modeling road traffic

15
Modeling Road Traffic Greg Pinkel Brad Ross Math 341 – Differential Equations December 1, 2008

Upload: garrett-stewart

Post on 31-Dec-2015

43 views

Category:

Documents


0 download

DESCRIPTION

Modeling Road Traffic. Greg Pinkel Brad Ross Math 341 – Differential Equations December 1 , 2008. Contents. Introduction Model Revised Model Solving our model Conclusions. Introduction. Modeling the behavior of cars in traffic Car Following Model - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Modeling Road Traffic

Modeling Road Traffic

Greg PinkelBrad Ross

Math 341 – Differential EquationsDecember 1, 2008

Page 2: Modeling Road Traffic

Contents• Introduction

• Model

• Revised Model

• Solving our model

• Conclusions

Page 3: Modeling Road Traffic

Introduction

• Modeling the behavior of cars in traffic

• Car Following Model

• Quick-Thinking-Driver (QTD) Car Following Model

• Major Results:– If lead car stops suddenly, system looks like a

damped harmonic oscillator– Temporal Following Spacing should be ≥ 2 sec

Page 4: Modeling Road Traffic

Overview• Purpose – traffic modeling• Models

– 2-car model, x0 position of leading car, x1 position of following car (meters)

– Acceleration of following car (m/s2)– Sensitivity coefficients, λ (1/sec.) and β (1/sec.2)– Reaction time T (seconds)– Preferred separation, d (meters) or temporal

separation, τ (seconds)

Original Model

dt

tdx

dt

tdx

dt

Ttxd )()()( 102

12

Quick Thinking Driver Model dtxtx

dt

txd )()(

)(102

12

Page 5: Modeling Road Traffic

Original Model

• λ : sensitivity coefficient, units s-1

• Impact of and

• Addition of T is a Delay Differential Equation, very difficult to solve analytically

dt

dx

dt

dx 10

dt

tdx

dt

tdx

dt

Ttxd )()()( 102

12

dt

dx

dt

dx 10

Page 6: Modeling Road Traffic

Quick Thinking Driver Model

dtxtxdt

txd )()(

)(102

12

• Ignores reaction time for driver

• β : sensitivity coefficient, units s-2

• Assuming d is a constant “preferred separation”

• Impact of x0 – x1<d and x0 – x1>d

Page 7: Modeling Road Traffic

Preferred Separation• Tends to depend on speed (Think of real world situation)

• Relate preferred spatial separation to temporal separation by

with a preferred temporal separation of τ seconds.

dt

tdxd

)(1

Page 8: Modeling Road Traffic

Review (and the 2nd order D.E.)

dt

tdx

dt

tdx

dt

Ttxd )()()( 102

12

)()()()(

011

21

2

txtxdt

tdx

dt

txd

dtxtxdt

txd )()(

)(102

12

1)

2)

Page 9: Modeling Road Traffic

Physically Reasonable Stopping

With initial conditions ,

Dxdt

dx

dt

xd 1

12

12

Udt

dx

)0(10)0(1 x

D = Separation of vehicles, x0 –x1, at time of x0 sudden stop

m/sm

Page 10: Modeling Road Traffic

Solving: Laplace Transforms and Mathematica!

Page 11: Modeling Road Traffic

The Solution: X1(t)

Page 12: Modeling Road Traffic

Major Results: Simple Harmonic Motion • Have three distinct cases for different values of τ:

)4( 2t

Overdamped,

(τ=1s)

Underdamped,

(τ=3s)

Critically Damped,

(τ=2s)

42

42

42

42

U=10 m/s, D=10m, β=1s-2

Page 13: Modeling Road Traffic

Significance of Results

• Always have collision with the underdamped situation

• To avoid collision: τ2 ≥ 4, τ ≥ 2 seconds

42 42

Page 14: Modeling Road Traffic

Minimum Stopping Distance• Collision is possible but will not necessarily

happen: a βτ2 > 4, βτ2 = 4

• Collision can be avoided by ensuring initial separation greater than Dstop

• Derived by setting the solution of x(t) equal to zero with given conditions, D becomes Dstop

• Having initial separation greater than this ensures that the collision will be avoided

4,4)(

2 2

2

UDstop

Page 15: Modeling Road Traffic

Conclusion

• Based on conditions: βτ2 > 4, βτ2 = 4, and βτ2 < 4 – collision is only avoidable for βτ2 ≥ 4

– collision only avoided if x0 – x1 > Dstop

– Because βτ2 < 4 always results in collision, maintain temporal time cushion at least 2 sec behind lead car