Coal Mine Accident Early Warning Knowledge Base Model And Application Based On Ontology

Although China’s coal mine deaths decreased year by year，the situation of safetyis still severe.Packaging the things in the underground coal mine by the internet ofthings technology and multi-agent theory into intelligent agents which contain coalmine accident early warning knowledge base can reduce coal mine accident rate.Butfor the present there is a lack of research on coal mine accident early warningknowledge base model based on accident causation mechanism. In recent years，theontology has achieved great development on knowledge representation and artificialintelligence. Therefore，the elements and relationships of the coal mine accident earlywarning knowledge base model and its application were studied based on ontology inthis dissertation.Firstly，the concept of danger source and hidden trouble are redefined and theirrelationship is analysed. Afterwards,combined with the temporal logic, the highabstract accident causation mechanism is based on the root danger source inspace-time perspective was proposed which is the theoretical basis of the coal mineaccident early warning knowledge base model. Then, space-time accident treeanalysis method which suits concrete accident casuation mechanism description wasdesigned. Furthermore，the coal mine accident early warning knowledge base modelwas constructed according to the root danger source agent’s early warning procedureand ontology. The model includes the accident causation mechanism based on theroot danger souce，space-time logic，coal mine danger source base，representationmethods of specific accident casuation mechanism，specific accident causationmechanism description and reasoning machines.Based on the present space-time logic, the concept hierarchy and interrelation oftime and space entity were studied in the space-time logic of accident early warningknowledge base which provides support for describing the accident causationmechanism and warning rules. The time logic was expressed by combining point andinterval，which divides the time entity into time point and interval and divides thetime relation into13major categories. The space logic uses subset of OGC whichincludes point,line and area to construct root danger source spatial representation. Thespatial relation was divided into three basic categories，namely，topological relation，directional relation and metric relation. The accidents early warning knowledge base ontology and reasoning alogrithmwere built to enable the intelligent agent to understand the accident early warningknowledge base model and reason. Coal mine accidents early warning knowledgebase ontology includes coal mine risk ontology which describes accident causationmechanism based on the root danger source and concepts in coal mine field，timeontology which describes time entity and its relation，space ontology which describesspace entity and its relation，space-time accident tree ontology which describesspecific accident causation mechanism.The research of accident intelligent early warning reasoning algorithms includesdescription logic reasoning algorithm based on ontology，time reasoning algorithm，space reasoning algorithm and space-time accident tree reasoning algorithm，whichprovides algorithm supports for accident intelligent early warning. The descriptionlogic reasoning was achieved by using Tableau algorithm. The time qualitative andquantitative reasoning was achieved by using the time constraint network. Thereasoning on topological space，direction and metric relation was achieved by usingcombination table. Moreover，the qualitative and quantitative reasoning calculation ofthe space-time accident tree was realized combining the description of space-timerestriction on the basis of the present qualitative and quantitative calculation of theaccident tree.Finally，the wanglou coal mine’s four layers architecture platform of accidentintelligent early warning was designed and applied to test and verify the coal mineaccident early warning knowledge base model. The application cases show that thecoal mine accident early warning knowledge base model based on ontology hasreasonable theory and effective application， which can provide the beneficialreferrence for the construction of coal mine accident early warning knowledge base.