Analysis of Truck Accident Severity in the Seoul Metropolitan Area Based on EPDO Sets

doi:10.15683/kosdi.2025.9.30.782

All Issue

2025 Vol.21, Issue 3 Preview Page Next Page

Original Article

Analysis of Truck Accident Severity in the Seoul Metropolitan Area Based on EPDO Sets EPDO 가중치에 따른 수도권 화물차 사고 심각도 분석에 관한 연구: 박민서¹ㆍ윤병조²*
Min Seo Park¹, Byoung Jo Yoon²*; ¹Master’s student, College of Urban Science, Incheon National University, Incheon, Republic of Korea
²Professor, College of Urban Science, Incheon National University, Incheon, Republic of Korea

30 September 2025. pp. 782-790

PDF

Abstract

Purpose: This study aims to analyze the severity of truck-related crashes and the factors that influence them. In particular, the study aims to compare the differences in crash severity based on two sets of Equivalent Property Damage Only (EPDO) weights, providing fundamental data for establishing safety measures tailored to the characteristics of truck crashes. Method: The dependent variables were divided into two datasets based on the differences between the two EPDO weight sets, 12:6:3:1 and 3:3:3:1. The optimal model was selected using the Light Gradient Boosting Machine (LGBM), and the differences in the performance evaluation metrics of the two datasets were analyzed. Furthermore, the variable importance of each model was presented using feature importance analysis. Result: The analysis of model performance revealed that the crash severity prediction error was lower in EPDO_B compared to EPDO_A, resulting in more reliable outcomes. Additionally, the key influencing factors on crash severity were identified and compared, confirming that predicting human casualties is relatively easier than predicting crash severity. Conclusion: Through the analysis of truck crash severity, this study provides insights into future traffic safety policies and infrastructure development. In particular, the study emphasizes the necessity of developing traffic safety countermeasures that focus on predicting human casualties and highlights that more precise data collection and analysis are essential for effective crash prevention.

Keywords

Truck

Accident Severity

EPDO

Machine Learning

LightGBM

Feature Importance

연구목적: 본 연구는 화물차 교통사고의 심각도와 영향을 미치는 요인을 분석하기 위해 수행되었다. 특히 EPDO(Equivalent Property Damage Only) 가중치 세트에 따른 사고 심각도 영향요인의 차이를 비교하였다. 개선된 EPDO 가중치에 따른 심각도 영향요인의 변화와 예측 정확도를 분석하여 화물차 사고의 원인에 따른 안전 대책 수립을 위한 기초자료를 제공하는 것을 목표로 한다. 연구방법: 종속변수를 12:6:3:1, 3:3:3:1의 두 가지 EPDO 가중치 세트의 차이에 따라 두 가지 데이터 세트로 구성하였으며 Light Gradient Boosting Machine을 기반하여 최적의 모델을 선정하여 두 데이터 세트의 모형 성능 평가 지표 결과 값의 차이를 분석하였다. 또한, 각 모델의 변수 중요도를 Feature importance를 통해 나타냈다. 연구결과: 모델 성능 분석 결과, EPDO_A보다 EPDO_B에서 사고 심각도 예측 오차가 낮아 보다 신뢰할 수 있는 결과를 제공하였다. 또한, 사고 심각도에 영향을 미치는 주요 요인을 비교하여 제시하였으며, 사고의 경중보다는 인적 피해 예측이 상대적으로 용이함을 확인하였다. 결론: 화물차 사고의 심각도 분석을 통해 향후 교통안전 정책 및 인프라 구축의 방향을 제시하였다. 특히, 인적 피해 예측을 중심으로 한 교통안전 전략 수립이 필요하며, 보다 정교한 데이터 수집과 분석이 사고 예방의 핵심이 될 수 있음을 강조하였다.

키워드

화물차

사고 심각도

EPDO

머신러닝

LightGBM

Feature importance

References

Korea Transport Database (2022). 2022 Nationwide Freight Data Update. Korea Transport Institute, KTDB-2022-506.
Korea Transport Institute (2022). Volume 6: Nationwide Freight O/D Supplement Update. Final Report of the 2022 National Transport Survey, Ministry of Land, Infrastructure and Transport.
Kwon, C.-W., Chang, H.-H. (2021). “Comparative analysis of traffic accident severity of two‑wheeled vehicles using XGBoost.” Journal of Information Technology Services, Vol. 20, No. 4, pp. 1-12. 10.12815/kits.2021.20.4.1
Lee, G.-J. (2023). A Study on the Analysis of Factors Influencing the Severity of Traffic Accident in Two‑Wheel Vehicles Using Machine Learning–Focus on Incheon Metropolitan City. Master Thesis, Incheon National University.
Ministry of Land, Infrastructure and Transport (2024). Status of Vehicle Registration in Korea. Department of Vehicle Operation and Insurance, Mobility and Automobile Bureau.
Seong, J., Yoon, B.-J. (2024). “Analysis of factors influencing accident severity during peak and off-peak hours using XGBoost.” Journal of the Korean Society of Disaster Information, Vol. 20, No. 2, pp. 440-447.
Yang, S.-R., Yoon, B.-J., Ko, E.-H. (2017). “A study of traffic accident analysis model on highway in accordance with the accident rate of trucks.” Journal of the Korean Society of Disaster Information, Vol. 13, No. 4, pp. 570-576. 10.15683/KOSDI.2017.12.31.570
Yoon, B.-J., Lee, S.-M., Park, H.-G. (2025). “Analysis of factors influencing the severity of rental car crashes in tourist areas using machine learning algorithms.” Journal of the Society of Disaster Information, Vol. 21, No. 1, pp. 260-272. 10.15683/KOSDI.2025.3.31.260
Yoon, B.-J., Lee, S.-Y., Jung, S.-Y. (2017). “A study on the factors of highway traffic accidents affecting the EPDO.” 2017 Annual Conference and Special Seminar of the Korean Society of Disaster Information, pp. 251-252.
Yoon, J., Kim, S., Kim, D., Lee, J. (2023). “Determining the optimal Equivalent Property Damage Only (EPDO) weight set for expressway hotspot identification.” Proceedings of the KOR-KST Conference, Busan, pp. 259-260.

Information

Publisher :The Korean Society of Disaster Information
Publisher(Ko) :한국재난정보학회
Journal Title :Journal of the Society of Disaster Information
Journal Title(Ko) :한국재난정보학회논문집
Volume : 21
No :3
Pages :782-790
DOI :https://doi.org/10.15683/kosdi.2025.9.30.782

[1] Korea Transport Database (2022). 2022 Nationwide Freight Data Update. Korea Transport Institute, KTDB-2022-506.

[2] Korea Transport Institute (2022). Volume 6: Nationwide Freight O/D Supplement Update. Final Report of the 2022 National Transport Survey, Ministry of Land, Infrastructure and Transport.

[3] Kwon, C.-W., Chang, H.-H. (2021). “Comparative analysis of traffic accident severity of two‑wheeled vehicles using XGBoost.” Journal of Information Technology Services, Vol. 20, No. 4, pp. 1-12. 10.12815/kits.2021.20.4.1

[4] Lee, G.-J. (2023). A Study on the Analysis of Factors Influencing the Severity of Traffic Accident in Two‑Wheel Vehicles Using Machine Learning–Focus on Incheon Metropolitan City. Master Thesis, Incheon National University.

[5] Ministry of Land, Infrastructure and Transport (2024). Status of Vehicle Registration in Korea. Department of Vehicle Operation and Insurance, Mobility and Automobile Bureau.

[6] Seong, J., Yoon, B.-J. (2024). “Analysis of factors influencing accident severity during peak and off-peak hours using XGBoost.” Journal of the Korean Society of Disaster Information, Vol. 20, No. 2, pp. 440-447.

[7] Yang, S.-R., Yoon, B.-J., Ko, E.-H. (2017). “A study of traffic accident analysis model on highway in accordance with the accident rate of trucks.” Journal of the Korean Society of Disaster Information, Vol. 13, No. 4, pp. 570-576. 10.15683/KOSDI.2017.12.31.570

[8] Yoon, B.-J., Lee, S.-M., Park, H.-G. (2025). “Analysis of factors influencing the severity of rental car crashes in tourist areas using machine learning algorithms.” Journal of the Society of Disaster Information, Vol. 21, No. 1, pp. 260-272. 10.15683/KOSDI.2025.3.31.260

[9] Yoon, B.-J., Lee, S.-Y., Jung, S.-Y. (2017). “A study on the factors of highway traffic accidents affecting the EPDO.” 2017 Annual Conference and Special Seminar of the Korean Society of Disaster Information, pp. 251-252.

[10] Yoon, J., Kim, S., Kim, D., Lee, J. (2023). “Determining the optimal Equivalent Property Damage Only (EPDO) weight set for expressway hotspot identification.” Proceedings of the KOR-KST Conference, Busan, pp. 259-260.

Journal of the Society of Disaster Information ISSN:1976-2208(Print) 2671-5287(Online) 한국재난정보학회논문집

All Issue