Model Based Testing of Automotive Control Functions · Confidential | Ankit Verma (RBEI/EEI5) | 10/7/2014 | © Robert Bosch Engineering and Business Solutions Limited 2014. All rights

Confidential | Ankit Verma (RBEI/EEI5) | 10/7/2014 | © Robert Bosch Engineering and Business Solutions Limited 2014. All rights reserved, also

regarding any disposal, exploitation, reproduction, editing, distribution, as well as in the event of applications for industrial property rights.

Model Based Testing of Automotive Control Function

Model Based Testing of

Automotive Control Functions

Matlab Expo, April 2015

Model-Based-Testing

System

Controls

Plant ModelAutomotive

Frontloading

Real Time

Closed-Loop

Code-genera

tion

Legacy

ECU

MBD

Calibration

Syste

m U

nder Test

Model Coverage

Validation

Verification

Test CaseCoding

Software

Hardware

QualityAccuracy

Functions

Feedback

Ankit Verma (RBEI/EEI5)

1

Simulation

Variants

Parameter

Consta

nt

Variables

Software-in-the-Loop

Model-in-th

e-LoopExperiments

Virtual

Concept

Simulation

Flashing




Agenda

1. Need for Model Based Testing

2. Conventional Vs MBD Approach

3. Test Environment Description

4. Testing of SUT using MBD Approach

5. System Simulation using Plant Model

6. Results

7. Benefits

8. Summary

2




1. Increasing Engine Control Complexity

Function Complexity increasingChallenges in maintaining quality

Increase in number of calibration More calibration effort

Increase number of ECU Increased inter domain connectivity

Need for Model Based Testing

Source: ETAS

New Functions &

VariantsIncreased inter domain connectivity

3




Conventional Vs MBD Approach

MBD Approach

System Analysis

Concept Eval

Concept Control Model

Plant Model

MiLOverall Control Model

Plant Model

SiLControl Model

Plant Model

RP/PiLControl Model

Plant Model

HiL

Plant Model

ViL

Vehicle

Conventional Approach

Function Model

Testing

Source Code Unit Testing Func validation

Compile/LinkCode generationFunction Design

Prototyping through H/W

Engine Test Bench

System Analysis

PC Environment Lab Environment

4




PLANT MODEL

Drive Train

*.dll from legacy

c-code

New Function

Model (ML/SL)

SUT (System

Under Test)

CONTROL MODEL

Software & Model in Loop

Test Output

Verification

Scheduler(DGS Lib)

Intake Vehicle

Closed Loop Test Environment Set-up

New

Control

Function -

SUT

(System

Under Test)

Engine

5




Testing of SUT (System Under Test)

using MBD Approach

6

Control Model




Closed loop Test Environment

MSE Plant Model

Control

Model Plant Model

Throttle

Angle [Deg]

Ignition

Angle [o CA]

Ignition

Timing [ms]

SoI [o CA]

Engine

Speed [rpm]

Manifold

Pressure [hPa] Gear [-]

Vehicle Speed [Kmph]

Lambda

Acc Pedal

Position

7




Test Case Description

Engine is in neutral gear

Check the engine behaviors with open drive train (neutral gear)

without consumer

S.No Test Case Description Observations

1. Unit Testing of Low Idle Speed Governor • Engine speed settles around the idle speed

• Low Idle governor working with speed deviation of

around -50 to 50 RPM

2. Engine in the Neutral Gear Engine Speed settles @ Idle Speed

3. Engine is in First Gear Engine speed increase during gear change from

neutral First

Engine speed settles to idle speed when gear

change from first zero.

8




Idle Speed Control – Test Results

Set Point Engine Speed [rpm]

Engine Speed [rpm]

Inference:

• Mass flow isn’t sufficient to achieve the desired idle

speed.

• Therefore, stepper configuration and by-pass valve

dimensions need to be calibrated

9




Next Step

1) Study of Engine behavior using simulation

Throttle & Ignition Sweep for different Engine Speed (800 – 2000

rpm)

Based on load variation to determine air mass flow through ISA

2) Component Dimensioning

Based on air mass flow – Area & Dia. of ISA to be fixed

10



MSE work package: Model based Testing

Idle Speed Actuator - Component Sizing

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

0 5 10 15 20 25 30 35

To

rqu

e [

Nm

]

Throttle Angle [%]

Torque @ Varying throttle and Fixed Speed

800 RPM 1000 RPM

1200 RPM 1400 RPM

1600 RPM 1800 RPM

2000 RPM

0

0.25

0.5

0.75

1

1.25

1.5

1.75

2

2.25

2.5

2.75

3

3.25

3.5

3.75

4

4.25

4.5

4.75

5 10 15 20 25 30

Air

Ma

ss F

low

[K

g/h

r]

Throttle Angle [%]

Air Mass flow @ Varying Throttle and Fixed Speed

800 RPM 1000 RPM 1200 RPM

1400 RPM 1600 RPM 1800 RPM

2000 RPM

Considering 0.5-1 Nm

load variation from

electrical loads

(magneto) ideal low

idle speed could be

above 1400 rpm

Based on the load

variation of 0.5 –

1Nm, mass flow of

1.25 Kg/hr through

ISA is required

Torque Air Mass Flow Area of By Pass Valve

11




Component Dimensioning

d

By pass Valve

Old By-Pass

Valve

New By-Pass

Valve

Diameter 2.8 mm 3.98 mm

12




New By-pass Vs Old By-pass valve - Results

Set Point Engine Speed

[rpm]Engine Speed [rpm]

Eng

ine

Sp

ee

d [rp

m]

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0 20 40 60 80 100 120 140 160 180 200

Stepper Motor Steps [-]

Stepper Area [cm2]

Eng

ine

Sp

ee

d [rp

m]

Set Point Engine Speed

[rpm]Engine Speed [rpm]

0

0.002

0.004

0.006

0.008

0.01

0.012

0 20 40 60 80 100 120 140 160 180 200

Ste

pp

er

Are

a [

cm

2]

Stepper Motor Steps [-]

Stepper Area [cm2]

By P

ass A

rea

[cm

2]

By P

ass A

rea

[cm

2]

Old Stepper New Stepper

13

Confidential | RBEI/EEI5 | 10/7/2014 | © Robert Bosch Engineering and Business Solutions Limited 2014. All rights reserved, also regarding any

disposal, exploitation, reproduction, editing, distribution, as well as in the event of applications for industrial property rights.


Pre-Calibration

14




Pre-calibration

0

0.0005

0.001

0.0015

0.002

0.0025

0.003

0.0035

-300 -200 -100 0 100 200 300

%T

orq

ue C

hange/r

pm

Engine Speed deviation [rpm]

P-gain (Air Path)

P-gain

0

0.05

0.1

0.15

0.2

0.25

-800 -600 -400 -200 0 200 400 600 800

%To

rqu

e C

ha

ng

e/r

pm

Engine Speed deviation [rpm]

P-gain (Air Path)

P-gain

Eng

ine

Sp

ee

d [rp

m] Set Point Engine Speed [rpm]

Engine Speed [rpm]

Eng

ine

Sp

ee

d [rp

m]

Deviation -100 to 100 rpm Deviation -50 to 50 rpm

15




Efficiency increase including development cost reduction

Early evaluation of control models and new control strategies frontloading

Reduce effort for validation of control models speed-up

Do more in virtual environment

Reduce dependence on dyno testing

System Level:

Concept evaluation before “make” decision

Investigate system behavior/function at early stage

Simulate extreme/rare environment conditions, safe experiments

Reusing plant models developed for other series project.

Benefit of Model Based Development

16




ECU-SW Level:

Easier debugging/reproducing of problems on PC

No reprogramming/flashing after every change in the functionality

Reuse ECU data

Benefit of Model Based Development

17




Desktop PC Simulation of implemented SW-functions together with the

available plant models allows a very early pre-calibration of these

functions and optimally supports the front-loading process in the area

of ECU-SW verification and validation.

The accuracy of the models is sufficient for concept evaluation w/o

measurement data and can be improved with measurement results to use

it for calibration tasks

Finally to deploy the virtual environment requires software as well as

system knowledge.

Summary

18




19

Documents

Model Based Testing of Automotive Control Functions · Confidential | Ankit Verma (RBEI/EEI5) | 10/7/2014 | © Robert Bosch Engineering and Business Solutions Limited 2014. All rights