Construction of Key Factor Model of Car Side-Impact Occupant Injury Based on Nais Deep Investigation of Traffic Accident Data

摘要

The scale of traffic accident data collected by NAIS(National Automobile Accident In-Depth Investigation System) is accumulating. After updating version 2.0, the data structure and related field parameters are optimized again, which can realize multi-angle and high-degree-of-freedom application of data. In this paper, the accident data in NAIS database are used to select a batch of car side-impact accident data with high data integrity and good quality, and try to construct the key factors model of passenger side-impact accident, which can extract the key parameters of side-impact accident. Based on the optimized screening results of many accident data, this study uses logistic regression mathematical method to analyze the correlation between various accident-related factors and human injury variables. From the point of view of the established model, the significant factors affecting whether the occupant in side crash is seriously injured mainly include five factors: the type of the object vehicle, the position of the far and near end of the occupant collision, the effective collision speed, the collision angle and the tangential velocity of the collision. Through the analysis and evaluation of the model, we can see that the key parameters are reasonable and the prediction accuracy index of the model is good. The results show that the regression parameters of the data logistic regression model are matched with the analysis of accident characteristics, and the model itself is accurate and effective. This shows that the quality and quantity of data in NAIS database can support such in-depth research.