Injury-Severity Analysis of Highway-Railroad Grade Crossing Crashes Based on a Random Parameters Logit Model With Heterogeneity in Means and Variances

Traffic accidents involving trucks at highway-railroad grade crossings (HRGCs) often result in catastrophic outcomes. It is crucial for transportation agencies to identify the underlying causes and develop strategies to mitigate the severity of these accidents. This study identifies the significant...

Full description

Saved in:
Bibliographic Details
Main Authors: Lan Wu, Jiayu Zhou, Haigen Min, Gen Li
Format: Article
Language:English
Published: Wiley 2025-01-01
Series:Advances in Civil Engineering
Online Access:http://dx.doi.org/10.1155/adce/6682353
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850171524515889152
author Lan Wu
Jiayu Zhou
Haigen Min
Gen Li
author_facet Lan Wu
Jiayu Zhou
Haigen Min
Gen Li
author_sort Lan Wu
collection DOAJ
description Traffic accidents involving trucks at highway-railroad grade crossings (HRGCs) often result in catastrophic outcomes. It is crucial for transportation agencies to identify the underlying causes and develop strategies to mitigate the severity of these accidents. This study identifies the significant factors contributing to the crash injury severity involving truck and automobile drivers at HRGCs. A comprehensive dataset of 14,415 crashes collected by the Federal Railroad Administration (FRA) between 2012 and 2021 was analyzed. The multinomial logit (MNL) model, the random parameters logit model, and the random parameters logit model with heterogeneity in means and variances were constructed, and the one with the best fit and accuracy to calculate the average marginal effect of each variable, thereby quantifying their impact on injury severity at HRGCs. The findings demonstrate that the random parameters logit model with heterogeneity in means and variances enhanced the fitting and accuracy of the model over both the standard one without considering heterogeneity and the binary logit model. This improved model better captures and accounts for unobserved heterogeneity in the data. There are clear distinctions between auto and truck crashes at HRGCs, specifically regarding the factors that impact crash severity and how those factors affect the severity level of crashes. The research highlights that these divergences primarily arise from dissimilarities in the characteristics of drivers and vehicles between autos and trucks. For instance, the injury severity for auto drivers can be significantly affected by features such as stopping and then proceeding and stopping at the crossing due to malfunctions. On the other hand, factors like passive control, open spaces, the presence of barriers, and warning devices emerge as crucial factors influencing truck driver injuries. This study proposes safety improvements tailored to the specific needs of each type of crash to improve HRGC safety and decrease accident severity.
format Article
id doaj-art-717c79cfd6f2469b95b923f6ee3e3a59
institution OA Journals
issn 1687-8094
language English
publishDate 2025-01-01
publisher Wiley
record_format Article
series Advances in Civil Engineering
spelling doaj-art-717c79cfd6f2469b95b923f6ee3e3a592025-08-20T02:20:16ZengWileyAdvances in Civil Engineering1687-80942025-01-01202510.1155/adce/6682353Injury-Severity Analysis of Highway-Railroad Grade Crossing Crashes Based on a Random Parameters Logit Model With Heterogeneity in Means and VariancesLan Wu0Jiayu Zhou1Haigen Min2Gen Li3College of Automobile and Traffic EngineeringCollege of Automobile and Traffic EngineeringSchool of Information EngineeringCollege of Automobile and Traffic EngineeringTraffic accidents involving trucks at highway-railroad grade crossings (HRGCs) often result in catastrophic outcomes. It is crucial for transportation agencies to identify the underlying causes and develop strategies to mitigate the severity of these accidents. This study identifies the significant factors contributing to the crash injury severity involving truck and automobile drivers at HRGCs. A comprehensive dataset of 14,415 crashes collected by the Federal Railroad Administration (FRA) between 2012 and 2021 was analyzed. The multinomial logit (MNL) model, the random parameters logit model, and the random parameters logit model with heterogeneity in means and variances were constructed, and the one with the best fit and accuracy to calculate the average marginal effect of each variable, thereby quantifying their impact on injury severity at HRGCs. The findings demonstrate that the random parameters logit model with heterogeneity in means and variances enhanced the fitting and accuracy of the model over both the standard one without considering heterogeneity and the binary logit model. This improved model better captures and accounts for unobserved heterogeneity in the data. There are clear distinctions between auto and truck crashes at HRGCs, specifically regarding the factors that impact crash severity and how those factors affect the severity level of crashes. The research highlights that these divergences primarily arise from dissimilarities in the characteristics of drivers and vehicles between autos and trucks. For instance, the injury severity for auto drivers can be significantly affected by features such as stopping and then proceeding and stopping at the crossing due to malfunctions. On the other hand, factors like passive control, open spaces, the presence of barriers, and warning devices emerge as crucial factors influencing truck driver injuries. This study proposes safety improvements tailored to the specific needs of each type of crash to improve HRGC safety and decrease accident severity.http://dx.doi.org/10.1155/adce/6682353
spellingShingle Lan Wu
Jiayu Zhou
Haigen Min
Gen Li
Injury-Severity Analysis of Highway-Railroad Grade Crossing Crashes Based on a Random Parameters Logit Model With Heterogeneity in Means and Variances
Advances in Civil Engineering
title Injury-Severity Analysis of Highway-Railroad Grade Crossing Crashes Based on a Random Parameters Logit Model With Heterogeneity in Means and Variances
title_full Injury-Severity Analysis of Highway-Railroad Grade Crossing Crashes Based on a Random Parameters Logit Model With Heterogeneity in Means and Variances
title_fullStr Injury-Severity Analysis of Highway-Railroad Grade Crossing Crashes Based on a Random Parameters Logit Model With Heterogeneity in Means and Variances
title_full_unstemmed Injury-Severity Analysis of Highway-Railroad Grade Crossing Crashes Based on a Random Parameters Logit Model With Heterogeneity in Means and Variances
title_short Injury-Severity Analysis of Highway-Railroad Grade Crossing Crashes Based on a Random Parameters Logit Model With Heterogeneity in Means and Variances
title_sort injury severity analysis of highway railroad grade crossing crashes based on a random parameters logit model with heterogeneity in means and variances
url http://dx.doi.org/10.1155/adce/6682353
work_keys_str_mv AT lanwu injuryseverityanalysisofhighwayrailroadgradecrossingcrashesbasedonarandomparameterslogitmodelwithheterogeneityinmeansandvariances
AT jiayuzhou injuryseverityanalysisofhighwayrailroadgradecrossingcrashesbasedonarandomparameterslogitmodelwithheterogeneityinmeansandvariances
AT haigenmin injuryseverityanalysisofhighwayrailroadgradecrossingcrashesbasedonarandomparameterslogitmodelwithheterogeneityinmeansandvariances
AT genli injuryseverityanalysisofhighwayrailroadgradecrossingcrashesbasedonarandomparameterslogitmodelwithheterogeneityinmeansandvariances