Driving behaviors at intersection are complex because drivers have to perceive more traffic events than normal road driving and thus are exposed to more errors with safety consequences. Drivers make real-time responsesin a stochastic manner. This paper presents our study using Hidden Markov Models (HMM) to model driving behaviors at intersections. Observed vehicle movement data are used to build up the model. A single HMM is used to cluster the vehicle movements when they are close to intersection. The re-estimated clustered HMMs provide better prediction of the vehicle movements compared to traditional car-following models. Only through vehicles on major roads are considered in this paper.
Zou, Xi and David Levinson (2006) Modeling Pipeline Driving Behaviors: A Hidden Markov Model Approach. Journal of the Transportation Research Board: Transportation Research Record 1980 16-23.
Zou, Xi; Levinson, David M.
Modeling Pipeline Driving Behaviors: A Hidden Markov Model Approach.
Transportation Research Board.
Retrieved from the University of Minnesota Digital Conservancy,
Content distributed via the University of Minnesota's Digital Conservancy may be subject to additional license and use restrictions applied by the depositor.