Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Design of a strategy to change lanes in dense traffic for (partially) automated cars based on human road behavior


S. Kammerer

Master Thesis, FS16

Increasing automation in today’s vehicles leads to a situation on the street where a mixture of autonomous vehicles and vehicles, that are manually driven, are coexistent. To cope with this situation, autonomous vehicles need to be accepted positively by surrounding vehicles. Using a cooperative driving strategy, it is possible to achieve the desired acceptance. Additionally, using cooperative driving strategies, the traffic flow can be increased as well as the positive perception of the autonomous vehicle as experienced by surrounding vehicles will be enhanced. Lane change maneuvers in dense and slow moving traffic on two lane streets are chosen in order to explore the advantage of the advanced algorithm developed in this work. Its behavior is perceived as a distinct sign that the vehicle intends to change lanes, as well as being cooperative during the lane change as judged by the following traffic participant. It is studied how a clear announcement of a desired lane change wish is required. Additionally, the resulting perception of the autonomous car as judged by surrounding traffic participants is studied, using subjective answers of probands and objective drive data originating from driving simulator studies. Therefore, the presented lane change algorithm is based on two proband studies illuminating the lane change from two perspectives. The first is from the lane changing vehicle and the second from the reacting vehicle on the new lane. The resulting strategy comprises of how to announce and execute a lane change where the announcement of the lane change is perceived more distinct and judged more cooperative regarding its execution than using state of the art behavior. The implementation is twofold, first a logistic regression as classification is used to determine, if the adjacent gap is appropriate to initiate a cooperative merging process or not. Second, optimal values from the basic studies are used to perform the announcement and execution of the decided lane change. These two steps are validated separately, the classification using accuracy and confusion matrix measurements, the performed action by conducting a proband study in the driving simulator with n=29 probands.

Supervisors: Nina Kauffmann (BMW AG), Nikolaos Kariotoglou, John Lygeros


Type of Publication:

(12)Diploma/Master Thesis

File Download:

Request a copy of this publication.
(Uses JavaScript)
% Autogenerated BibTeX entry
@PhdThesis { Xxx:2017:IFA_5634
Permanent link