Identifying Features of Traffic Hysteresis on Freeways

Delay in recovering the speed of the vehicle results in occurrence of the hysteresis phenomenon in traffic flow the disturbance. The fundamental theories to analyze the hysteresis phenomenon based on driver behavior asymmetry during acceleration, and deceleration phases are open research area and need more developments. In this paper, which is the review of recent efforts in understanding hysteresis; using a Newell`s car following model, are reviewed to study the effects of different parameters on the magnitude of the motioned phenomenon. Moreover, the hysteresis phenomenon based on aggressive and timid driver behavior was identified. Microscopic sensitivity analysis and its results are shown by considering a conducted research based on the modeling of the phenomenon by neural network, and optimizing hidden layers of the neural network using genetic algorithm. Furthermore, in non-stationary traffic flow by using the kinematic wave model in variable wave speed; the magnitude of the hysteresis was estimated by a gradual analysis of the speed- density relationship. The results showed that the shape of traffic hysteresis loops depend onthe driving behavior. Also, Spacing and acceleration at the end point of the phenomenon are considered as the most effective parameters in the sensitivity analysis related to driver behavior. Last but not least, it was found that the hysteresis magnitude in the non-stationary condition resulted in the lowest amount.