Research on Identification of Causes and Prevention and Control Measures of Railway Freight Accidents Based on Complex Network

For the railway operation system, it is vital to find the fundamental causes and their associated chain of rail freight accidents and reduce the chance of accidents from the source. First, we use structured railroad freight accident data based on Derailment, Collision, Crossing, Obstruction and other five types of accidents, and subdivide the causes of accidents into 31 categories from the perspective of human, machine and environment. We use the A priori algorithm to find the relationship between accident types, causes and consequences of accidents. Second, according to the obtained correlation relationship, we construct a complex network model of railway freight accident causation, and mine the fundamental causes and their associated chain of accidents by calculating the topological characteristics of network node degree, path length and clustering coefficient. Finally, we made suggestions for preventing and controlling railroad freight accidents. The results show that the fundamental causes and consequences of different types of accidents are different, and accident prevention and control should focus on the influence of human factors, the integrity of line structure factors such as roadbed and regular maintenance of railroad locomotive equipment.