Development and Evaluation of a Novel Traffic Friendly ... · Novel Traffic Friendly Commuter Vehicle ... – Automated highway systems ... zSteer-by-wire system

Development and Evaluation of a Novel Traffic Friendly Commuter Vehicle

Rajesh Rajamani

Department of Mechanical Engineering

University of Minnesota

2

Highway Congestion

Traffic congestion is a significant - and growing -problem in the country’s major metropolitan areasSome statistics

– Average traffic delay due to congestion grew 235% between 1982-1997

– Uncongested travel fell from two-thirds of travel in 1982 to one-third travel in 1997

– The increase in traffic demand every year exceeds the increase in capacity due to new construction

Traffic congestion will continue to become worse

3

Approaches to Addressing Highway Congestion

What solutions do researchers in the automotive industry and researchers in the vehicle dynamics community offer to highway congestion?

– Automated highway systems

– Intelligent adaptive cruise control systems

4

Automated Highway Systems

AHS lanes will have three times the capacity of regular highway lanes -Vehicles will travel together in closely-packed “platoons”.Dedicated to automated vehicles -regular passenger cars will have to be specially instrumented to travel on AHS lanes.

AHS is “dual mode” transit– Your instrumented car can travel on AHS and can also

take you on regular roads

5


NAHSC demonstration in 1997Demonstration operated continuously several hours a day for 3 weeks– Passenger rides given to over a 1000 visitors

Video

6


A variety of challenges need to be addressed before AHS can become a reality– technical – economic– social– legal

The AHS concept received a setback with the demise of the NAHSC

7

Adaptive Cruise Control

Without preceding vehicleMaintain constant speed

With preceding vehicleMaintainsafe distance

radar

8

Adaptive Cruise Control

Typical ACC System– Constant time-gap (CTG) spacing between

vehicles

Research Questions– Should ACC systems be designed to maintain a

constant time-gap between vehicles ? – Can traffic capacity and safety be simultaneously

improved by clever ACC design ?

L

hv

hvL +Desired spacing =

9

Narrow Commuter Vehicles

Motivation– Double highway capacity by using a half-width lane

instead of a regular highway lane– Motorcycles

take up less space on the highwayneed less space for parkingare fun to driveare energy efficient

Can we develop a narrow vehicle that can– provide the benefits of a motorcycle– the safety of a car– be as easy to drive as a car ?

10

Narrow Commuter Vehicles

A narrow vehicle needs to tilt into a curve for stability

Front Viewθ

mg

mVR

2

h

11

Project Objective

Develop an automatic control system that

– Keeps the vehicle balanced while it is traveling straight

– Tilts the vehicle into a curve during cornering

– Reduces acceleration experienced by the driver

– Always maintains vehicle stability

12

Some interesting prototype narrow vehicles

The Volkswagen 1-liter car

Volkswagen’s objective: Develop a prototype vehicle that can provide 100 km/liter (250 miles / gallon !)

• Solution: A narrow vehicle two-seater platform, aerodynamic design, lightweight carbon-fiber reinforces composite body

• Very agile handling, provides driving pleasure combined with a fuel economy never seen before

13

Some other interesting prototype narrow vehiclesIntroduction

Control system

Prototype

Simulation

Experiments

Conclusion

Safe

and

Sta

ble

Nar

row

Com

mut

er V

ehic

les

Fig. 1 Ford “Gyron” (1961)

Fig. 2 GM “Lean Machine” (1983)

Fig. 4 “Carver” (2003)

Fig. 3 Daimler “Life Jet” (1997)

14

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Control system - Dynamic Model

z

Tc.g.

β

θ

mgRz

Ry

y

h2θ&

Vy ψ&&& +

hθ&&

θβ

y’

z’

h

Safe

and

Sta

ble

Nar

row

Com

mut

er V

ehic

les

y

ψY

X

a) top view b) front view

θ

Fig. 5 Degrees of freedomFig. 6 Bicycle model

(1)ThFFhmghmhmI

FlFlI

mgFFhmmhVmym

rfx

rrffz

rf

++−+−−=

−=

++=−++

)cos()()sin())cos()cos()sin((

)sin()sin()cos( 2

θθβθθθθθ

ψ

βθθθθψ

&&&&&

&&

&&&&&&

15

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Control system – Driver Model

Driver Model: Look-ahead-error based model (Guldner et al. 1996)

Safe

and

Sta

ble

Nar

row

Com

mut

er V

ehic

les

ds

e2ds

e1e1

e2

Desiredtrajectory

Look-ahead error

Fig. 7 Look ahead error

*)(*_ 21 ff

erroraheadLook

dseekinputDriver δ++−=−−43421

(2)

* Parameters can be defined as a function of velocity to account for driver steering behavior change due to velocity variations.

16

Tilt Control Systems

Direct tilt control (DTC)– An independent actuator is used to

provide tilt torque– Tilt and lateral dynamics can be

independently controlled– For low speed operation (less than 5 m/s

or 11.18 mph).

Steering tilt control (STC)– Steering angle is used to control both tilt

and lateral position– Challenge: One control input, two control

tasks– For high speed operation (higher than 10

m/s or 22.37 mph).

Possible actuation systems

DriverVehicle

STC

Throttle

Brake

Driver steering

θ

Actual steering

Throttle

DriverVehicle

DTC

Brake

Driver steering

θ and alatTilt

torque

17

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Introduction – Proposed control systemSa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

4 mph 10 mph 55 mph0 mph

Vehicle Speed

High speed controller

(STC)

Low speed controller

(DTC)

Tilt brake

31 2

Fig. 8 Control scheme

18

Direct Tilt Control (DTC)

a) Simple approach (stand alone):

RgV

gV

des

2

==ψθ&

MG

hFf + Fr

mg

θ

FN

ghmIIVy rotationrotwheelwheel

des

)/()(2)( __ ψωψθ ψ &&&& −−+

=

b) Proposed approach I (stand alone):

- Accounts for lateral acceleration of vehicle in addition to centripetal acceleration.

- Accounts for gyroscopic moment due to rotating wheels.

- Results in a significant reduction in transient tilt torque requirement.

- Results in zero tilt torque requirement at steady state.

Moments at c.g.

θθθ &21 )( kkT des −−−=

c) Proposed approach II (Integrated):))(()_( 21 ffsdes edekksinputDriverks δθ ++−==

19

Steering Tilt Control (STC)

Steer-by-wire system– Driver steering input is modified before being used

to steer the wheels– Driver input is interpreted as a desired tilt angle

With appropriate definitions, leads to the same steady state value of steering angle

DriverVehicle

STC

Throttle

Brake

Driver steering

θ

Actual steering

))(()_( 21 ffsdes edekksinputDriverks δθ ++−==

44344214444 34444 21&&

statesteady

des

transient

desdesdesekkk )(

43 )()( θθνθθθθθδ −−+−+−=

20

Control system - Combined system (SDTC)

Fig. 12 SDTC

• Both STC and DTC used in parallel.

-DTC active at low vehicle speed.

-STC active at high vehicle speed.

Driver steering

DriverVehicle

STC

Throttle

Brake

θ

Actual steering

DTC

θ

Tilt torque

β

α

SDTC

Safe

and

Sta

ble

Nar

row

Com

mut

er V

ehic

les

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

0 5 10 15 20 25

0

0.2

0.4

0.6

0.8

1

Velocity [m/s]

Mag

nitu

de

α

β

Vo Vf

Fig. 13 Weighting functions

⎪⎪⎪

⎩

⎪⎪⎪

⎨

⎧

>

<<⎥⎥⎦

⎤

⎢⎢⎣

⎡+

−−

+

<

=

f

fofo

o

o

VVif

VVVifVVVV

VVif

0

)2

))()(sin(1

21

1

ππα

αβ −= 1

])([

)()]())([

21

43

θθθα

θθθθθθβδ

&

&&

kkT

dtkkk

des

desidesdes

−−−=

−+−+−= ∫

(9)

(10)

21

Control system - Tilt BrakeIntroduction

Control system

Prototype

Simulation

Experiments

ConclusionStart

V

Brake offθdes = ks*driver_input

V > Vmin ?

|driver_input| < ε1 ?

f fdesdes kk δθθθθδ +−+−= )()( 43&&

Brake onθ = 0

δ = driver_input

V < Vmin ?

θdes = 0

|θ| < ε2 ?

Brake onδ = driver_input

No

driver_input

V

θ

No

No

Yes

Yes

Yes

Yes

Safe

and

Sta

ble

Nar

row

Com

mut

er V

ehic

les

Fig. 14 Tilt brake algorithm

22

Prototype

Fig. 15 Second generation prototype

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Safe

and

Sta

ble

Nar

row

Com

mut

er V

ehic

les

23

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Prototype - DesignSa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

• Engine: 125cc Yamaha

• Longitudinal acceleration: + 0.7 g

• Tilt limit: +30o

• Bump limit: +15o (0.1 m)

• Steering limit +30o

• Rake angle: 9.5o

• Front wheel trail: 0.06 m

Vehicle Specifications

Fig. 16 Vehicle dimensions

24

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Prototype - Tilt mechanismSa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

Video

Fig. 17 Vehicle tilt mechanism

25

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Prototype - Tilt Mechanism actuation

Fig. 18 Tilt motor

26

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Prototype - Steering Mechanism

(a) Direct steering (b) Steer-by-wire

Fig. 19 Steering system

27

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Prototype - Tilt Brake System

Fig. 20 Tilt brake

28

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Prototype - InstrumentationSa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

PC - 104

IBM computer

Tilt servomotorIncremental encoder 1

Steering servomotor

Tilt brake

electromagnet

Incremental encoder 2

Absolute

encoder

Hall effect sensor

Xbow IMU

Driver input

Tilt angle

Steering angle

Speed

ay, ax, az

yaw, pitch, roll

Sensors

States

Ref. input

Actuators

Control inputs

Fig. 21

29

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Simulation – DTC (Simple Vs. Proposed)

Operating conditions:

• Vehicle velocity: 5 m/s (11 mph).

• A desired trajectory of a straight line for the first 50 seconds followed by a

circular path of radius 8 m (26 ft).

Safe

and

Sta

ble

Nar

row

Com

mut

er V

ehic

les

Fig. 22 Tilt torque requirement

30

Simulation – DTC (Simple Vs. Proposed)

a. Proposed approach b. Simple approach

Fig. 9 States

31

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Simulation – SDTC

The SDTC system was simulated for:

• Vehicle velocity varying between 1 m/s (2.24 mph) and 25 m/s (55.92 mph).

• A desired trajectory of a straight line for the first 50 seconds followed by a

circular path of radius 500 m (1640.42 ft).

Safe

and

Sta

ble

Nar

row

Com

mut

er V

ehic

les

0 20 40 60 80 100 120 140-5

0

5

10

15

20

25

30

t [sec]

y[

]

1

Vel

ocity

[m/s

] STC

Transition

DTC

Fig. 29 Velocity profile of vehicle

32

SDTC Simulation

Trajectory tracking

0 200 400 600 800 1000 1200 1400-200

0

200

400

600

800

1000

1200

x [m]

y [m

]

Actual trajectoryDesired trajectroy

A

Circular path

Straight path

Zoomed view of ‘A’

832 832.2 832.4 832.6 832.8 833 833.2

7.4

7.6

7.8

8

8.2

8.4

8.6

x [m]y

[m]

Actual trajectoryDesired trajectroy

33

Simulation – SDTC Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Transition DTCSTC

Safe

and

Sta

ble

Nar

row

Com

mut

er V

ehic

les

0 20 40 60 80 100 120 140 160 180 200-10

-8

-6

-4

-2

0

2

4

6

8

10

Tilt

torq

ue [N

m]

t [sec]0 20 40 60 80 100 120 140 160 180 20

-1

0

1

2

3

4

5x 10-3

t [sec]

[rad]

[N-m

]

0

Fig. 32 SDTC - Tilt torque (T) Fig. 33 SDTC - Steering input (δ)

DriverModel

DTC

Roadδ

PLANTSTC T

θ

34

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Simulation – SDTC Sa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

Tilt Zone 1.4 m

Track Width 1 m

Allo

wa b

le E

rro r

0.2

m

All o

wa b

le E

rro r

0.2

m

Half Lane 1.8 m

0 50 100 150-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

t [sec]

Sta

tes

Vehicle tilt [rad]

Position error [m]

Heading error [rad]Look ahead error [m]

Fig. 30 SDTC – Lateral states Fig. 31 Allowable position error

35

Introduction

Control system

Prototype

Simulation

ExperimentsConclusion

Experiments

Validate the tilt stability and trajectory tracking.• Turn maneuver experiments

• Lane change maneuver experiments

Validate the tilt brake system.• Sinusoid steering input experiments

Vehicle handling investigation.• Under steering experiment

• Neutral steering experiments

• Constant yaw rate experiments

System limits.• Steering subsystem vibration

Safe

and

Sta

ble

Nar

row

Com

mut

er V

ehic

les

36

Introduction

Control system

Prototype

Simulation


Experiments - General stabilitySa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

Video

Fig. 34 Tilt stability

37

Introduction

Control system

Prototype

Simulation


Experiments – Trajectory tracking Sa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

Video

Fig. 35 Slalom course

38

Introduction

Control system

Prototype

Simulation


Experiments – Turing maneuver Sa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

Fig. 36 Desired tilt angle tracking

39

Introduction

Control system

Prototype

Simulation


Experiments – Turning maneuver Sa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

Fig. 37 Vehicle yaw rate

40

Introduction

Control system

Prototype

Simulation


Experiments – Lane change maneuver Sa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

Fig. 38 Desired tilt angle tracking

41

Introduction

Control system

Prototype

Simulation


Experiments – Lane change maneuver Sa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s


42

Introduction

Control system

Prototype

Simulation


Experiments – Tilt brake validationSa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

Fig. 40 Tilt angle tracking and tilt brake status

43

Introduction

Control system

Prototype

Simulation


Experiments – Tilt brake validationSa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s


44

Introduction

Control system

Prototype

Simulation


Experiments – Neutral steeringSa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

Fig. 42 Neutral steering

45

Introduction

Control system

Prototype

Simulation


Experiments – System limitationsSa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

Video

Fig. 43 Steering subsystem instability

46

Introduction

Control system

Prototype

Simulation


Experiments – System limitationsSa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

0 2 4 6 8 10 12 14 16 18-10

-5

0

5

10

15

20

25

time [s]

Steering motor current rating [Amp]

Steering motor current [Amp]

Desired tilt [deg]

Actual tilt [deg] Speed [mph]

0 2 4 6 8 10 12 14 16 18-10

-5

0

5

10

15

time [s]

oscillation

Steering motor current rating [Amp] Steering motor current [Amp]

Actual steering [rad]

Desired steering [rad]

(a) Steering motor current (b) Actual steering

Fig. 44 Steering subsystem instability

47

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

Conclusion

A control system that stabilizes an NTV from start up to highway speed while accurately following a desired trajectory was successfully designed.

– A DTC system that is suited for an integrated SDTC (STC + DTC) system was proposed.

– A method for smoothly shifting control efforts between the STC and DTC system was developed.

– A Tilt brake system and algorithm was developed for low speed operation.

A new stand alone DTC system with minimal transient torque requirement and zero steady state torque requirement was developed.

A second generation tilting vehicle was fabricated and the proposed compound control system was successfully implemented.

Safe

and

Sta

ble

Nar

row

Com

mut

er V

ehic

les

Limitations

• Actuator saturation

Future work

• Replace steering system motor with at least 10 Amp motor and investigate high speed performance.

• Driver Haptic feedback.

48

Introduction

Control system

Prototype

Simulation

Experiments

Conclusion

ConclusionSa

fe a

nd S

tabl

e N

arro

w C

omm

uter

Veh

icle

s

Acknowledgement:

This project was partly supported by funds from the National Science Foundation under Grant CMS-0411455 and by funds from the ITS-Institute, University of Minnesota.

Research Team:

Samuel Kidane

Lee Alexander

Patrick Starr

Rajesh Rajamani

Max Donath

49

Safe

and

Sta

ble

Nar

row

Com

mut

er V

ehic

les

~ The End ~

Documents

Development and Evaluation of a Novel Traffic Friendly ... · Novel Traffic Friendly Commuter Vehicle ... – Automated highway systems ... zSteer-by-wire system