An Exploration Of Car Following Model Under Emergency Evacuation Situation

In recent years, with the number of all kinds of emergencies at home and abroad have been gradually rising, governments at all levels have stepped up the investment on emergencies. Besides making early warning to emergencies from the aspect of prevention, it is necessary to focus on studying on emergency evacuation plan and establishing scientific emergency evacuation system. At present, in the environment of emergency evacuation, the quantitative analysis of ecacuation indicators still have been conducted based on normal driver behavior characteristics. However, affected by the emergency evacuation management measures, drivers unavoidably have the feeling of tension with the change on psychological and physiological indicators from the normal condition. The emotion of tension has an influence on driver's decision and behavior to a certain degree which will also affect the whole traffic flow characteristics of evacuation vehicles.In this paper, the methodology of scenario simulation is adopted with using simulator for the purpose of establishing car-following model under emergency evacuation situation. A real scenario of emergency evacuation is provided for drivers through playing video clips about incidents and evacuation warnings. In the whole process of experiment, the heart rate of driver and the data of car-following driving are recorded to provide data base for establishing car-following model. The contents include the following aspects:(1) The emergency characteristics of drivers were analyzed from two aspects.The questionnaire with evaluation on the degree of tension was provided for drivers to fill in from the aspect of subjectivity. The heart rate of driver was also recorded to evaluate the tension from the aspect of objectivity.(2) Perception reaction time of driver was devided into three types: the reaction time under incidents, the one in the condition of acceleration and the one in deceleration. They were analyzed in terms of statistical features and influence factors so that the change rule of perception reaction time can be summarized.(3) Through making an analysis on the relationship between the headway and relative speed, the critical headway of following condition under emergency evacuation situation can be determined. The regression models of spacing in both acceleration and deceleration condition were established respectively.(4) Based on quadric response surface analysis, the multivariate nonlinear regression of spacing was conducted in the software of SAS. It can be concluded that the safe spacing model under emergency ecacuation situation was established. Through model validation, this calibrated model of safe spacing is in a good agreement with the measured data. Therefore, the safe spacing model based on quadric response surface analysis has a better reflect on the characteristics of following vehicles under emergency evacuation situation. This conclusion can contribute to quantitative calculation on evacuation indicators and providing a guide for calibrating parameters about driving behavior in simulation softwares.