Cluster-size dependent randomization traffic flow model

In order to exhibit the meta-stable states, several slow-to-start rules have been investigated as modification to Nagel-Schreckenberg (NS) model. These models can reproduce some realistic phenomena which are absent in the original NS model. But in these models, the size of cluster is still not considered as a useful parameter. In real traffic, the slow-to-start motion of a standing vehicle often depends on the degree of congestion which can be measured by the clusters' size. According to this idea, we propose a cluster-size dependent slow-to-start model based on the speed-dependent slow-to-start rule (VDR) model. It gives expected results through simulations. Comparing with the VDR model, our new model has a better traffic efficiency and shows richer complex characters.