Risk Warning for Special work Based on an Event Knowledge Graph

The number of accidents occurred in special work for petrochemical enterprises was still very high. Controlling the number related to special work from the source was an interesting research direction to reduce the accident rate. The data of permit to work derived of a permit to work system was processed and an event knowledge graph about permit to work was constructed, then from the three aspects of essential defect work, sub-quality work and unplanned work, data mining was used for analyzing unreasonable work objects and work methods. Through the study, the cumulative probability of the permit to work for the spherical body or the specific parts and components of the spherical tank was a logarithmic function of working time interval, and 57.27% of the permit to work tickets were operated repeatedly within 30 days.