Analysis method of multi-attribute railway accident cause weight

Abstract

The present invention provides a multi-attribute railway accident cause weight analysis method, including: based on the railway accident cause classification CREAM-RAs model, decomposing and encoding all railway accident reports, and establishing a multi-attribute railway accident data set MARA-D ;Based on the self-organizing map algorithm SOM and K-Means integrated clustering method for railway accidents, MARA-D is clustered and analyzed to obtain different accident clusters; according to the association rule algorithm AL-Apriori of railway accident levels, different According to the association rule analysis of the accident clusters, the strong association rules between the causes of different accident clusters are obtained; based on the strong association rules and the decision laboratory analysis method DEMATEL, the comprehensive weight of the cause is obtained, and then according to the comprehensive weight of the cause Sort by weight. This method can find out the key causes of railway accidents, and provide reference for further improving the safety level of railways.

Description

technical field [0001] The invention relates to the field of railway accident data mining, in particular to a multi-attribute railway accident cause weight analysis method. Background technique [0002] With the high demand for my country's railway system safety assurance capabilities and the advent of the era of big data, it is urgent to conduct comprehensive study and analysis of past railway accidents, and transform tacit knowledge in railway accident data into explicit knowledge. The establishment, mining and analysis of railway accident data sets will be of great significance to railway accident prevention. [0003] At present, commonly used accident causation classification models include SHEL (Software, Hardware, Environment, Human Software Hardware Environment Liveware) model, Human Factors Analysis and Classification System (The Human Factors Analysis and Classification System, HFACS) model and Cognitive Reliability and Error Analysis Method (Cognitive Reliability a...

AI Technical Summary

Problems solved by technology

However, as a general classification framework, the CREAM classification model does not cover all the contributing factors that lead to railway accidents. Therefore, it is necessary to improve the existing CREAM classification model according to the characteristics of my country's railway traffic accidents, in order to analyze the More scientific extraction and classification of causes

[0004] In addition, the current mining methods of accident data sets generally use clustering and association rule analysis, but they are still in the initial stage of theoretical exploration and method research, and most of the research focuses on the field of road traffic accidents and ship traffic accidents, and relatively few in the field of railway accidents. , so it is necessary to do further exploration and research on the mining and analysis of railway accident data sets

Images