Research of the left-turn vehicles lane-changing behaviors at signalized intersections with contraflow lane

The design of contraflow left-turn (CL) lane has effectively improved the capacity of left-turn vehicles at signalized intersections and has been widely used. Existing studies have focused more on the coordination and optimization of the main signal and pre-signal as well as capacity, delay time and queue dissipation. However, the vehicles lane-changing behaviors for CL lane design have not been adequately studied. Therefore, it is necessary to study of left-turn vehicles lane-changing behaviors at signalized intersections with contraflow lane and the traffic evolution law of the road. In this study, a cellular automaton (CA) model that simulates left-turn traffic flow at signalized intersections with CL lane was developed and verified by field data. Various factors are considered, including the length of CL lane and lane-changing area, as well as the location of the vehicles. The simulation results show that the density distribution of traditional left-turn (TL) lane is more influenced by the length of the lane-changing area when the probability of vehicle injection is small (alpha <= 0.1). Furthermore, when the probability of vehicle injection is larger (alpha > 0.1), the density distribution of CL lanes is more influenced by length of the lane-changing area. In addition, the departure ratios of CL and TL lanes is close to 50% when the probability of vehicles injection and length of the lane-changing area are larger (alpha > 0.1and LB >= 20cells), which is almost consistent with the peak hour vehicle departure ratios in the real scenario. Finally, this study can provide a reference for the actual application of CL lane design.