Premium
Estimation of Driver's Intention to Stop or Pass Through at Yellow Traffic Signal
Author(s) -
MABUCHI RYUKI,
YAMADA KEIICHI
Publication year - 2015
Publication title -
electronics and communications in japan
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.131
H-Index - 13
eISSN - 1942-9541
pISSN - 1942-9533
DOI - 10.1002/ecj.11649
Subject(s) - intersection (aeronautics) , signal (programming language) , dilemma , computer science , moment (physics) , computer security , traffic signal , advanced driver assistance systems , signal timing , simulation , transport engineering , control theory (sociology) , engineering , real time computing , artificial intelligence , mathematics , control (management) , physics , geometry , classical mechanics , programming language
SUMMARY At signalized intersections, a driver is forced to decide whether to stop or pass through when the signal turns yellow from green. It has been pointed out that this dilemma sometimes leads to accidents at intersections. This paper proposes two methods based on vehicle behavior for predicting whether a driver approaching a signalized intersection decides to stop or pass through the intersection when the signal turns yellow, in order to support the drivers' decision‐making to prevent accidents at and near intersections. The first method predicts whether a driver will decide to stop when the signal turns yellow, based on the idea that whether a driver decides to stop should depend on the situation and the vehicle behavior at the moment the signal turns yellow. The second method estimates whether a driver has decided to stop, in other words, whether the driver has the intention to stop, based on the idea that the driver's intention will appear in the vehicle behavior when the driver has decided to stop at the intersection. This method can detect the driver's oversight or neglect of traffic signals on the way to the intersection. The effectiveness of these two methods is demonstrated with experiments using real‐world data.