Development of an accident prediction model for freeways systems

摘要

This paper presents a model using log-logistics accelerated failure time approach based on the traffic accident records collected in the Ministry of Road Transport and Highways, Government of India in 2018-2019. This research work provides a series of approaches for identifying different critical elements to model accident prediction and clearance timing in freeways systems. Specifically, the innovative solutions are aimed at not only developing a prediction model but also analysing severity indicators like property damage, number of injuries, fatalities, etc. At the process level, the models are concerned with different variables, which play vital roles in real-life scenarios. The proposed model also keeps track of different random variables and is able to forecast possible accidents that may save several lives. Under influence of those random keys, this analysis also provides few suggestive proposals to take necessary precautions to reduce accident impacts and enhance traffic safety and the accident management process.