Traffic paradox on a road segment based on a cellular automaton: Impact of lane-changing behavior

•Novel cellular automaton is used considering macroscopic characteristics of traffic flow.•The results are universally applicable for unsignalized and signalized conditions.•Figures for dual functions of dependent variables are obtained.•The critical boundaries of the paradox is located.

The traffic paradox “faster is slower” does not always apply. To study when and where it is valid, a simulation for a real road segment is performed using a novel cellular automaton. This simulation is used to analyze the change in global traffic flow status during free lane-changing behavior under general urban traffic conditions. The impact of lane-changing behavior is quantified into two aspects, time and space, and are described by average delay and transitable flow, respectively. Then surfaces are obtained, which adopt the arriving probability of vehicles and the green ratio as dual independent variables. Thus by the comparison of two surfaces, free lane-changing and straight proceeding, the horizontal projection of the intersecting lines is solved. Finally, the range of occurrence and reasons for the paradox are analyzed.