Developing accident prediction model for railway level crossings

•A sound model for predicting train-car collisions at LXs is developed.•The model quality is validated through thorough statistical means.•The predictive accuracy of distributions combined with the prediction model is fine. Railway level crossing (LX) safety continues to be one of the most critical...

Full description

Saved in:
Bibliographic Details
Published in:Safety science 2018-01, Vol.101, p.48-59
Main Authors: Liang, Ci, Ghazel, Mohamed, Cazier, Olivier, El-Koursi, El-Miloudi
Format: Article
Language:English
Subjects:
Accident prediction
Accident prediction modeling
Accidents
Data processing
Least squares
Level crossing safety
Level crossings
Prediction models
Railroad accidents & safety
Railroad crossings
Railroads
Railway accidents
Railway networks
Risk assessment
Statistical analysis
Statistics
Traffic
Traffic models
Train-car collision
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•A sound model for predicting train-car collisions at LXs is developed.•The model quality is validated through thorough statistical means.•The predictive accuracy of distributions combined with the prediction model is fine. Railway level crossing (LX) safety continues to be one of the most critical issues for railways, despite an ever-increasing focus on improving design and application practices. Accidents at European LXs account for about one-third of the entire railway accidents and result in more than 300 deaths every year in Europe. Due to the non-deterministic causes, the complex operation background and the lack of thorough statistical analysis based on accident/incident data, the risk assessment of LXs remains a challenging task. In the present paper, some LX accident prediction models are developed. Such models allow for highlighting the influence of the main impacting parameters, i.e., the average daily road traffic, the average daily railway traffic, the annual road accidents, the vertical road profile, the horizontal road alignment, the road width, the crossing length, the railway speed limit and the geographic region. The Ordinary Least-Squares (OLS) and Nonlinear Least-Squares (NLS) methods are employed to estimate the respective coefficients of variables in the prediction models, based on the LX accident/incident data. The validation and comparison process is performed through statistical means to examine how well the estimation of the models fits the reality. The outcomes of validation and comparison attest that the improved accident prediction model has statistic-based approbatory quality. Moreover, the improved accident prediction model combined with the NB distribution shows relatively high predictive accuracy of the probability of accident occurrence.
ISSN:0925-7535
1879-1042
DOI:10.1016/j.ssci.2017.08.013