Method of Predicting Slowing Down Behavior of Preceding Vehicle at Intersection

Various types of information about traffic streams, congestion, safety, environmental conditions, and the like can be obtained by analyzing data from vehicles used under a wide range of conditions. A method for predicting the slowing down behavior of vehicles from this information would be extremely useful for safety and environmental applications. Therefore, this research started to examine the feasibility of using statistical methods to predict the slowing down behavior of a preceding vehicle. As one aspect of this research, this paper studies the feasibility of predicting whether the preceding vehicle will slow down to turn left at an intersection using statistical significance found in data of different driving behavior, based on the velocity and acceleration of vehicles that slowed down or did not slow down at the same location in the past.