GSI-TOPSIS Method for Quantization of Railway Safety Situation Based on Incident and Accident Data

Abstract

This study holds significant theoretical and practical importance for enhancing the safety management and operational efficiency of railway systems. Currently, there is a notable gap in standardized safety-level measurement methods across diverse railway operating entities. Conventional safety assessment approaches predominantly rely on qualitative analysis frameworks, which often fail to comprehensively address the multifaceted risk factors inherent in complex railway operating environments. To address these limitations, this study leverages historical operational data from participating railway companies to propose an advanced integrated quantitative methodology: the Global Safety Index (GSI)-Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) method. This innovative approach quantifies railway safety conditions by systematically analyzing incident and accident data, integrating statistical modeling frameworks, data imputation algorithms, and comprehensive analytical protocols. The method enables detailed examination and interpretation of large-scale operational datasets within railway systems. The implementation of this quantitative framework has demonstrated substantial improvements in the accuracy of safety performance metrics. Furthermore, it offers robust technical support for developing risk mitigation strategies and optimizing safety performance in the railway sector. By employing systematic risk factor identification and data-driven safety quantification, this approach facilitates accident prevention, enhances risk early-warning systems, and provides evidence-based decision-making support. As research progresses and the accessibility of Chinese railway safety data increases, the analytical precision of this methodology can be further refined. Future applications may include in-depth analyses of Chinese railway risk event datasets, thereby offering strong technical support for the continuous elevation of safety standards in China’s railway operations.

Keywords

Risk accident data / Correlation analysis / Global Safety Index (GSI) / Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS)

References

[1] LiuHL. Research on the safety risk of high-speed railway train control system equipment based on complex network theory. Railw Transp Econ, 2022, 44(7): 82-89

[2] UIC (2020) UIC safety report: significant accidents 2019. Public Report, Paris

[3] DavariN, VelosoB, CostaGA, et al.. A survey on data-driven predictive maintenance for the railway industry. Sensors, 2021, 21(17): 5739

[4] Gao H, Shi Y (2024) Implementation of fault diagnosis technology of EMU bearing based on VMD and neural network. In: International symposium on intelligent robotics and systems (ISoIRS), Changsha, China, pp 253–257. https://doi.org/10.1109/ISoIRS63136.2024.00056

[5] WangP, WeiQ, ZhaoG, et al.. Safety risk identification method for railway construction in complex and dangerous areas. Sustainability, 2022, 14(21): 13698

[6] LinM, ShanM, ZhouJ, et al.. A data-driven fault diagnosis method using modified health index and deep neural networks of a rolling bearing. ASME J Comput Inf Sci Eng, 2022

[7] XinG, ZhangY, WangX, et al.. Autonomous bearing fault diagnosis based on fault-induced envelope spectrum and moving peaks-over-threshold approach. IEEE Trans Instrum Meas, 2024, 73: 1-12

[8] AlawadH, KaewunruenS, AnM. Learning from accidents: Machine learning for safety at railway stations. IEEE Access, 2020, 8: 633-648

[9] PhamMT, KimJM, KimCH. Deep learning-based bearing fault diagnosis method for embedded systems. Sensors, 2020, 20(23): 6886

[10] HuangWC, ZhangY, YuYC, et al.. Historical data-driven risk assessment of railway dangerous goods transportation system: comparisons between entropy weight method and scatter degree method. Reliab Eng Syst Saf, 2020, 205: 107236

[11] CaoY, AnYT, SuS, et al.. A statistical study of railway safety in China and Japan 1990–2020. Accid Anal Prev, 2022, 175: 106764

[12] WangHX, TianYLD, YinH. Correlation analysis of external environment risk factors for high-speed railway derailment based on unstructured data. J Adv Transp, 2021

[13] WangHX, GuoLT, YinH. Association mining for operation and maintenance safety risks of EMUs based on unstructured event data. Urban Rail Transit, 2025

[14] ZhangM. On how to strengthen the railway business line construction safety management. Inner Mongolia Sci Econ, 2012, 263(13): 42-44

[15] UIC (2023) UIC safety report: significant accidents 2022 Public Report, Paris

[16] Laarhoven T (2017) L2 regularization versus batch and weight normalization. arXiv. https://arxiv.org/abs/1706.05350

[17] OlshenRA, RajaratnamB. Successive normalization of rectangular arrays. Ann Stat, 2010, 38(3): 1638-1664

[18] LiaoY, LiuL, XingC. Investigation of different normalization methods for TOPSIS. Trans Beijing Inst Technol, 2012, 8(871–875): 880

[19] Panda M, Jagadev AK (2018) TOPSIS in multi-criteria decision making: a survey. In: 2018 2nd international conference on data science and business analytics (ICDSBA), Changsha, China, pp 51–54. https://doi.org/10.1109/ICDSBA.2018.00017

[20] ÇelikbilekY, TüysüzF. An in-depth review of theory of the TOPSIS method: an experimental analysis. J Manag Anal, 2020, 7(2): 281-300

[21] KongQ. Urban rail transit operation safety evaluation method based on improved TOPSIS method. Adv Transp Stud, 2022, 2022(1): 47-56

[22] WuHW, LiEQ, SunYY, et al.. Research on the operation safety evaluation of urban rail stations based on the improved TOPSIS method and entropy weight method. J Rail Transp Plan Manag, 2021, 20: 100262

[23] QiYK, LiXY, LiuJH, et al.. Research on evaluation of basic railway safety management capability based on combination weighting—TOPSIS. Railw Sci, 2024, 4(1): 108-122

[24] Qi ZX (2013) China implements radical railway reform. Int Railw J 53(8)

[25] LugerK. Chinese railways: reform and efficiency improvement opportunities. Physica, 2008

[26] TangT, ChaoL, RuN. Research on safety management system for China Railway. Proc ASME/IEEE Joint Rail Conf, 2012

[27] YangHR, WangC, YouYY. The spatial structure evolution of China’s high-speed rail network and its impacts on real estate investment. Appl Spatial Anal, 2022, 15: 49-69

[28] Li H, Xia X (2024) Social benefit impact analysis of high-speed rail from policy perspective. In: The social benefits of high-speed rails in China. Contributions to Public administration and public policy. Springer, Singapore. https://doi.org/10.1007/978-981-97-1695-1_7

[29] YuanZ, YuanX, YangY, et al.. Greenhouse gas emission analysis and measurement for urban rail transit: a review of research progress and prospects. Digit Transp Saf, 2023, 1(1): 37-52

[30] ChangN, DongC, ZhuS, et al.. Towards recognizing cognitive distraction levels with low-cost and high-sensitive measures: the effectiveness of sample, approximate, and traditional steering entropies. Transp Res F: Traffic Psychol Behav, 2025, 109: 1063-1079

[31] WangW, YangY, YangX, et al.. A negative binomial lindley approach considering spatiotemporal effects for modeling traffic crash frequency with excess zeros. Accid Anal Prev, 2024, 207, ArticleID: 107741

Funding

China Railway(RD2024T001)

RIGHTS & PERMISSIONS

The Author(s)