Search

In this paper, a data-driven two-lane traffic flow model based on cellular automata is proposed. Long Short-Term Memory (LSTM) and Support Vector Machine (SVM) are used to learn the characteristics of car following behavior and lane changing behavior, respectively, from real operation data of vehicles. Under optimal network parameters, the mean absolute errors of the LSTM network for training and testing data are only 0.001 and 0.006, respectively; while the prediction accuracy of the SVM classifier for both data reaches higher than 0.99. Moreover, forward rules and lane changing rules which are more consistent with actual situation are designed. The simulation results show that: (1) the new model can reflect the first-order phase transition from free flow to synchronized flow; (2) the frequency of unsuccessful lane changing is near zero in low-density traffic areas, but increases sharply in high-density regions; and (3) the lane changing duration and unsuccessful lane changing frequency display similar trends as traffic densities increase.

In this paper, a two-lane cellular automata traffic flow model based on regular hexagonal cells is proposed. The realistic lane changing maneuvers with the lateral movement process, the lane changing duration, and the unsuccessful lane changing are considered by the new model. The vehicle trajectories from NGSIM dataset are used to extract the realistic lane changing maneuvers and calibrate the lane changing angle. The numerical results showed that the flow in the intermediate density range of the new model is lower than that of the traditional model. While the lane changing frequency in the same density range of the new model is higher than that of the traditional model. The lane changing duration will increase as the density grows. And the unsuccessful lane changing rate grows to large values in high density range. As to the heterogeneous traffic flow, the lane changing frequency is obviously greater than that in homogeneous condition. The trucks are more likely to make unsuccessful lane changing. And the lane changing duration of trucks is longer than that of cars in free flow region, but this is not the case in congested flow.

An alternative floor field (FF) model is proposed by incorporating the agitated behavior and elastic characteristics of pedestrians. The agitated behavior which is regarded as an important factor to pedestrian dynamics is depicted by introducing a parameter to revise the transition probability of pedestrians to move to the neighboring cells. To characterize elasticity of pedestrians, it is assumed that a cell can hold more than one pedestrians in crowd condition, while it can hold only one pedestrian in normal condition. In addition, a method to deal with conflicts is employed by considering the effects of agitated behavior and desired velocity. Numerical simulations are carried out to investigate pedestrian evacuation from a room. The results show, that as the value of agitated parameter increases, the evacuation time decreases to the minimum value and then increases gradually. Also, the faster-is-slower effect which is obtained by some other simulation models can be reproduced by the proposed model. Finally, the influence of exit width and the corresponding mechanism on evacuation process is investigated which is expected to be helpful to the exit design of public rooms.

In this paper, a new two-dimensional car-following model is proposed to depict the features of mixed traffic flow consisting of motorized vehicles ( m -vehicle) and non-motorized vehicles ( n m -vehicle), based on the two-dimensional optimal velocity (OV) model by Nakayama et al. [A. Nakayama, K. Hasebe, Y. Sugiyama, Phys. Rev. E 71 (2005) 036121]. In the proposed model, velocity difference terms are introduced, which are regarded as important factors for traffic behavior. Numerical simulations are carried out to investigate the interaction between left-turning n m -vehicle flow and straight-going m -vehicle flow at a typical unsignalized interaction. The results show that the straight-going m -vehicle flow just next to n m -lane is disturbed more seriously than others. In addition, a well-known phenomenon in reality is observed that groups of m -vehicles and n m -vehicles pass through the intersection alternately.