Project 9 Traffic Flow Simulation

1
Project 9Traffic Flow Simulation

• Team 5

2
Contents

• Introduction Motivation
• Environment Modeling (Roadways)
• Simulation
• Statistics
• Visualization
• Conclusion

3
Introduction

• Real-time image processing and Robotics for
  automotive applications
• Driver assistance using Intelligent feedback on
  traffic conditions, cross-roads accidents, etc.
• Collision warning and avoidance system
• Intelligent parking assistance

4
Environment Modeling- Roadways

• Preparing the map
• The map is stored in a text file
• It is represented by
• Lines
• Curves

5
Representing lines

• ( x1, y1, z1 ) -gt ( x2, y2, z2 )

line 125 400 10 125 375 20
6
Representing curves

• ( x, y, z, r, anglestart, anglestop, direction)

arc 125 300 20 75 270 180 -1
7
Sampling

• Sampling distance
• Line
• length of line
• number of sampling points
• allocating the points
• Curve
• angle
• length of arc
• number of sampling points
• angle of sampling points
• allocating the points

8
Simulation of Car Traffic Flow

• Based on physics laws (kinematics)
• The control parameters
• the number of cars
• the size of the histogram interval (for
  generating statistics)
• Equations
• the velocity law
• the position law

9
The parameters of the cars

• Position on the road index and lane
• Current speed and cruising speed
• Maximum acceleration
• Maximum deceleration
• Reaction time of the driver

10
How the simulation works

• The time is divided into small intervals (
  10ms)
• In one time slice a car can only
• maintain speed
• accelerate
• decelerate

11
The algorithm

• for each car
• must decelerate ?
• yes decelerate !
• no current speed lt cruising speed can
  accelerate ?
• yes accelerate !
• no maintain speed.

12
When the car must decelerate ?

• We search for the cars in front of the current
  car
• When the cars in front of the current car are
  close together (20m), they block the road.
• We assume that in the next time-slice the cars in
  front will maintain speed (a reasonable
  assumption)
• After that, we check if we maintain the speed of
  the current car, the time distance between cars
  will be decreased below the time_dist (a constant
  value).

13
The car can accelerate?

• a car can accelerate when the time distance to
  the car in front is less than the time_distance
  or when an overtake is performed (not yet
  implemented)
• the maximum allowed acceleration is the greatest
  between the maximum possible acceleration of the
  car and the acceleration that will not lower the
  time distance under time_distance

14
Acceleration and Deceleration

• We update the position and the velocity

15
Utilization of road

• Dividing the road to smaller intervals
• Counting the number of cars on each interval -gt
  histogram
• Drawing a chart minimal, average, maximal
  utilization

16
Example
17
Example
18
3D Visualization

• Actually a 2.5D representation
• It can be rotated and zoomed in and out
• Still to do more !!

19
3D example
20
3D example
21
Thank you for your attention!