Abstract: Scenario simulation is employed to create a driving environment in the emergency evacuation and the questionnaire investigation and the Miriam ECG are used to verify the validity of the driving environment in the emergency evacuation from subjective and objective aspects.Gray-correlation Analysis is undertaken to determine which factors have more impact on the acceleration of the following car.Acceleration of the following car is set as the mother factor series and the speed of the following car,speed of the front car,speed difference between vehicles,headway between vehicles,acceleration of the front car are set as the sub-factor series.Finally acceleration of the front car,speed difference between vehicles,speed of the following car are determined as three most important factors.BP neural network is designed and the three factors mentioned above are selected as input variables and the acceleration of the following car is selected as output variable.Simulation of the BP neural network with data gathered from driving simulator reveals that BP neural network has a high precision in the prediction of the car-following model.