Adaptive Cruise Control (ACC)

1
Adaptive Cruise Control (ACC)
Group Memebers Mirza Abdel Jabbar Baig
(3256498) Mohammad Ali Akbari (3299852) Navid
Moazzami (3413826) Hasan Ashrafuzzaman (3384661)

• ELG 4152 Project
• Professor Riadh Habash
• TA Fouad Khalil

2
Reference

• 1 A Safe Longitudinal Control for Adaptive
  Cruise Control and Stop-and-Go ScenariosMartinez,
  J.-J. Canudas-de-Wit, C. Volume 15,  Issue 2, 
  March 2007 Page(s)246 258
• 2 Modeling a Cruise Control
• http//www.library.cmu.edu/ctms/ctms/examples/cru
  ise/cc.htm
• 3 Highway Speed Controller
• http//www.site.uottawa.ca/misbah/elg4392/HC12Co
  deWarriorC/HighwaySpeedController/project.c
• 4 W. Jones, Keeping cars from crashing, IEEE
  Spectrum, vol. 38, no.
• 9, pp. 4045, Sep. 2001.
• 5 M. A. Goodrich and E. R. Boer, Designing
  human-centered automation
• Tradeoffs in collision avoidance system design,
  IEEE Trans. Intell.
• Transp. Syst., vol. 1, no. 1, pp. 4054, Mar.
  2000.

3
Problem Statement

• The main problem regarding the normal Cruise
  Control technology is that it is not aware of
  other vehicless movement
• The driver must be always aware. Hence,
  possibility of mistakes
• Possibility of collision with the leading car if
  not manually slowed down

4
Proposed Solution

• Introduce Adaptive Cruise Control for
  longitudinal control of the vehicle
• Speed would be automatically adjusted for safe
  inter-distance
• Once safe inter-distance is reached, the speed
  would return to the desired speed set by the
  driver

5
Technical Objectives

• To design a control system for ACC.
• No overshoot
• Settling Time of about 4-7 seconds.
• No oscillation (because no overshoot)
• A steady-state error of 0

6
Vehicle Characteristics

• If the inertia of the wheels is neglected, and it
  is assumed that friction (which is proportional
  to the car's speed) is what is opposing the
  motion of the car, then the problem is reduced to
  the simple mass and damper system shown in the
  next slide.

7
Vehicle Characteristics
8
System Block Diagram 2
9
Controller Selection

• Which kind of Controller is the best?
• No controller.
• P controller.
• PI controller.
• PID controller.
• PD controller.

10
Controller Selection

• P Controller

No Controller
Kp 10000 Settling Time 0.389s Steady state
error 2
Settling time 76.7 s Steady state error gt 98
11
Controller Selection
PI Controller
Final choice is PI Controller
Kp800, Ki40 Settling time 4.89 s Steady state
error 0
12
Distance Checking 1

• Three scenarios
• dr gt d0, cruises at desired speed, ACC inactive
• dr lt dc, danger zone, ACC enables to slow down
• d0 lt dr lt d0, ACC is enable to reach safe
  inter-distance

13
Implementation of Distance Checking 3

• The distance checking algorithm only requires a
  minimum distance and a range.
• The algorithm calculates the actual minimum
  distance (gt provided distance) and maximum
  distance and then outputs the new speed of the
  vehicle.
• The user can also provide a maximum and minimum
  speed for the vehicle.

14
Implementation of Distance Checking
temp(300(speedmax-speedmin))/(12range)
minimum_Distance(minimum_Distance32)/10
max_Distance minimum_Distance (3range)
if (distance gt (max_Distance))
speed speedmax if (distance lt
minimum_Distance) speed 0
if ((distance lt max_Distance) and
(distancegtminimum_Distance)) if
leader_speed gt 0 speed
((100speedmin-(kvit(minimum_distance))) temp
distance)/100 else speed
((100speedmin(kvit(max_Distance))) temp
distance)/100
15
Simulation
The following parameters were used for the
simulation

• Maximum follower vehicle speed 100 m/s
• Minimum follower vehicle speed 0 m/s
• Minimum distance 40 m
• Range 20 m
• Initial distance 80 m
• Kp 800
• Ki 40
• b 50
• m 1000

16
Final Model (simplified)
17
Simulation
Yellow Distance between two vehicles Blue Speed
of the leader vehicle Purple Speed of the
follower vehicle
18
Limitations/Conclusion

• Not a complete transfer function of the vehicle
  and environment.
• Linear distance-checking model.
• No limitations on the acceleration and jerk.
• Our model is simplified compared to real-time
  models, but can be used to implement a practical
  ACC.