111

Bo Chen, Ph.D.Dave House Associate Professor of Mechanical Engineering and

Electrical EngineeringDepartment of Mechanical Engineering - Engineering Mechanics

Department of Electrical and Computer EngineeringMichigan Technological University

July 4th, 2013

Model-Based Design and Hardware-in-the-Loop Simulation

for Clean Vehicles

1

Outline of Presentation

Background and Motivation Model-Based Approach for Embedded Control

System Design Hardware-In-the-Loop (HIL) Simulation Single Shaft Parallel Hybrid Electric Vehicle (HEV)

and Electronic Control Unit (ECU) Network HIL Simulation of Parallel HEV Conclusions

2

Background and MotivationVehicle Population• 750, 000, 000+ vehicles in the world• Expected 2 billions by the year of 2020

Fuel Consumption• The transportation sector in the U.S.

accounts for two-thirds of our petroleum use

Environmental Emission• Transportation produces 33% of US CO2

output

Technologies & Challenges• Hybrid, electric, fuel cell , and renewable energy

vehicles• Complex powertrain and control system

Fuel Economy& Emission Standards

3

Model-Based Approach for Embedded System Design

Conventional V-model: a plan-driven process. The design process follow the defined development stages in order

Model-based Design: provide an integrated environment for design, simulation, automatic code generation, and validation.

4

Salient Features of Model-Based Embedded System Design Approach

Platform for representing entire system: control strategy and plant models Graphical representation: use graphical language to

describe implementation details Time saving: minimize software development time and

maximize software re-use. No hand coding, production quality code is automatically generated Integrated development and validation cycles Communication among the team members is made easier

5

Hardware-In-the-Loop (HIL) Simulation

ActuatorsSensors

Automotive Plant Model

Signal Conditioning

Output Drivers

Micro-controller

Electronic Control Unit

Real World

dSPACE HIL Simulator

MotoTronEngine ECU

MicroAutoBoxMotor ECU

Electrical Signal Interface

HIL System

• Generate an environment where ECU assumes that it is running with a real physical system.

• HIL simulator simulates the physical system that is under test. It generates plant sensor signals and capture actuator signals from ECU.

• Used to test control strategies to be implemented on ECUs.

6

dSPACE HIL I/O Interface

PHS-Peripheral High Speed

7

Model Development and HIL Simulation Process

Model Development Parameterization Updating

ModelAuto-code Generation

Downloading to ECUs

Online CalibrationRecording

Colors indicate the different softwares used in the steps. Blue : Simulink/Stateflow/dSPACE blocksets/Real-time Workshop

Model building, modification, and auto-code generation Yellow : dSPACE ModelDesk

Model Parameterization Red : dSPACE ControlDesk Next Generation

Real Time Calibration, Data Recording, File Export

8

Single Shaft Parallel HEV PowertrainArchitecture and ECU Network

ICE PMSM

Battery

Inverter

Gearbox

Final Drive

Clutch1 Clutch2

Automatic Transmission

Motor ECU Transmission ECU

Hybrid ECU

Engine ECU

Driver Environment• Vehicle level

ECU:

Hybrid ECU

• Low level ECU:

Engine ECUMotor ECUTransmission ECU

9

dSPACE Parallel HEV Model

Actu

ator

sig

nals

from

EC

U

Sens

or s

igna

ls fr

om E

CU

From RTI Blocks

To RTI Blocks

dSPACE HIL Simulator

Sensor SignalsActuator Signals

From MotoTron ECM:• Ignition angle • Injection angle and

duration• Throttle valve, EGR

valve, pressure control valve position, and PWM control signal

To MotoTron ECM:• Engine torque request• Engine speed• Engine keys• Intake manifold

pressure and temperature

• Coolant temperature• Rail pressure• ……

From MicroAutoBox II:• Three phase PWM

signals

To MicroAutoBox II:• Motor Torque

Request• Motor Speed• Three phase

current• DC bus voltage

10

Parameterize HEV Model Using ModelDesk

Environment

Vehicle DynamicsDrivetrainEngine

11

HIL Setup for the Parallel HEV

dSPACE HIL simulator MicroAutoBox II for motor controller MotoTron 128 pin ECM for engine

controller

12

Driver Controller

Vehicle Plant

Hybrid ECU

MicroAutoBox II

Motor Controller

HIL Simulator

APP

BPP

ICE Torque Request, Key

EngineActuator Signals

SimulatedEngineSensor Signals

Vehicle Key

PWM Signals

Vehicle Reference Speed Actual Vehicle Speed

Motor Torque Request, Motor Speed, 3 phase current, DC bus voltage

PWM Signals

Motor Speed, 3 phase current, DC bus voltage

Motor Torque Request

SOC, Transmission Speed, Current Gear Level, Engine State

Transmission ECU

Gear Level, Clutch Command

Mechanical Brake Torque Request

Gear Ratio,Clutch Command

Shifting Maneuver

Simulation Models

Engine Controller

MotoTron

Measurable Engine Signals

CapturedEngineActuator Signals

Signals between Engine Controller and Vehicle Plant

Sensor signals:• AFM through throttle• engine speed• intake manifold

pressure and temperature

• EGR valve position• coolant temperature• injection pressure…

Actuator signals:• Ignition angle and

duration• injection angle and

duration• throttle valve, EGR

valve, and pressure control valve position PWM control signal

Signal for the Parallel HEV

13

Hybrid ECU

Electric Only

Engine Only

BatteryCharging

Hybrid

Reg

ener

ativ

e B

raki

ngBPP>0

APP>0

SOC>Max_SOC and T_Bat<Max and TrqRequest<MaxMotorTrq

SOC<Min_SOC or T_Bat>Maxor TrqRequest>MaxMotorTrq

SOC>Max_SOC

T_Bat>Max

T_Bat<Max

SOC<Min_SOC

• Vehicle operating mode control• Split powertrain torque between engine and electric machine to achieve

maximum fuel economy• Control regenerative braking to recover as much energy as possible

and ensure braking performance at the same time

14

Vehicle Mode and Energy Flow

(a) Electric only mode

(b) Battery charging mode

(c) Engine only model

(d) Hybrid mode. (e) Regenerative

braking mode

(a) (b)

(c)

(d) (e)

15

Overview of Engine ECU

Homogenous Mode

Stratified Mode

Engine Operation

Starter switch

Ignition switch

Engine Speed

Relative Airmass

Intake Pressure

Intake Temperature

Engine Torque Setpoint

Engine State

Coolant Temperature

Engine State

Relative Airmass

Engine Torque setpoint

Engine Torque

Combustion Mode Switch Airpath

Rail Pressure Control

Angular Processing

Unit

RailPressure

λ setpoint

ṁthrottle

Rail Pressure

ṁthrottle

Injection Signals

Spark Signal

Throttle position

Spark Signal

Throttle position

Intake PressureAmbient Pressure

EGR %

Turbocharger Control Signal

EGR Control Signal

Rail PressureMean Injection

quantity setpoint

Fuel Metering Unit

Pressure Control Valve

Cylinder wiseCrank angle

ENGINE ECU SUBSYSTEMS

Injection Signals

• Injection signals

• Ignition signals

• Throttle position

• Turbocharger control signal

• EGR control signal

• Rail pressure control signals

• Engine position

16

Motor Controller

1717

dSPACE Experiment Software - ControlDesk

18

# Vehicle Mode0 Battery charging1 Engine only2 Hybrid mode3 Motor only4 Regenerative breaking

Simulation Results: Vehicle Mode Control• Hybrid ECU control logic is

modified to avoid frequent mode change.

• Improved vehicle performance is shown in enlarged subfigures.

Vehicle operation modes in UDDS drive cycle

19

Simulation Results: Power and Energy Distribution

Wind drag

∑ −−−== GfWdrollTx FFFFmaF

rroll fmgF θcos= 2

21 AvcF dWd ρ= θsinmgFGf =

vFFFmavFP GfWdrollTreq ⋅+++=⋅= )(

Rolling resistance

Gravitational force

Vehicle power request, rolling resistance power, and wind drag power over a UDDS drive cycle

20

Simulation Results: Regenerative Breaking

• Designed to recover as much as possible

• Disabled when v < 0.1 km/h or SOC reached charging limit

• Limited by maximum motor power

• 81% braking energy was recovered, total 0.4638 kWh

0 100 200 300 400 500-45

-40

-35

-30

-25

-20

-15

-10

-5

0

Time(s)

Pow

er(k

W)

Total Braking PowerRegenerative Braking Power

Braking power in US06 drive cycle

21

Simulation Results: Engine StatesState value 1 2 3 4

Engine state Idling Engine Traction only

Traction &charging Battery

Traction withmotor

Engine On

30.8%

Engine idling

53.3%

0 200 400 600 800 1000 1200 14000

20

40

60

80

100

Veh

icle

Spe

ed (k

m/h

)

0 200 400 600 800 1000 1200 14000

1

2

3

4

5

Engi

ne S

tate

Time(s)

Engine states in UDDS drive cycle

22

Simulation Results: Engine Operating Region

• Engine operating points: HEV (left), conventional vehicle (right)

• Motor provides traction when torque demand below 100 Nm.

242

261

280

299

318

356

394432

Engine speed (rpm)

Tor

que

(Nm

)

BSFC Speed Torque Map

1000 1500 2000 2500 3000 3500 4000 4500 5000 55000

50

100

150

200

250

300

300

400

500

600

700

800

900

idlingmotor

242

Engine speed (rpm)

Torq

ue (

Nm

)

BSFC Speed Torque Map

1000 1500 2000 2500 3000 3500 4000 4500 5000 55000

50

100

150

200

250

300

300

400

500

600

700

800

900

23

Conclusions

Model-based design allows ECU development with vehicle plant models in the same environment.

Hardware-in-the-loop simulation enable very short development times with parallel control system validation.

• Model-based design HIL simulation are suitable for the development of complex control systems for the clean vehicles.

24

Acknowledgements

This research is supported by the National Science Foundation under Grant NO. OISE-1157647.

Laboratory of Intelligent Mechatronics and Embedded SystemsMichigan Technological University

http://me.sites.mtu.edu/chen/

25