Training Part 1 Overview

Sponsored by Lee SlezakSponsored by Lee Slezak

“This presentation does not contain any proprietary or confidential information”

PSAT Used to Support R&D and Management Decisions After a thorough assessment, PSAT was selected in 2004 as the primary 

vehicle model for all FreedomCAR and 21 CTP activities by the U.S.DOE, stating that “All future code development and enhancements for OFCVT

Management Decisions

stating that  All future code development and enhancements for OFCVT shall focus on PSAT and PSAT‐PRO”

PSAT was awarded a R&D100 Award in 2004 and a Technology Transfer Award in 2007Award in 2007

PSAT is currently used by more than 130 companies and 700 users worldwide, including GM, Ford, Chrysler, Hyundai, Toyota

2

Reliance on Modeling & Simulation is Continuously Increasing Reduce cost and time to production

Provides math‐based environment for more thorough multidisciplinary 

Continuously Increasing

integration and testing in the virtual environment before hardware builds

Sorts technologies quickly to reduce hardware build iterations

Promotes parallel and integrated virtual development of control systems p g p yand hardware

Reduces/eliminates duplicate modeling and analysis work and activities

Enables fast to market with new technologies and real fuel Enables fast to market with new technologies and real fuel economy Delivers better‐integrated, initial designs that balance 

( )Fuel Economy, Emissions and Drivability (FEED) requirements.

Provides common methods and tools for comparing/evaluating technologies.

Facilitates efficient, seamless link from research to production to maximize reuse of work products and eliminate waste.

3

AUTONOMIE –

Taking it to the Next Level

The objective is to accelerate the development and j pintroduction of advanced technologies through a Plug&Play architecture that will be adopted by the 

entire industry and research community. 

4

Accelerating the Development and I t d ti f Ad d T h l iIntroduction of Advanced Technologies

Why is Autonomie Unique?Why is Autonomie Unique?

Develop/Evaluate Component/Control Technologies

Develop/Evaluate Vehicle ArchitecturesDevelop/Evaluate Vehicle Architectures

5

AUTONOMIE/PSAT Comparison p

Capability PSAT PSAT‐PRO AutonomieArchitecture

Plug & Play Architecture

Hierarchical Architecture Standards  (Vehicle, syst…)

Model Reusability through System Experts (Concept to Production)Establish Standard Interfaces (Industry‐wide)

FeaturesCapability PSAT PSAT‐PRO Autonomie

Model/data Customization

Features

Powertrain Configuration Customization 

Select Appropriate Level of Modeling

GUI Customization (process, post‐processing…) 

6

(p , p p g )

Database Management

AUTONOMIE/PSAT Comparison p

Capability PSAT PSAT‐PRO AutonomieUsage*

Capability PSAT PSAT PRO AutonomieEvaluate Fuel Consumption Benefits(technology, size, powertrain configuration…) Evaluate and Balance FEED in Simulation(Fuel Economy, Emissions & Drivability)Develop Component Requirements

Simulate Single Component

Develop System/Subsystem Requirements

Develop Vehicle Level Control

Develop System/Subsystem/Component ControlDevelop System/Subsystem/Component Control 

Component‐in‐the‐Loop

Software‐in‐the‐Loop, Hardware‐in‐the‐Loop…

7

*Final usage depends on the level of details of the models available

Develop a Software Architecture & Environment to Plug-and-Play Math-Based Models of Hardware & Control System Software Early or Up Front in the Design & Analysis ProcessSoftware Early or Up-Front in the Design & Analysis Process

Maximum  User AccessDatabase

Maximum Flexibility

Reusability

EnterpriseWid S l ti

ControlDatabase

Flexibility 

Selectable Complexity

Wide Solution

VersionControl

Interface

Complexity

CodeNeutrality

Control

DatabaseSearch

Models, Data

DatabaseModels, Data

y Search

Graphical User Interface

Linkage withOther Tools

ResultsVisualization

GenericProcesses

SetupSimulation

8

Other ToolsVisualizationProcessesSimulation

Model & Data Requirements

■ Automated integration of existing models / controls / data ■ All models for a specific area of expertise in a single location■ Systems duplicated using Matlab API

MaximumReusability

Maximum■ Any system can be built automatically■ User can add their own configurations

Flexibility  ■ Single components or entire vehicles can be simulated

■ Common nomenclature (i e naming I/O )Selectable Complexity

■ Common nomenclature (i.e., naming, I/O…)■ Common model organization (i.e, CAPS)■Model compatibility checked

CodeNeutrality

■Matlab / Simulink main environment■ Use S‐functions■ Co‐simulation (i.e., CoSimate*)

9

( , )

* Not yet implemented

Graphical User Interface Requirements

SetupSimulation

■ Select architecture, model and data ■ Check compatibilities ■ Select simulation outputs to Workspace■ Select simulation type (i e fuel efficiency performance )

Generic 

■ Select simulation type (i.e., fuel efficiency, performance…)

■ Calibration, Validation, Tuning■ Parametric study, including Monte Carlo analysis

Processes  ■ Optimization algorithms■ Predefined set of simulations & report (i.e., Drive Quality)

■ Predefined calculations (i e fuel economy efficiency power )Results 

Visualization

■ Predefined calculations (i.e., fuel economy, efficiency, power…)■ Predefined plots, both time based and lookup tables■ Energy balance■ Specific plots available for different models (users defined)

Linkage withOther Tools

■ Co‐simulation (i.e., CoSimate*)■ Specialty Tools (i.e, GT‐Power, CarSim, AMESIM, AVL Drive*…)■ Database Management (i.e., SourceSafe…)

10

g■Well‐to‐Wheel (i.e., GREET*)

* Not yet implemented

Database Requirements

User AccessControl

■ Prevent unauthorized users from accessing restricted or proprietary data■ Allow authorized users to download all necessary files■ Ensure model is documented before integration into database

Enterprise 

■ Ensure model is documented before integration into database

■ Allow users to collaborate (i.e. share models)■Main database accessible anywhere

Wide Solution  ■ Consistent process for interacting with files

■Maintain traceability of all changesVersion Control

■Maintain traceability of all changes■ Keep linked files together through entire vehicle process (i.e. design, simulation and test)

DatabaseSearch

■ Use keywords to search data, models, controls related to    specific projects■ Quickly find the correct model with the correct fidelity of 

11

Q y ymodeling and all related files

Key Benefits

Plug&Play

■ Flexibility & Reusability■ Customizable architectures■ Common Nomenclature■ Code Neutral

R d■ Common Methods to sort technologies quickly t d h d b ild it tiReduces 

Cost & Time to 

Production

to reduce hardware build iterations■ Reduces/eliminates duplicate modeling and analysis work■ Delivers designs that balance Fuel Economy, E i i d D i bilit (FEED) i t

Enterprise

Emissions and Drivability (FEED) requirements

Enterprise Wide 

Solution

■ Database Management■ Provides common methods and tools for comparing/evaluating technologies

12

No other tool currently allows the linkage with any legacy code!

Unique Feature – Plug & PlayImplement any language

Legacy Code

Implement any languageAutomated process to import legacy code (data, model, control, process)…

Calibration CarSim*

, p )

Plug & Play

Specialty Software (COTS)

Legacy

Processes

CalibrationValidationTuningDrive Quality…

CarSimGTPower*Amesim*AVL Drive…Drive Quality…

Database Management

Version ControlDatabase Search

13* Already linked

No other tool currently builds the model automatically!

Key New Feature – Software CustomizationModel organizationModel organizationSingle or multiple plantsController location

Model Configuration

Fuel economyModelsDataControls…

Fuel economyValidationDrive qualityControl…

Customization ProcessesProprietary Information

CalculationsPlotsR t

Post‐processing

14

Reports…

Integrate Legacy Code

Example: Integrate Legacy Simulink ModelsLegacy Code

GUI Automatically Renames Variables, Creates Necessary Support FilesCreates Necessary Support Files

Ready for use in

15

Model Plug&Play Capability Provided by “W ” “Wrapper”

Legacy Plant

First block automatically built to select the required input parameters change

Last block automatically built to change units (SI) and data type before sending

16

the required input parameters, change units and data type

units (SI) and data type before sending the information to a bus to make them available to other systems

Experts Develop Systems1 Wh t i S t i A t i T i l ?1- What is a System in Autonomie Terminology?

Expert System

Initialization 

DocumentationExpert System

Model

File

Pre‐processingFile

Test Data

File

Post‐processingFile

ReportFile (HTML)

17

Experts Develop Systems2 E h S t C fi ti C B C t i d2- Each System Configuration Can Be Customized

Each System is OptionalEach System is Optional

ActuatorController Plant Sensor

Example: GM 2 Mode HEVTransmission Plant

Electric Machine #1

Any System can have Subsystems To Accurately Represent 

Transmission Plant

Electric Machine #2

Gearbox

18

y pHardware

Experts Systems Can Be Reused3 H C W R it?3- How Can We Reuse it?

Expert #9

Vehicle

Expert #1

Expert #3

BatteryExpert #7

Expert #2

Controller

Expert #7

ChassisPlant

Expert #8

Expert #4 Expert #5

Expert #6

EngineDriveline

Expert #4

Controller

Expert #5

Plant

19

Autonomie Designed to Be Used For All Steps in the Development ProcessDevelopment Process

Easy selection & implementation of data, models, control or cycles

Run batch mode +Distributed

Build and compare large number of of data, models, control or cycles Distributed

computinglarge number of technology, powertrain, options

Ensure simulation t bilit

Analyze and compare test and

traceability, model compatibilities

psimulation data

Database M t

Enables MIL, SIL,

Generic ProcessesManagement RCP, HIL, CIL

20

Accelerating the Development and I t d ti f Ad d T h l iIntroduction of Advanced Technologies

Why is Autonomie Unique?Why is Autonomie Unique?

Develop/Evaluate Component/Control Technologies

Develop/Evaluate Vehicle ArchitecturesDevelop/Evaluate Vehicle Architectures

21

Develop Plant Models with Different Level of C l itComplexity

Steady State Model

Physical Model

Mean Efficiency Model

Highly Dynamic Model with Production Code

Different models are necessary to study different phenomena

Detailed models are required when technology

22

Detailed models are required when technology cannot be tested or does not exist!

Link with Commercial Off the Shelf Tools (COTS)#1 – Develop Model in Native Environment

#2 – Integrate Model in Simulinkin Simulink

#3 – Use Model in Autonomie

COTS  (1) can be used to generate maps(2) can be run in the Simulink environment(3) can be run in their own environment (co‐simulation)

23GT Power, CarSim have already been linked

(3) can be run in their own environment (co simulation)

Select the Appropriate Level of Modelingllers

Control or ororor

ts

oror

Plan

oror

oror

or

or

EngineExperts

TransmissionExperts

ChassisExperts

Battery, MotorExperts

24

Experts Experts Experts

Vehicle Experts

Experts

Main Steps of Model Based Design

Hardware‐in‐the‐LoopSoftware‐in‐the‐Loop

Rapid Control Prototyping Component‐in‐the‐LoopRapid Control Prototyping Component in the Loop

25

Develop Control Algorithm Software in the LoopSoftware-in-the-Loop

Specific configuration defined to develop and test new control algorithmsnew control algorithms

New algorithm(s) to be testedProduction Code

Real Time Operating System (RTOS) ensures call of functions at specific 

intervals (such as CAN)intervals (such as CAN)

26

Hardware input/outputSends and receives CAN signals

Test Low Level Control AlgorithmHardware in the Loop and Software in the Loop

Automatic building save days of development

Hardware-in-the-Loop and Software-in-the-LoopHardware‐in‐the‐Loop

(Control Hardware / Emulated Plant)Automatic building save days of development

All the lines connecting the blocks are built automatically

Block built automatically‐Change unitsChange data type‐Change data type‐Extract all the proper sensors

27

Evaluate Fuel Consumption BenefitsModel in the LoopModel-in-the-Loop

Example “Evaluation of HCCI Engine Fuel Savings for Various Powertrain”E i f U fMEngine maps from UofM

Vehicle & ControlDevelopment

1.000

Ratio Fuel Consumption of  HCCI vs Regular Engine 

Fuel Consumption Analysis

p

0.850

0.900

0.950

Unad

justed

 Fuel Con

sumption Ra

tio

Conventional

Micro_HEV

Mild_HEV

Split_HEV

Split_PHEV10

Split_PHEV20

Series_PHEV30

Series PHEV40

Detailed Analysis of Reasons Behind Benefits

0 950

1.000

HCCI Fuel Savings vs PI According to Hybridization Degree

0.750

0.800

1

U Series_PHEV40

0.800

0.850

0.900

0.950

Una

djusted Fuel Con

sumption Ra

tio 

UDDS

HWFET

Combined

0.900

0.950

1.000

1.050

on Ratio

Fuel Consumption Ratio (High Case) of HCCI vs PI

28

0.750

0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 70.00%

Hybridization Degree (%)

0.600

0.650

0.700

0.750

0.800

0.850

0.900

0.00% 20.00% 40.00% 60.00% 80.00% 100.00% 120.00%

Una

djusted Fu

el Con

sumptio

Percent of Time Engine is ON

Evaluate Non Modeled Phenomena With HardwareComponent-in-the-LoopComponent-in-the-Loop

Example#1: Impact of battery cold start on PHEVs Fuel Consumption

Sensors Battery behaves as if in vehicle

Example #3: Engine and Battery are Coupled

Example #2: Impact of emission and engine

Rest of the Vehicle Modeled

y p

emission and engine cold start on PHEVs Fuel Consumption

29

Engine behaves as if in vehicle

Evaluate Non Modeled Phenomena With HardwareComponent-in-the-Loop - ExamplesComponent-in-the-Loop - Examples

Impact of Vehicle Level Control on Battery Life

vehicle speedl l t d CO2 /

Impact of Vehicle Level Control on Engine Emissions

10

15

20

25

O2

gram

s/se

cond

calculated CO2 g/secengine torque/10 measured

200 250 300 350 400

0

5

Time in seconds

CO

Impact of Charger Rating on Battery Roundtrip Efficiency

2, 95.151.4, 95.56, 21.5

94

95

96

cien

cy

20

25

rent

-

Impact of Temperature on All Electric Range 

8 0

9 0

1 0 0

eed

I n itia l te m p e rature 2 0 d e g re e s C ve hic le s p e e d in m e te rs p e r s e c o ndInitia l te m p e rature 0 d e g re e s CInitia l te m p e rarture -7 d e g re e s C

6 89 11

2, 71.4, 5

89

90

91

92

93

94te

ry ro

undt

rip e

ffic

(%)

5

10

15

tter

y ch

argi

ng c

urR

MS

(A)

4 0

5 0

6 0

7 0

Sta

te o

f Cha

rge,

Veh

icle

spe

30

6, 89.11

88

89

0 2 4 6 8

Charger Power (kW)

Bat

t

0

Ba

0 5 0 0 1 0 0 0 1 5 0 0 2 0 0 0 2 5 0 0 3 0 0 0 3 5 0 02 0

3 0

T im e in s e c o nd s

S

Define Component Requirementsen

t Data 

D Teams

Testing

Battery AccessoriesCamry A/C Power

Prius A/C Power

Electric Machine2004 Prius

Compo

nefrom

 R&D

IL Ro

Rp1 Rp2

OCVVLModel

Prius A/C Power

hicle

ations

Ip1 Ip2

Vehicle Classes

Vehicle  DOE / USABC

Sizing & Simulation

Veh

Simula

Requirements DOE / USABC Requirements

Validation Battery RCP Vehicle Testing

31

V

JCS VL41M

Accelerating the Development and I t d ti f Ad d T h l iIntroduction of Advanced Technologies

Why is Autonomie Unique?Why is Autonomie Unique?

Develop/Evaluate Component/Control Technologies

Develop/Evaluate Vehicle ArchitecturesDevelop/Evaluate Vehicle Architectures

32

Validate Vehicle Models to Ensure Studies I t itIntegrity

Analyze Data Develop Control

Collect Test Data

Engine Torque

Validate Model

d d b d l b

APRF

Understand Gearbox Model Gearbox

Source : GM

33Example of GM 2 Mode Tahoe

Develop Vehicle Control to Maximize Fuel DisplacementGl b l O ti i ti

tSOCtP

tJtSOCtP '0,000 ,1',1

tU 0 Optimal Control

Teng Tmc

ωmc

Global Optimization

tSOCtP nm ,

tJtSOCtP MMMM ',' ,1,1

0

tU M

TX 0X.

.

.

.

.

.

Optimal Control, Minimal Fuel Cons.

Backward model Bellman Principle

ωeng

Backward model p

Various Control St t iVarious 

Control Design

StrategiesControl Principles Rule‐Based Control Design

Heuristic Optimizationp

Optimally tunedParameters

34

Existing Control Logic DIRECT Algorithm

Develop Vehicle Control Taking into Account N M d l d P tNon Modeled ParametersExample: Impact of Vehicle Level Control on Engine Emissions for PHEVs(collaboration with ORNL)(collaboration with ORNL)Step #1 – Evaluate Different Controls in Simulation

Vehicle Model Detailed After‐Treatment ModelEvaluation ofEvaluation of Different Vehicle Control Logic and Tuning

Step #2 – Verify the Trends with Engine‐in‐the‐LoopAnalyze Results

35

Rest of the Vehicle Modeled Engine behaves 

as if in vehicle

Match Technologies, Configurations to S ifi A li tiSpecific Applications

15Comparison of Aerodynamic Fuel Savings for Drive Cycles vs Steady States

Impact of Aerodynamics for Different Line Haul Applications

30.0

Impact of Mild and Full HEV for Line Haul Applications

5

10

15

mpr

ovem

ent f

rom

hig

h to

low C

d

Drive CyclesSteady States

16.0

13.716.3

24.022.4

15.513.1

15.6

21.319.2

15.112.5

15.0 14.015.5

10.0

15.0

20.0

25.0

sumption (gal/100

mi) CONV MILD‐HEV FULL‐HEV

50% load

0

5

Per

cent

age

Im

HTUF Class 6/

21 mph SSUDDS Truck/26 mph SS

HHDDT Cruise/42 mph SS

HHDDT High Speed/53 mph SS

0.0

5.0

10.0

HHDDT65  HHDDT Cruise HHDDT High Speed

HHDDT Transient

udds_truck

Fuel Con

s

h S f h C bi d h l i h S f h h l

16.0%

18.0%

20.0%

d

Impact of All Technologies on Fuel Consumption

Engine

Transmission

The Sum of the Combined Technologies < The Sum of Each Technology 

Class 2B Pickup

14.9% 15.9%

1.2%1.6%1.7%1.4%

8.6%8.9%

4 0%

6.0%

8.0%

10.0%

12.0%

14.0%

Percen

t Fue

l Saved Transmission

RR

Cd

Weight

Hybrid

36

1.4% 1.4%

3.0% 3.8%

0.0%

2.0%

4.0%

Each technology (Conv)

Combination(Conv)

Each technology(Hybrid)

Combination(Hybrid)

Hybrid  improvementsBaseline improvements

Evaluate Uncertainties Through Monte Carlo A l iAnalysis■ Uncertainty is modeled by a probability density function (pdf)■ How is the uncertainty propagated? 

■ PHEV 10 miles All Electric Range (AER) midsize used as reference case■ PHEV 10 miles All Electric Range (AER) midsize used as reference case

Inputs Sampling Results

CdMonte Carlo (MC), Latin hypercube (LHS),Median Latin hypercube (MLHS)Quasi Monte‐Carlo

FA

Crr

Weight

37

A f i d d d f k

SummaryA software environment and standard framework

Unifies the entire engineering organization for efficient operation

Focuses analysis on total system engineering and integration

Provides complete user customization by an open architecture

Simulates from single components, subsystems to entire vehicles

Manages models, data, processes, results and control code from research to production by configuration and database management 

38