Evaluation of a behavioral cellular-based traffic model

A cellular-based model that improves cell-based models’ ability to replicate actual traffic flow on high speed highways is proposed in this study. Behavior variations were introduced in the model by incorporating both driver aggression and responsiveness to the action of the leading vehicle. The deceleration and acceleration of the following vehicles were based on their speed and distances relative to leading vehicles. The evaluation of the model performance indicated that the proposed model was able to reproduce traffic conditions observed in the field at an acceptable level of accuracy as a MAPE statistic of 9.9% was obtained. A low (0.002) Theil’s inequality coefficient also confirmed the validity of the model specifications. Correlations of average speeds and traffic flow from the simulation and actual traffic data verified model’s output match real world data. Further examination of individual vehicle trajectories from simulation results showed that the model can reproduce backward forming traffic jams consistent with empirical observations.