Parameter calibration of micro traffic simulation model is the precondition of the application of micro traffic simulation technology. This paper is based on the calibration of parameters of existing simulation models, refining the parameter classification of micro-simulation model, a new calibration method of micro simulation model parameters is proposed, which combines engineering measurement with intelligent optimization. Selecting the signalized intersection and based on VISSIM simulation experiment platform, the genetic algorithm is applied to calibrate the model parameters, and the existing methods and the new method are compared and analyzed. The experiment results show that compared with the real traffic data, the proposed method and the existing methods can achieve a lower error level in the output of the model, but the proposed method has obvious explanatory power in the actual physical meaning of the parameters, and the vehicle running characteristics in the simulation scenario are more in line with the actual condition.

CICTP