Study On Construction Of Safety Management Knowledge Graph For Road Transport Of Dangerous Goods

In recent years,the country has issued a number of new standards in line with the development trend of road transportation of dangerous goods,and put forward the requirements for systematic safety management for the road transportation of dangerous goods.Meanwhile,safety management personnel should learn from accident knowledge to ensure the scientific safety management.However,there are still difficulties in the popularization of standard norms and accident experience.The huge knowledge volume and complex knowledge structure increase the knowledge burden of personnel,which is not conducive to the safety management of dangerous goods road transportation enterprises development.(without special instructions,hereinafter referred to as "dangerous transport enterprises").How to obtain the required safety management knowledge from the large amount of knowledge is of great significance to improve the safety management level of dangerous transportation enterprises.This paper takes the knowledge of road transportation safety management of dangerous goods as the research object,introduces and constructs the knowledge management method of knowledge graph,uses the graph to obtain domain safety management knowledge,and enhances the safety knowledge reserve of personnel in dangerous transportation enterprises.The specific research work is as follows:Combined with the overall structure of the dangerous goods road transportation project and the safety management process and work content,a knowledge requirement model for safety management in the DRSM field is proposed.Clarify the knowledge graph and knowledge structure elements,introduce the related technologies of knowledge graph construction,and support the knowledge graph construction from the theoretical and technical aspects.Based on the scope of domain knowledge,a model-level knowledge modeling process is proposed.Integrate the knowledge classification system in multiple dimensions in the road transportation of dangerous goods,and finally form a multi-dimensional related knowledge structure system.Propose a conceptual model of sub-domain knowledge in the DRSM domain,condense the key knowledge,obtain the core concept of the DRSM domain knowledge graph and expand the relationship,and construct the DRSM domain knowledge graph model layer of a multi-dimensional related knowledge system.Multivariate data is used to construct the DRSM domain knowledge graph data layer,and the data is sorted according to the data structure.Based on the conceptual structure design in the pattern layer,and according to different data characteristics,a multi-method fusion method of domain entity extraction is proposed.The accuracy of model extraction basically reaches 85%,and a total of 988 professional domain entities and 4,518 accident corpus entities are obtained.Construct the corpus seed template extraction relationship,extract it through Bootstrapping algorithm,and obtain the domain entity relationship in 852 accident corpus.Use the graph database Neo4j to perform knowledge management on the extraction results,and complete the construction of the DRSM domain knowledge graph.The relationship between the sub-modules of the DRSM domain knowledge graph is analyzed,and the operation mode of the DRSM domain knowledge graph is explained.Taking a loading and unloading accident as an example,the DRSM domain knowledge graph was used to obtain related knowledge,and the risk control measures in the loading and unloading process were proposed,and the safety management responsibility list was listed for the enterprise.System development is carried out based on the DRSM domain knowledge graph.