Modeling of car-following behavior at signalized intersection considering vehicle-vehicle and vehicle-road interactions

To analyze the effects of vehicle-vehicle and vehicle-road interactions on traffic flow at signalized intersections, this study proposes an improved car-following model based on the Full Velocity Difference (FVD) model, which considers the impact of lane width and the percentage of heavy vehicles on car-following behavior. The longitudinal space headway in the base FVD model was transformed into a comprehensive space headway with lateral and longitudinal weighted spacing to account for the vehicle-road interaction effect. The model was also segmented into four scenarios (car-car, car-bus, bus-car, and bus-bus) to account for the vehicle-vehicle interaction effect. The performance of the proposed model was validated using measured trajectory data, showing improvements in the average Root Mean Square Error (RMSE) by up to 31 % and Relative Entropy (RE) by up to 63 % compared with the base FVD model. Similarly, compared to the OV-Rong model, RMSE improved by an average of 84 % and RE improved by an average of 63 %; compared to the FVD-Liu model, RMSE improved by an average of 30 % and RE improved by an average of 52 %. Numerical simulations across 24 scenarios with different lane widths and varying percentages of heavy vehicles validated the precision of the model in capturing traffic flow characteristics. The results suggest that the improved model can accurately analyze macroscopic capacity changes based on microscopic car-following behavior under different traffic compositions and lane widths.