Knowledge Base Construction And Early Warning Inference Of Gas Accidents Based On Time-Space Constraints

With the development of coal mine accident prevention technology,mining companies have developed a series of early warning systems for gas accidents to prevent accidents,which has reduced the gas accident rate to a certain extent.However,the existing warning methods for gas accidents in coal mines which mainly based on artificial prevention through inspection data perform poorly in efficiency and accuracy.In order to realize the intelligent analysis and early warning of gas accidents and the reuse and sharing of related knowledge,based o n the ontology and accident tree theory,this paper proposes to build a dynamic gas accident warning knowledge base and reasoning model to meet the gas accidents under the constraints of time and space coexistence of various disaster factors Knowledge description and analysis and early warning can effectively calculate the dynamic gas accident occurrence probability under space-time constraints,and has made useful explorations in the field of gas accident knowledge description and intelligent early warning.The specific research content is as follows:(1)Constructing the spatiotemporal knowledge base of mine gas accidents.Firstly,use a combination of accident tree analysis and ontology description and incorporating the internal relationship between the spatiotemporal characteristics of the mine and the occurrence of gas accidents.Furthermore,protege is used to construct a gas accident knowledge base of space-time constraints,and describe the knowledge logic of gas accidents under space-time constraints.Then,based on the knowledge logic of the built gas accident knowledge base,the automatic generation algorithm of the mapping file is designed based on the R2 RML algorithm,describing how to automatically write the mapping file between different types of tables in the relational database,and finally parse the mapping file through a custom ontology mapping processor To map the data in the relational database to the specified gas accident knowledge base to form instance data in the knowledge base.(2)Design gas accident early warning reasoning rules.Firstly,research and design Jena inference rules based on hazards to basic events to realize reasoning from hazards to basic disaster-causing events.Then incorporate the time-space constraint rules that meet the various disaster-causing factors,based on the spatiotemporal accident tree analysis method,the probability prediction of gas accidents is realized by a quantitative method.According to the predicted results,we will warn whether the gas disaster event will occur,and at the same time give the combination of basicevents leading to the gas accident to prevent the accident.(3)Develop a gas accident intelligent reasoning and early warning system based on space-time constraints.The system is designed to realize the deduplication of relational database data,semi-automatic construction of instance data in the gas accident knowledge base,early warning of gas accident probability,and the combination of basic events that lead to gas accidents.At the same time,the functions of each module are visually displayed.Initially realize the early warning of gas accidents based on the knowledge logic level,and verify the feasibility and practicability of the intelligent early warning system of coal mine gas disasters.The paper has 49 figures,19 tables,78 references.