A warning algorithm for Lane Departure Warning system

Single Vehicle Road Departure (SVRD) is a main cause of enormous loss of life and property on highways. In total, SVRD accidents amount to around 20 percent of traffic accidents and 40 percent of fatalities in total. Recent developments of Advanced Driver Assistant System (ADAS) such as Lane Departure Warning (LDW) system which perceives road boundaries through optical sensors that trigger a warning when the driver leaves the lane without using their turn signal. However, a robust LDW system can not be realized with only turn signal usage to suppress warnings during an intended lane change. This paper will analyze not only driving position, but also driving tendency and driver state to determine whether an unintended lane departure is detected. The paper will explore how the warning algorithm summarizes information from the environment, vehicle and driver and establishes a model so that real unintended lane departure can be detected and a warning will be triggered accordingly.

[1]  Dean A. Pomerleau,et al.  Driver-adaptive lane departure warning systems , 1999 .

[2]  Michael A. Goodrich,et al.  Maintaining driver autonomy with collision avoidance systems: a satisficing approach , 1997, Proceedings of Conference on Intelligent Transportation Systems.

[3]  T. Pilutti,et al.  Fuzzy-logic-based virtual rumble strip for road departure warning systems , 2003, IEEE Trans. Intell. Transp. Syst..

[4]  M. Gonzalez-Mendoza,et al.  A comparison of road departure warning systems on real driving conditions , 2004, Proceedings. The 7th International IEEE Conference on Intelligent Transportation Systems (IEEE Cat. No.04TH8749).

[5]  Paul Milgram,et al.  The Development of a Time-Related Measure to Describe Driving Strategy , 1984 .

[6]  Sukhan Lee,et al.  Experiments on decision making strategies for a lane departure warning system , 1999, Proceedings 1999 IEEE International Conference on Robotics and Automation (Cat. No.99CH36288C).