A new behavioral model for microscopic freeway traffic-flow simulation is presented. As a key feature, the model is subdivided into small modules, each dealing with one distinct problem or intention of the current driving task. The modules operate independently of each other. Each module provides a situation-based vote for the longitudinal and the lateral movement of the vehicle. Subsequently, the individual votes are combined by applying a procedure that can be described as a virtual fuzzification. The model also incorporates the chronological sequence of the individual votes; decisions depend not only on the current evaluation of a driving task but also on the trend of prior evaluations. Thus, it emulates actual driver behavior. The behavioral model was implemented into the simulation software BABSIM and field tested. It is shown that the model is valid for various applications ranging from rather simple freeway sections over a lane reduction up to complex multilane on ramps and off ramps. Thus, the new approach is a valuable contribution to freeway traffic-flow modeling and simulation. It provides more accurate results and a more flexible use of the simulation software. At the same time, the straightforward and concise structure of the model simplifies its maintenance and ensures the expandability of the concept. All this promotes further use of traffic-flow simulation and leads the way for promising future developments.
New Behavioral Model for Microscopic Freeway Traffic-Flow Simulation
Transportation Research Record
Transportation Research Record: Journal of the Transportation Research Board ; 2088 , 1 ; 10-17
2008-01-01
Article (Journal)
Electronic Resource
English
New Behavioral Model for Microscopic Freeway Traffic-Flow Simulation
Online Contents | 2008
|Numerical Simulation of Freeway Traffic Flow
British Library Online Contents | 1997
|Numerical Simulation of Freeway Traffic Flow
Online Contents | 1997
|