Applied research of data-driven multi-level reliability model in transportation safety risk evaluation

In view of the complexity and unpredictability of risk data in safe production of transportation enterprises, this paper introduces the concept of risk source metadata, proposes the definition and extraction method of risk source metadata for transportation enterprises, and realizes the static description and dynamic process description of data source characteristics; a safety risk evaluation method based on multi-level reliability model is established, and the comprehensive evaluation and analysis of risk sources are realized; finally, a safety risk evaluation system based on Internet + for transportation enterprises is constructed, and improves the management process of safety risk evaluation for transportation enterprises.