Models to simulate individual driver behavior offer the possibility to investigate human-machine interaction in early stages of driver assistance system development. Many driver behavior models were published with regard to the simulation of longitudinal and lateral vehicle control, but the number of detailed models which simulate gap acceptance behavior preceeding a lane change (overtaking) maneuver is comparably small.
In this thesis, the influence of different rear view mirror types on driver's gap acceptance behavior during a simulated lane change scenario on a two-lane German Autobahn was investigated and a driver behavior model was built to simulate this behavior. Individual behavior differences were also considered.
As a first step, a simulation of different rear view mirror types (Planar, Convex C20 / C14) was implemented in a research driving simulator. Afterwards, two driving simulator studies were conducted: 1) To validate the implementation of the three mirror types, the results of a distance estimation study were compared against already published field studies with real mirror. 2) Results of a lane change study were then used for the development of a driver behavior model to simulate gap acceptance behavior: to estimate gap size and closing speed of approaching vehicles from behind, the model relies on visual angles and their rate of change which are well known concepts in psychology of visual perception.