基于 SOM 与 K-Means 的铁路事故聚类可视化分析方法

打开文本图片集

DOI：10.19981/j.CN23-1581/G3.2026.12.041

中图分类号：U298.5

文献标志码： A

文章编号：2095-2945（2026）12-0164-04

Abstract： To more clearly identify accident groups with high causal correlation and core accident-causing factors， and provide scientific references for various departments in the railway industry to formulate safety management and control strategies， thereby contributing to the systematic improvement of railway transportation safety levels. The SOM and K-Means clustering visualization analysis methods adopted in this paper can accurately identify and aggregate accident samples with similar accident-causing characteristics， and the clustering effect is significant and its effectiveness has been verified. In addition， through the construction of a railway accident topological distribution mapping model and a thermal map of accident-causing attributes， the visual presentation of railway accident clustering results has been completed.

Keywords： railway accident; railway cause; clustering; data mining; visualization

面对铁路安全保障能力迭代升级与大数据资源深度整合的双重驱动，挖掘事故数据蕴含的深层价值，将其中内隐的致因逻辑转化为可形式化的风险表征，并形成主动防御导向的事故抑制策略，已成为提升行业本质安全水平的关键突破口[1]。（剩余3844字）