ABSTRACT
Microscopic traffic simulation models can represent individual vehicles and account for the correlative attributes, among which the cellular automata (CA) model can simulate and explain complex traffic phenomena with high simulation efficiency by defining a series of simple rules. However, at present, there are not enough reasonable multi-lane cellular automata models accompanied by the proper calibration methods for the simulation and calibration simultaneously of expressway traffic flow. Therefore, this paper develops a multi-lane cellular automata model from the microscopic aspect, and meanwhile uses the simulation-based optimization (SO) method to calibrate the parameter of the model. For verification, the vehicle data of a super bridge and a segment of an expressway are simulated, and the related parameter is calibrated. The results show that the model can simulate and provide a reasonable value of the probability of vehicles’ random slowdown behavior to better reflect the traffic situation.
