1 of 12

4 min presentation in DD2438

Group 16

Farouk Miled and Rantao Dong

https://docs.google.com/presentation/d/1AEhrvdoOA-Nr_5EoMbjozHjY8whdiub3kXRUeMwCLQE/edit?usp=sharing

2 of 12

Overview of our plan so far...

  • Use A* algorithm
  • Use pure pursuit (Rantao)
  • Use Collision avoidance (sensors on the car with Raycast).
  • Have completed all five terrains

3 of 12

Overview of our plan so far...

4 of 12

Overview of our plan so far...

5 of 12

Details of our approach

  • Improvement of A* (to avoid possible collisions):

Inflate obstacles

but the inflation radius can not be too large, or some narrow ways may be considered as blocked

6 of 12

Details of our approach

  • Improvement of A* (to avoid possible collisions):

Inflate obstacles + ‘obstacle penalty’

Original A*: F = G + H (for H we use Manhattan distance)

Our improvement: F = G + H + ‘Obstacle Penalty

Obstacle Penalty: closer to the obstacles

→ higher penalty

→ higher F value

→ the final path will be more distant from the obstacles

7 of 12

Details of our approach

  • Pure Pursuit:

Steering: lager angle to the target point, larger steering

Acceleration: more distant from the target point, larger acceleration

& larger steering, smaller acceleration (slow down at corners to avoid collisions)

  • Collision avoidance with Raycast :

Steering : adjusted depending on which sensors are activated.

8 of 12

9 of 12

Problems and plan for next week

  • The algorithm is not robust enough

Search for some papers to get more ideas

Use sensors to avoid obstacles

  • The car is not fast enough

Try to adjust parameters

Try different tracking methods

10 of 12

Problems and plan for next week

  • The collision avoidance is unstable.

Work on smoothing the path and the steering.

  • Sensitivity to the sensor length or range.

11 of 12

12 of 12

Progress Status Week 5

  • Comment to customer paying 250 000kr for the report:
    • All is well. We have spent a bit more time than planned, and progress is as expected
  • Planned Time spent: 50%
    • (Out of the combined 200h)
  • Actual Time spent: 50%
    • Out of the combined 200h
  • Actual Progress: 40%
    • (estimate progress towards completing assignment)
  • Risk of not completing assignment: 3%