on the optimum efficiency curve and
whenever possible at its maximum efficiency
ICE Efficiency
ICE Power
32
SHM Operating Modes /1
Parking
PV Panels
with sunlight
VMU
ICE
EG
EM
Battery
33
SHM Operating Modes /2
Hybrid
PV Panels
with sunlight
VMU
ICE
EG
EM
Battery
34
SHM Operating Modes /3
Electric Driving
PV Panels
with sunlight
VMU
ICE
EG
EM
Battery
35
SHM Operating Modes /4
Regenerative Braking
PV Panels
with sunlight
VMU
ICE
EG
EM
Battery
36
SHM Operating Modes /5
Recharge from grid
PV Panels
with sunlight
VMU
ICE
EG
EM
Battery
37
SHM Operating Modes /6
We believe that the most plausible vehicle of
the future is a plug-in hybrid... (Center for
Energy and Climate Solutions, 2004)
Power to grid (V2G)
PV Panels
Thermal load
with sunlight
heat
VMU
ICE
EG
EM
Battery
38
Flow Chart
DESIGN SPECIFICATION Power demand Insolation
HSV Structure
CONTROL VARIABLES Control Strategy for EG MPPT
for PV
DESIGN VARIABLES PV Panel Area and Position EG
and EM Power Car dimensions Materials
EXHOGENOUS VARIABLES Fuel Price Panel
Efficiency Unit weight and costs
MODELS Energy Flows for HSV/CC Car sizing -
Weight - Cost
OUTPUT Car Stability Fuel Savings Weight -
Payback
Objective Function and Constraints
39
Payback Optimization
? Objective Function minimum Payback
? Inequality Constraints

• Design variables X
• Electric Generator Power PEG
• Electric Motor Power PEM
• Horizontal panel area APV,H
• Vertical panel area APV,V
• Car length l
• Car width w
• Car Height h
• Weight reduction factor of car chassis with
  respect to base value CWf

Solved by Sequential Quadratic Programming
(Matlab routine FMINCON)
40
Constraint Specification
EG Power within lower and upper bounds
41
Optimal design results
cf /kg cPV /m2 /W ?P / APV,H m2 PEG kW PB yrs
1 1.77 800/6.15 0.13 0 35.5 6.1
2 1.77 800/6.15 0.13 3 35.5 9.9
3 1.77 200/2.15 0.13 4 37 5.6
4 3.54 200/2.15 0.16 5.6 38.4 2.4
Fuel Price 2.1 /KG Italy, June, 2008
PV Retail Price June 2008 4.70 /W
42
MPPT Techniques
Uniform working conditions

• Due to changing sun irradiance, PV source must be
  matched to the load to draw maximum power.
• Maximum Power Point Tracking (MPPT) techniques
  are adopted.
• The presence of local maxima occur during
  mismatched conditions, due to shading effects and
  temperature variations in different parts of the
  panel.
• The characteristic may change rapidly during
  driving conditions, required advanced MPPT
  control.

Mismatched PV field
Power vs. voltage characteristic of a PV field
43
Sources of mismatching

• Different solar irradiation levels due to
• Clouds
• Shadows
• Different orientation of parts of the PV field
• Dirtiness
• Tolerances (due to manufacturing and/or ageing)
• Different types of panels (different models,
  photo-glass, coloured) in the same string

44
MPPT management of PV array

• MPPT strategy are implemented to maximizing PV
  efficiency throughout the day.

Max Allowable Power
Power given to the battery
P
Vi
MPPT
Vi
45
Outline

• Introduction
• HSV models and results
• Optimization of Management Strategies
• The Prototype
• Conclusions

46
HSV Modeling

• Longitudinal model of the HSV protoype

? Power at wheels
? EM Power
? Battery recharge power
experimentally characterized
47
Experimental characterization EG and EM

• The electric generator was characterized
  connecting a pure resistive electrical load.
• A 4 order polynomial regression was obtained.

• EM efficiency is modeled by a 3rd order
  polynomial regression identified vs. manufacturer
  technical data.

48
Experimental characterization the Battery pack

• Battery is modeled applying the Kirchoff law to
  an equivalent circuit.
• The internal resistance was modeled as a
  nonlinear function of state of charge.
• Model accuracy was checked against experiments.

49
Experimental characterization the PV array

• The PV array has been characterized by connecting
  the converter output to a resistive load.

hPV 10 (390 W/m2 irradiation)

• The average PV daily energy was derived from an
  experimental year-thorough distribution

50
ICE thermal transients
Engine temperature dynamics is estimated by a
first order dynamic model
Steady state temperatures and time constants are
assigned for ICE on and ICE off events
51
Thermal effects on power and SFC
Specific Fuel Consumption and power are related
to the ratio between actual temperature and its
steady state value, starting from experimental
data for a SI engine
52
Modeling of HC emissions

• Due to the ICE intermittent use, HC emissions
  occurring during warm-up have to be accounted
  for.

Experimental warm-up HC dynamics
53
Energy management strategy

• In case of ICE intermittent use, energy
  management for HSV can be addressed via an
  optimization analysis.

• The decision variables X include number of ICE
  starts, starting time, duration and ICE power
  level.

Minimum and maximum values considered for state
of charge

• Day through charge sustaining is achieved
  constraining SOC variations.

54
Simulation of HSV prototype scenario analysis

• The prototype was simulated on a driving cycle
  composed of 4 ECE-like modules.

HSV Specification
ICE power kW 46
Fuel gasoline
PEG kW 43
PEM kW 90
Number of battery modules / 27
PV horizontal surface APV,H m2 1.44
Coefficient of drag (Cd) 0.4
Frontal area m2 2.6
Weight kg 1465
55
Control optimization results (DBM) 1/3
PEG,N4 gt PEG,N2

• Initially fuel economy increases with engine
  starts due to the higher degrees of freedom.
• After 4 ICE-on, fuel economy tends to decrease
  due to the increasing impact of thermal
  transients.

N
56
Control optimization results (DBM) 2/3

• HC emissions show an increasing trend with number
  of starts.
• A local minimum occurs at N 4.
• Such a behavior is due to the different
  temperature trajectories.

57
Control optimization results (DBM) 3/3

• On average EG operating conditions fall in a high
  efficiency region.

• SOC excursions are satisfactorily bounded
• Final SOC leaves room for PV charging during
  parking phases

58
Energy management optimization by means of
genetic algorithm (GA) search

• As both integer and real variables are involved,
  the GA search method was selected for such an
  analysis.
• HC emissions and fuel consumption for cranking
  energy have been also included in the objective
  function.

Binary representation of the optimization problem
GA parameters
Decision variable Definition range Precision Number of bits
NEG 1 8 1 3
tEG (min) 0 78/ NEG 0.073/ NEG 10
DtEG (min) 0 78/ NEG 0.073/ NEG 10
PEG (kW) 0 43 0.040 10
Population size 70
Number of generations 100
Crossover probability 0.8
Mutation probability 0.033
59
Control optimization results (GA) 1/2
60
Control optimization results (GA) 2/2
To be recovered in the parking phase
61
Comparison between GA and DBM results
Optimization outputs DBM GA 1
NEG 4 3
Fuel consumption kg and saving () 2.41 (14.8) 2.48 (12.4)
HC emissions 1 (g) 1.13 0.85
Average engine temperature C 65 68
Max SOC / 0.79 0.88
Min SOC / 0.65 0.58
HC emissions 2 (g/km) 0.025 0.018
A further optimization analysis was run
considering an increase in PV horizontal area
from 1.44 m2 to 3 m2. Such configuration upgrade
results in a fuel consumption reduction from 2.48
kg to 2.28 kg (19.4 saving).
() conventional vehicle fuel consumption 2.83
Kg
62
Outline

• Introduction
• HSV models and results
• Optimization of Management Strategies
• The Prototype
• Conclusions

63
HSV Prototype
Vehicle Piaggio Porter
Length 3.370 m
Width 1.395 m
Height 1.870 m
Drive ratio 14.875
Electric Motor BRUSA MV 200 84 V
Continuous Power 9 KW
Peak Power 15 KW
Batteries 16 6V Modules Pb-Gel
Mass 520 Kg
Capacity 180 Ah
Photovoltaic Panels Polycrystalline
Surface 1.44 m2
Weight 60 kg
Efficiency 0.13
Electric Generator Diesel Yanmar S 6000
Power COP/LTP 5.67/6.92 kVA
Specific fuel cons. 272 g/kWh
Weight 120 kg
Overall weight (with driver)
Weight 1950 kg
A prototype of hybrid solar vehicle with series
structure has been developed at the University of
Salerno, within the EU Leonardo Program Energy
Conversion Systems and Their Environmental
Impact (www.dimec.unisa.it/leonardo)
64
A research and educational project
Leonardo Program (I05/B/P/PP-154181) Energy
Conversion Systems and Their Environmental Impact
http//www.dimec.unisa.it/leonardo
Sponsored by ACS, Salerno (I), Lombardini (I),
Saggese (I).
65
The web site www.dimec.unisa.it/Leonardo
A multi-lingual web site has been developed. The
site has more than 1000 visits per week and is at
the top positions on Google.
66
Participation to the FIA Alternative Energies Cup
race ECO-TARGA FLORIO (Palermo, Italy)
67
Outline

• Introduction
• HSV models and results
• Optimization of Management Strategies
• The Prototype
• Conclusions

68
Conclusions

• Hybrid Solar Vehicles can represent a valuable
  solution for energy saving and environmental
  issues, but accurate re-design and optimization
  of both vehicle and powertrain with respect to
  HEV are required.
• Economic feasibility could be achieved in a near
  future, with realistic assumptions for component
  costs, fuel price and PV panel efficiency.
• Significant fuel savings can be obtained by
  proper ICE management strategies. Thermal
  transient effects on fuel consumption and HC
  emissions must be considered in case of
  intermittent use.
• The use of optimization techniques (GA, DBM) has
  allowed to select the best management strategies,
  to be used as benchmark for real-time
  implementable control.
• Interdisciplinary research is needed, but also a
  systematic dissemination of results and
  potentialities, in order to remove the obstacles
  to the diffusion of such vehicles.

69
On-going activities

• Development and implementation of real-time
  control strategies and comparison with benchmark
  solutions.
• On road tests on the prototype to validate both
  simulation results and control strategies.
• Installation of an automated sun-tracking roof to
  further enhance solar energy contribution.

70
HSV An artistic point of view
Thank you for your kind attention