Control of Powertrain Systems at the High Efficiency Limit Anna G. - - PDF document

Powertrain ! Control!

(1)/46

Anna G. Stefanopoulou, University of Michigan annastef@umich.edu National Science Foundation US Department of Energy US Army with Ford, Bosch, GE, A123/Navitas, & Daimler Thanks to the

Control of Powertrain Systems at the High Efficiency Limit

American n Co Cont ntrol Co Conf nferenc nce ,June une 2014

Powertrain ! Control!

(2)/46

Outline

• - Chaotic Engines
• - Stressed-out Batteries
• - Dead-ended Fuel Cells

Powertrain ! Control!

(3)/46

Passenger 15%

Commercial 5%

Transportation 20%

Electric Power 22% Residential 3% Commercial 2% Industrial 10%

Global Temperature Change

Carbon Dioxide (fossil fuel use) 57%

Carbon Dioxide (other) 20% Methane, NO, F-gases 23% Powertrain ! Control!

(4)/46

A Glance around the Globe

Powertrain ! Control!

(5)/46

History Lessons (US-focused)

Powertrain ! Control!

(6)/46

Slow ..? No .., just better! Slender … ?

Powertrain ! Control!

(7)/46

Enabling Act..uat..ors Behind-the-Scenes And some of the real Actors

Port Fuel Injection Multi-Valve Variable Valve Carburetor Direct Injection Port Fuel Injection Variable Valve Timing

gal/100mi/HP

Anna’s PhD Mrdjan’s diVCT Jessy’s Air Charge Estimator Ilya @ Ford

Turbocharging Downsizing

Powertrain ! Control!

(8)/46

Gasoline Engine Efficiency

A A Midsize US US Car: 3800 lb lbs 1700 1700 kgs kgs

Km/h sec

• - Ford, with its Ecoboost

turbocharging technology, makes a small engine act big.

• - GM, with its Active Fuel

Management cylinder deactivation system, makes a

big engine act small.

• R. Truett, Auto News, Jan 6, 2014

1000 1500 2000 2500 3000 3500 50 100 150 200 250 300 350

Engine Speed [RPM] FTP−75 Engine Load [Nm]

Time / 1877s [%] 0.0001 0.001 0.01 0.1 1 10

V6, 3.6L

Sweet spot

Powertrain ! Control!

(9)/46

Efficiency Improvement: Downsizing

3500

−5 −15 −20 − 1 − 5 −25 − 5

Engine Speed [RPM]

− 2 5 − 2 5 − 5 − 5 5 − 1 −5 10 −5 − 1 −20 − 1 5 −15

FTP−75 Engine Load [Nm]

1000 1500 2000 2500 3000 3500 50 100 150 200 250 300 350

20% 10% 5% Efficiency Improvements % Worst Controlling the Dynamics Original 3.6L V6 Turbocharged 2.0L I4

• 5. Cnv
• 4. v/eTC
• 3. thr/wg
• 2. TC-Dnsz
• 1. Dnsz

Powertrain ! Control!

(10)/46

Efficiency Improvement: Downsizing

Efficiency Improvements % Worst Controlling the Dynamics Original 3.6L V6 Turbocharged 2.0L I4

• 5. Cnv
• 4. v/eTC
• 3. thr/wg
• 2. TC-Dnsz
• 1. Dnsz

20% 10% 5% Worst

Powertrain ! Control! 12V BAS Micro-HEV **

Cost effectiveness

TRBDS--2 TRBDS--1 GDI VVL VVT FR PHEV MPG DIESEL HEV

PHEV MPGe EV (estimated)

Powertrain ! Control! 12V BAS Micro-HEV **

Cost effectiveness

TRBDS--2 TRBDS--1 GDI VVL VVT FR PHEV MPG DIESEL HEV

PHEV MPGe EV (estimated)

?

Powertrain ! Control!

(13)/46

Hot Flame Region: NOx Hot Flame Region: NOx&Soot

Gasoline Diesel versus

Spark Ignition (SI) Compression Ignition (CI)

Gasoline Diesel

Stoichiometric Lean

TWC SCR+ …$ ...$ Exhaust After- Treatment

Powertrain ! Control!

(14)/46

Gasoline Diesel HCCI

SI HCCI CI Peak Temperature (K) >2000 1600 1800 NOx emission High Low Medium Combustion Duration (CAD) 40 2-10 40

Car Makers Seek New Spark In Gas Engines The Wall Street Journal 09/28/04 … engineers call homogenous-charge compression-ignition, or HCCI and expected to provide 80% of the efficiency of a hybrid or a diesel for 20% of the cost, …

Powertrain ! Control!

Gasoline Systems - HCCI

Actuators, Sensors, & Performance Objective

Variable Valve Injector

In-cylinder pressure sensor

Controller θ50

50 ref"

θ50

50"

Combustion phasing controlled through trapped dilution

Powertrain ! Control!

(16)/46

HCCI Model

Combustion

Homogeneous Charge Chemical Kinetics= Arrhenius Integral

Powertrain ! Control!

(17)/46

Stable, Unstable, and Limit Cycle Behavior

Automotive Engineering SAE 2002-01-0111– Lund Regions with Stable and Unstable operation ASME ICE 2000– Caterpillar Limit cycle behavior SAE 892068– Southwest Research Institute Very Stable and Unstable behavior at different regions

Stability in auto-thermal reactors

Heerden 1953, Liljenroth 1918 Combustion Intake Temperature, Tivc (K) Blow-Down Temperature, Tbd (K)

Early Combustion Phasing

Chiang CDC 2004 & TCST 2004

Stable Unstable Limit Cycle

Powertrain ! Control!

(18)/46

Drive around Stable Points!

Intake Temperature, Tivc (K) BlowDown Temperature, Tbd (K)

Early Combustion Phasing

Chiang CDC 2004 & Chiang, IEEE-TCST 2004 Clean or Efficient? An Engine Goes for ‘Both of the Above’

By LINDSAY BROOKE August 19, 2007

SAE-2009-01-1131

Combustion

Powertrain ! Control!

(19)/46

CDC 2004 SAE 2009 NYT 2007

Gerdes’s team @ Stanford Switching gains (2011) Anna’s team @ UMICH Nonlinear Cntr Lyapunov functions (2006) Tunestal’s team @ Lund Nonlinear MPC (2009) @Alberta, CA @MSU @Cambridge, UK @Chalmers, Se @Univ. of Minnesota

Controlling Stable, Unstable, and Limit Cycle Behavior

Powertrain ! Control!

(20)/46

Heat Release Analysis

Observations from the high variability points

Powertrain ! Control!

(21)/46

Detailed Heat Release Observations Key factors for describing CV Nonlinear coupling between ! the recycled thermal energy ! the recycled chemical energy in the unburned fuel

Powertrain ! Control!

(22)/46

Model that captures the global behavior

! Period doubling bifurcations ! Thermal runaway ! Noisy simulations match the data

Powertrain ! Control!

Controlling Combustion at its Limit

Injection Timing (usoi) Combustion Phasing (θ50) Powertrain ! Control!

(24)/46

HCCI Control Toolbox

~Torque Phasing

Avoided Misfire Moderate Transient

Powertrain ! Control!

HCCI Control Toolbox

Reduced Ringing Did not slow down

~Torque Phasing

Powertrain ! Control!

(26)/46

Mode Transitions

Significant number of mode transitions during driving cycle!

Powertrain ! Control!

(27)/46

To Jump … or not to Jump

Powertrain ! Control! PHEV MPG * DIESEL

Data Sources: 1. Assessment of Fuel Economy Technologies for Light-Duty Vehicles (2011) National Research Council 2. * www.fueleconomy.gov DOE & EPA website (MPGe : 1 Gallon of Gasoline = 33.7 kWh) 3. **MPG baseline 2008 midsize cars. NHTSA stats (2014)

HEV 12V BAS Micro-HEV **

Cost effectiveness of engine technologies

PHEV MPGe * EV (estimated)

HCCI

TRBDS--2 TRBDS--1 GDI VVL VVT FR

2025 Target

Powertrain ! Control!

(29)/46

Fuel Price, $ per gallon Battery Prices, $/kWh

*8 years payback 2012 posting in http://www.washingtonpost.com/

2014

Tesla R 04/08

The Rational Business Perspective

Powertrain ! Control!

(30)/46

from Dr. Ilan Gur, ARPA-E Program Director

Removing the Blinders Model-Based Estimation

What we Control

Current Coolant Flow and Temp

Powertrain ! Control!

(31)/46

The Operating Principle

Graphics from K.Smith, CSM 2010

6C+LiCoO2"LixC6+Li1-xCoO2

+Heat

+Swelling

LiCoO2 LiCoO2

Powertrain ! Control!

(32)/46

Models for Electrical State Estimation

Fidelity Computation time Single-Particle Model Electro-Chemical Coupled Diffusion-Reaction

Fuller et al., 1994; Ramadass et al., 2003; Fathy & Moura, 2011 … Ning and Popov, 2004; Subramanian et al., 2005; Di Domenico et al., 2010 CY Wang, 2014 …

Parameter identification is difficult

Current collector Current collector Negative electrode (-) Positive electrode (+) Separator

C1,n C1,p r r C12,n C12,p Li+

e- e-

V I

x= 0 x= Ln x= Ln+Ls x= Ln+Ls+Lp Current collector Current collector Negative electrode (-) Positive electrode (+) Separator C1,n C1,p r r C12,n C12,p Li+

e- e-

V I

x x= 0 x= Ln x= Ln+Ls x= Ln+Ls+Lp

Equivalent-Circuit Model

Yurkovich, 2009 Perez et al., 2012; Prasad and Rahn, 2012; Hu et al., 2012

Powertrain ! Control!

(33)/46

Models for Thermal State Estimation

Fidelity Computation time

Gu and Wang, 2000; Kumaresan et al., 2008; Lee et al., 2010; Fleckenstein et al. 2011; Muratori et al., 2010; Muratori et al., 2012

Distributed Parameter

• 0.015
• 0.01
• 0.005

• 0.015
• 0.01
• 0.005

Lumped Parameter Equivalent-Circuit

Mahamud and Park, 2011; Park and Jaura, 2003; Forgez et al., 2010; Lin et al., 2013

Reduced-Order

Powertrain ! Control!

(34)/46

Fidelity Reduced-Order

Gu and Wang, 2000; Kumaresan et al., 2008; Lee et al., 2010; Fleckenstein et al. 2011; Muratori et al., 2010; Muratori et al., 2012

Distributed Parameter

• 0.015
• 0.01
• 0.005

• 0.015
• 0.01
• 0.005

Lumped Parameter Equivalent-Circuit

Mahamud and Park, 2011; Park and Jaura, 2003; Forgez et al., 2010; Lin et al., 2013

Pack Level

Models for Thermal State Estimation

Powertrain ! Control!

(35)/46

Powering at the Operational Limits

T1 T4 T3 T2 Powertrain ! Control!

(36)/46

Powering at the Operational Limits

T1 T4 T3 T2

Powertrain ! Control!

(37)/46

Neutron Imaging: Lithium Concentration & Expansion

Nationa nal Ins nstitut ute of Stand ndards and nd Techno hnology (NIST) T)

24 MW Reactor

Reactor Core

Powertrain ! Control!

(38)/46

Reactor Core Dark Areas= High Lithium Concentration

Neutron Imaging: Lithium Concentration & Expansion

Powertrain ! Control!

(39)/46

Neutron Imaging: Lithium Concentration & Expansion

Powertrain ! Control!

(40)/46

Sensor Multi-Sensing with Adaptive Models for better estimation

• f State of Charge (SOC), State of Power (SOP), State of Health (SOH)

and bigger operating battery window in HEVs

Battery Multi-Sensing, Estimation, and Controls

Powertrain ! Control!

(41)/46

5 10 15 20 25 30 10 15 20 25 30 35 40 45 50 55

t((min) T(( oC)

Sensor Multi-Sensing with Adaptive Models for better estimation

• f State of Charge (SOC), State of Power (SOP), State of Health (SOH)

and bigger operating battery window in HEVs

Battery Multi-Sensing, Estimation, and Controls

Thin Film Sensor Array Collocated Temp & Strain Powertrain ! Control!

(42)/46

PEM Fuel Cell

Phosphoric Acid Fuel Cell Solid Oxide Fuel Cell

Ragone Plot (Energy versus Power)

Powertrain ! Control!

(43)/46

A Dead-Ended Fuel Cell

Purges no flow control

Fuel Cell Challenge: Control the Water

Powertrain ! Control!

(44)/46

Energy-Lean Information-Rich Transportation

Electrical Grid Gas Grid Transportation Grid

National Highway System

Information Grid

Powertrain ! Control!

(45)/46

Per Capita Energy Use

Powertrain ! Control!

Thanks to

Jason Siegel, Erik Hellstrom Xinfan Lin, YoungKi Kim, Shankar Mohan, Boyun Wang Shyam Jade, Jacob Larimore, Sandro Nuench, Pat Gorzelic, Mike Hand Dan Hussey, David Jacobson (NIST), David Gorsich, Yi Ding, Matt Castanier (TARDEC) Danny King, Patrick Hagans, Maha Hammoud, Mike Wixon (A123) Dyche Anderson, Mrdjan Jankovic, Jim Waldecker (Ford) Li Jiang, Hakan Yilmaz (Bosch), Aaron Knobloch (GE) William Lim, Kathie Wolney, Charlotte Bowens (ARC) ….

Denise McKay Vasilis Tsourapas Jixin Chen Ardalan Vahidi

• J. Chiang

Buz McKain Amey Karnik Aswin Shalvi Laura Olesky … Jeff Cook Jing Sun Jessy Grizzle Jim Freudenberg Ilya Kolmanovsky Galip Ulsoy Dawn Tilbury Huei Peng Tulga Ersal Jason Martz … and Lino Guzzella (ETH) Zoran Filipi (Clemson) …