An improved car-following model considering the influence of multiple preceding vehicles in the same and two adjacent lanes

To investigate the driver's car-following behavior in multiple lanes, this study develops an improved multi-lane full velocity difference (FVD) model. This model considers the positions and widths of multiple preceding vehicles in the same lane in the longitudinal direction, as well as the lateral offset distances of the immediately preceding vehicles in the same and two adjacent lanes in the lateral direction. Based on the linear stability analysis, it is found that an increase in the lateral offset distances between a vehicle and its predecessors in the same lane is negative to the traffic flow stability, whereas an increase in the widths of front vehicles in the same lane within a certain range is beneficial to the traffic flow stability. These theoretical findings are verified by numerical simulations. With the use of a publicly available dataset of vehicle trajectories from Southeast University, this study extracts the microscopic traffic flow parameters of the following vehicles and their associated vehicles on urban roads in China. Through parameter calibration and model comparison, it is found that the developed improved multi-lane FVD model has greater adaptability than the conventional FVD model and existing improved FVD models, and its mean absolute error and root mean square error are to some extent optimized. With the rapid development of connected vehicle technology, the developed multi-lane FVD model can be used to assist drivers in making rational behavioral decisions and improve traffic flow stability.