development an intelligent system for discretionary lane change behaviors in the real traffic flow

Lane changing behaviors, due to the hidden aspects of factors involved, are considered as one of the most complicated behaviors in traffic flows. This paper proposes an intelligent system for discretionary lane change behaviors in the real traffic flow. The introduction presents a review of the research into the models of lane change behaviors. Then, a new method is proposed to determine the initial and final points of lane changing behavior in terms of time parameter. Hypotheses are then introduced to extract data on lane change behaviors from data recorded by NGSIM. An intelligent system is put forward using two control sub-systems based on fuzzy logic. Basically, this paper brings about innovative features by the utilization of time parameter during the execution of lane change behaviors and by the consideration of the impact of human and environmental factors through measurable data such as relative distance and relative speed in terms of time parameter. Finally, the results of the proposed intelligent system are compared with the behaviors of the driver in real situations. The framework of the result analysis employs error criteria and standard deviation (Variance). The said comparison indicates that the developed intelligent system outperforms the human behaviors in real situations.