关键词: 交通工程, 累积Logistic模型, 青藏高原, 事故严重程度预测

Abstract: Based on the analysis of traffic accident data of Qinghai-Tibetan plateau, it is found that most of the traffic accidents are of high casualties, which means the proportion of accidents with deaths and injuries is relatively larger. The cumulative Logistic regression was used to research the factors which affected the severity of traffic accidents in plateau area. The collected accident severity was selected as the dependent variable. The altitude of the environment, the oxygen content, the weather, the linear and environment of the accident site, the type of the vehicle, the traffic volume, carts ratio, the speed difference between cart and car were selected as independent variables. Therefore, the cumulative Logistic regression model was established. The results show that five independent variables including altitude, carts ratio, the vehicle type and the speed difference between the cart and the car are significantly correlated with the severity of accident. This prediction model can provide theoretical basis for the related highway management departments to make safety measures.

Key words: traffic engineering, cumulative Logistic model, Qinghai-Tibet plateau, accident severity prediction