Power Electronics Architecture R&D

learnedbawledΗλεκτρονική - Συσκευές

24 Νοε 2013 (πριν από 3 χρόνια και 6 μήνες)

76 εμφανίσεις

Power Electronics
Architecture R&D

Omer C. Onar

This presentation does not contain any proprietary, confidential,
or otherwise restricted information

2013 U.S.
DOE

Hydrogen and Fuel
Cells Program and Vehicle
Technologies Program Annual Merit
Review and Peer Evaluation Meeting

Oak Ridge National Laboratory

May 14, 2013

Project ID: APE056

2

Overview


Start


FY13


Finish


FY15


22% complete


Boost ratio/range and efficiency that can
drive dc
-
dc converter architectural choices,


Potential cost addition due to choice of dc
-
dc converter and hybrid energy storage
systems



T
otal project funding


DOE

share


100%


Funding
for
FY13:
$ 375K


Timeline

Budget

Barriers

Partners


ORNL
-

Burak
Ozpineci,
Bradley
Brown,
Jianlin Li, Lixin Tang, Tim Burress, Madhu
Chinthavali,
Cliff White, Larry Seiber, Steven
Campbell




Targets Addressed


Traction Drive Power Electronics DOE 2020
targets


Power density: >13.4 kW/l


Specific power
:

>14.1 kW/kg


Service life: >15 years or 13000 hours

3

Project Objective


Overall Objective


Develop a bi
-
directional buck/boost dc
-
dc converter between the
regenerative energy storage systems (RESS) and the dc link (traction drive
inverter),


Active energy management, inverter efficiency improvement, better RESS utilization.


Design a hybrid battery/ultra
-
capacitor energy storage system architecture,


Improved regenerative braking performance, improved overall fuel economy (all
-
electric range), improved power density, peak power capability, and improved battery
lifetime.


Design a modular reconfigurable battery and dc
-
dc converter architecture,


Lower overall power electronics kVA rating and cost reduction on dc
-
dc converter.


FY13 Objective


Model, simulate, and analyze a modular reconfigurable dc
-
dc converter
architecture.

4

Milestones

Date

Milestones and Go/No
-
Go Decisions

Status

April 2013

Milestone
:
Model and simulate the modular reconfigurable
dc
-
dc converter structures that are best utilized to meet the
vehicle power demand.

On track

August
2013

Milestone
:
Model and simulate various hybrid

RESS
architectures.

On track

August
2013

Go/No
-
Go decision
:
CHANGE

if simulation results show
that hybrid RESS approach outperforms the reconfigurable and
modular RESS battery approach.




September
2013

Milestone
:
Prepare a summary report that documents the
results, findings, performance comparisons, and
recommendations to be incorporated into the annual VTO
report.

On track


5

Approach/Strategy

Regenerative energy
storage system
DC link
+
-
V
DC
M
Traction drive
inverter
+
-
V
Batt
RESS
-
DC link
interface
(
200
-
450
V
DC
)
(
720
-
1200
V
DC
)

Current State
-
of
-
the
-
art Traction Drive System


Single
battery pack utilized regardless of the power
demand,


Single
energy storage system (batteries)


䍯C灬敤⁰e睥爠慮a⁥湥牧礠
requirement,


Regenerative
Energy Storage System (RESS)
-

Motor drive inverter
interface converter High inverter current


䱯睥爠
敦晩捩敮捹


Future traction drive system layout

6

Approach/Strategy


Develop bidirectional dc
-
dc
converter
architectures:


Dual active bridge
dc
-
dc
converter,


Two
-
quadrant dc
-
dc buck
-
boost converter,


Resonant/improved dual active bridge converter,


Single stage dual active half
-
bridge dc
-
dc converter,


Current boosted active clamp forward dc
-
dc converter,


Bi
-
directional four switch buck
-
boost dc
-
dc converter,


Bi
-
directional dc
-
dc CUK converter,


Integrated buck/boost converter,


Multi
-
phase interleaved ZVS dc
-
dc converter.


Battery
T
1
T
2
L
1
AC
Grid
DC
/
DC
Converter
T
1
T
3
C
R
L
T
2
T
4
L
2
C
Traction
drive
inverter
Two
-
quadrant
dc
-
dc
buck
-
boost

Traction
drive
inverter
T
1
T
2
C
1
T
3
L
1
L
2
L
3
T
5
T
6
T
7
Battery
Interleaved 3
-
phase
dc
-
dc
converter

Battery
T
1
T
3
T
2
T
4
L
1
L
2
Trf
.
T
5
T
7
T
6
T
8
Traction
drive
inverter
Dual active
-
bridge
dc
-
dc
converter

Battery
+
+
-
-
S
1
S
2
C
1
C
2
C
3
L
1
L
2
Traction
drive
inverter
Bi
-
directional CUK

7

Approach/Strategy (Cont’d)

Modular reconfigurable RESS and
dc
-
dc
converters


Develop control systems for active energy management that have
potential for cost reduction and efficiency improvement.


The RESS and modular
dc
-
dc
converters will be
best utilized
to meet the vehicle power
demand.


Utilize wide
bandgap

devices,


Design for lower
overall kVA
rating and
cost,


Improved service life,


Reduce the thermal
management burden.

dc
-
dc
converter
1
Battery module
1
Battery module
2
Battery module N
dc
-
dc
converter
2
dc
-
dc
converter
N
Battery pack
Traction
drive
inverter
M
V
DC
+
-
8

Technical Accomplishments and
Progress


Overall



Reviewed and modeled bi
-
directional dc
-
dc converter architectures that
interface the RESS to the traction drive inverter and created a summary report
discussing on the operation principles, controls, advantages and drawbacks.


Reviewed and modeled hybrid RESS architectures and created a summary
report based on the advantages, drawbacks, control systems, performance,
number of parts, etc.


Selected and modeled four battery/UC hybridization strategies.


Built simulation models of the battery and UC.


Modeled hybridization architectures.


Due to simulation time constraints, a portion of the UDDS drive cycle, t=[690, 760] that
includes acceleration, braking, and idling simulated for these hybridization architectures.


Collected and compared simulation results.


Created a comparison results table for these RESS hybridization architectures.

9

Technical Accomplishments and
Progress


Developed power electronic interfaces for hybrid RESS:


Decoupled energy and power: battery/ultra
-
capacitor (UC)


Active power and energy management based on the drive train power demand

Bi
-
directional
dc
-
dc
Converter
DC link
Traction
Drive
Inverter
M
Battery
pack
UC
+
-
V
ess
+
-
V
dc
Passive Parallel
A
rchitecture (PPA)

Cascaded Converters Architecture (CCA)

DC link
Traction
Drive
Train
M
Battery
pack
UC
+
-
V
DC
+
-
V
UC
Bi
-
directional
dc
-
dc
Converter
Bi
-
directional
dc
-
dc
Converter
Parallel with Multi
-
Converters
Architecture (PMCA)

DC link
Traction
Drive
Inverter
M
Battery
pack
UC
+
-
V
DC
+
-
V
UC
Bi
-
directional
dc
-
dc
Converter
Bi
-
directional
dc
-
dc
Converter
UC


dc
-
dc Converter


Battery Architecture
(UCDCBA)

Bi
-
directional
dc
-
dc
Converter
Traction
Drive
Inverter
M
Battery
pack
UC
+
-
V
UC
+
-
V
DC
10

Technical Accomplishments and
Progress



Implemented a cell pack model needed in dynamic system simulations

+
-
Controlled voltage
source
R
int
V
b
i
b
Sgn fnc
.
)
BattType
,
Exp
,
i
,
s
/
i
(
f
V
)
BattType
,
Exp
,
i
,
s
/
i
(
f
V
b
b
*
e
arg
disch
_
b
b
b
*
e
arg
ch
_
b
2
1


s
1


1
1


s
)
t
(
i
.
B
/
A
)
s
(
Sel
)
s
(
Exp
b
*
b
V
Sel
Exp
s
/
i
b
b
i
+ (
charge
)
-

(
discharge
)

Utilized the governing charge and
discharge equations of the model


A
controlled voltage source
was
used
(current controlled voltage
source to represent
V=f(I) cell
characteristics)


Voltage was calculated
with a non
-
linear equation based on the
state
-
of
-
charge (SOC)
of the
cell;
V=f(I,SOC
)


Reference: O
. Tremblay and L.

A.
Dessaint
, “Experimental Validation of
a Battery Dynamic Model for EV
Applications,”
W
orld
Electric Vehicle
Journal
, vol. 3, May 13
-

16, 2009.


Cell pack model

Example discharge curves

11

Technical Accomplishments and
Progress



Completed
simulations for the
passive parallel,
cascaded, modified
cascaded, and the
multiple parallel
converters
architectures under
same conditions.


Due
to simulation
time constraints, a
portion of the UDDS
drive cycle, t=[690,
760] that includes
acceleration,
braking, and idling
simulated for these
hybridization
architectures
.

UDDS drive cycle power
demand

12

Technical Accomplishments and
Progress



Completed PPA (Passive Parallel Architecture) Simulations

S
R
Q
sensed
cnv
i
_
cnv
i
*
2
,
1
G
Switching
Oscilator
Comparator
S
-
R Latch
Σ
Peak Current Mode Controller
PID Controller
Converter reference current
ref
dc
V
_
load
V
+
-
s
K
sK
K
s
i
P
D


2
kHz
f
s
10

Voltage control loop
Current Control Loop
PPA Control System

690
700
710
720
730
740
750
760
-
30
-
20
-
10
0
10
20
30
40
50
60
70
Time
[
s
]
Load Current
[
A
]
690
700
710
720
730
740
750
760
0
10
20
30
40
50
60
70
Battery Current
[
A
]
Time
[
s
]
690
700
710
720
730
740
750
760
-
120
-
100
-
80
-
60
-
40
-
20
0
20
40
60
Ultra
-
capacitor Current
[
A
]
Time
[
s
]
690
700
710
720
730
740
750
760
86
.
5
87
87
.
5
88
88
.
5
89
89
.
5
90
90
.
5
Time
[
s
]
Battery SoC
[%]
690
700
710
720
730
740
750
760
75
80
85
90
95
100
Time
[
s
]
Ultra
-
capacitor SoC
[%]
Load current

Battery current

UC current

Battery SOC

UC SOC

690
700
710
720
730
740
750
760
300
320
340
360
380
400
420
440
460
480
500
DC Link Voltage
[
V
]
Time
[
s
]
DC link voltage

13

Technical Accomplishments and
Progress



Completed CCA (Cascaded Converters Architecture) Simulations

690
700
710
720
730
740
750
760
-
30
-
20
-
10
0
10
20
30
40
50
60
70
Time
[
s
]
Load Current
[
A
]
Load current

Battery current

UC current

Battery SOC

UC SOC

DC link voltage

690
700
710
720
730
740
750
760
-
10
0
10
20
30
40
50
60
70
80
90
Time
[
s
]
Battery Current
[
A
]
690
700
710
720
730
740
750
760
-
100
-
80
-
60
-
40
-
20
0
20
40
60
80
100
Time
[
s
]
Ultra
-
capacitor Current
[
A
]
690
700
710
720
730
740
750
760
86
86
.
5
87
87
.
5
88
88
.
5
89
89
.
5
90
90
.
5
Time
[
s
]
Battery SoC
[%]
690
700
710
720
730
740
750
760
75
80
85
90
95
100
Time
[
s
]
Ultra
-
capacitor SoC
[%]
690
700
710
720
730
740
750
760
300
320
340
360
380
400
420
440
460
480
500
Time
[
s
]
DC Link Voltage
[
V
]
14

Technical Accomplishments and
Progress



Completed CCA
(Cascaded Converters Architecture)

Simulations with modified controls

CCA Modified Control System

Battery current

UC current

Battery SOC

UC SOC

DC link voltage

batt
i
*
Π
)
(
s
G
LP
Low
-
pass
Bessel Filter
Battery reference current computation
load
i
batt
load
V
V


'
i
batt
*
Rate limiter
Saturation
690
700
710
720
730
740
750
760
0
5
10
15
20
25
30
35
40
45
50
Time
[
s
]
Battery Current
[
A
]
690
700
710
720
730
740
750
760
-
150
-
100
-
50
0
50
100
Time
[
s
]
Ultra
-
capacitor Current
[
A
]
690
700
710
720
730
740
750
760
86
.
5
87
87
.
5
88
88
.
5
89
89
.
5
90
90
.
5
Time
[
s
]
Battery SoC
[%]
690
700
710
720
730
740
750
760
83
84
85
86
87
88
89
90
91
Time
[
s
]
Ultra
-
capacitor SoC
[%]
690
700
710
720
730
740
750
760
300
320
340
360
380
400
420
440
460
480
500
Time
[
s
]
DC Link Voltage
[
V
]
15

Technical Accomplishments and
Progress



Completed MPCA (Multiple Parallel Converters Architecture) Simulations

Battery current

UC current

Battery SOC

UC SOC

DC link voltage

690
700
710
720
730
740
750
760
-
10
0
10
20
30
40
50
Time
[
s
]
Battery Current
[
A
]
Time
[
s
]
Ultra
-
capacitor Current
[
A
]
690
700
710
720
730
740
750
760
-
60
-
40
-
20
0
20
40
60
Time
[
s
]
Battery SoC
[%]
690
700
710
720
730
740
750
760
87
87
.
5
88
88
.
5
89
89
.
5
90
90
.
5
Time
[
s
]
Ultra
-
capacitor SoC
[%]
690
700
710
720
730
740
750
760
84
85
86
87
88
89
90
91
Time
[
s
]
DC Link Voltage
[
V
]
690
700
710
720
730
740
750
760
300
320
340
360
380
400
420
440
460
480
500
690
700
710
720
730
740
750
760
-
30
-
20
-
10
0
10
20
30
40
50
60
70
Time
[
s
]
Load Current
[
A
]
Load current

16

Technical Accomplishments and
Progress
(Cont’d)



Created Results Table

Criteria

Passive
Parallel
(PPA)

Cascaded
Converters
(CCA)

Cascaded
(manipulated
controls
) (CCA)

Parallel
Converters
(MPCA)

Control simplicity

1

2

3

3

Structure complexity

1

2

2

2

Number of converters

1

2

2

2

Number of inductors

1

2

2

2

Total inductor mass

2

3

3

2

Number of transducers

5

6

6

6

Cycle
-
end battery
SoC

86.7194%

86.2462%

86.6598%

87.0305%

Cycle
-
end ultra
-
capacitor
SoC

89.8995%

91.9084%

90.4502%

87.1022%

Maximum battery current ripple

~7 [A]

~9 [A]

~1.7 [A]

~1.8 [A]

Cycle based
energy efficiency

(RESS and dc
-
dc

converters combined
)

95.25%

90.34%

90.72%

95.25%

Maximum DC link voltage variation
percentage

2.52%

2.42%

2.51%

0.77%

17

Conclusions



According to the simulation results, PPA has the simplest structure and the
least number of parts and components.


The high efficiency of PPA is mainly due to the simple configuration and to
the fact that there is no additional dc
-
dc converters used for hybrid RESS.


Although PPA has high efficiency, it does not provide control flexibility and
the battery current ripple and DC link voltage ripple values are not as good
as MPC architecture.


MPCA provides the highest efficiency (as high as the PPA) and the best DC
link voltage and battery current ripples.


Improving the controls with an additional current rate limiter for the CCA
improved the efficiency and the overall performance.


Efficiency is computed as the cycle based energy efficiency; i.e., the input
and output power of the system is integrated over the time period of
simulation.


18

Collaboration and
Coordination

Organization

Type of Collaboration/Coordination

Maxwell,
IOXUS

Fast response electrochemical capacitor
development

Chrysler

Power electronics dc
-
dc interface trends for
RESS

ORNL Energy Storage

Program

Design guidelines and research on modular
battery pack configuration

ORNL Battery Manufacturing Facility

Manufacturing research on modular battery
development

19

Proposed Future
Work


Remainder
of
FY13


Modeling, simulations, and analysis of modular reconfigurable dc
-
dc
converter architectures. Share results with APEEM team members.


FY14


Fabricate and test
a
candidate 10 kW reconfigurable dc
-
dc
converter
architecture for experimental
validation of models and
simulations.
Share results with APEEM team members.


FY15


Fabricate and
test a
full rated (55 kW)
reconfigurable dc
-
dc converter
architecture.
Share results with APEEM team members.

20

Summary



Relevance:
This project is targeted toward active energy management and reduced size and
cost of the power electronic converters that interface RESS and traction drive inverter.


Approach:



Develop a bi
-
directional buck/boost dc
-
dc converter between the regenerative energy
storage systems (RESS) and the dc link (traction drive inverter),


Design
a hybrid battery/ultra
-
capacitor energy storage system architecture,


Design
a modular reconfigurable
dc
-
dc
converter architecture,


Collaborations:

Collaborations with Chrysler, Maxwell, IOXUS, and ORNL’s Energy Storage
Programs and Battery Manufacturing Facility are being used to maximize the impact of this
work.


Technical Accomplishments:


Reviewed
and modeled bi
-
directional dc
-
dc
converters.


Reviewed
and modeled hybrid RESS architectures and created a summary report based
on the advantages, drawbacks, control systems, performance, number of parts,
etc.


Selected
and modeled four battery/UC hybridization strategies
. Run the simulations and
compared the performance results.


21

Reviewer
-
Only Slides


22

Responses to Previous Year
Reviewers’
Comments



Recommendation/Comment
:


Response/Action
:



New Start for FY13, no previous comments



23

Publications and Presentations


N/A at the time of presentation submission



24

Critical Assumptions and Issues


Current traction drive architectures are variable voltage
(325
±
125V), resulting in lower power electronics efficiency.


Optimal battery and ultra
-
capacitor sizing.


Boost
ratio/range and efficiency that can drive
dc
-
dc
converter architectural
choices.


Potential cost addition due to choice of
dc
-
dc
converter
and hybrid energy storage
systems.