Research On Knowledge Representation And Reasoning Of Miner’s Unsafe Behavior Based On Ontology

The unsafe behavior of miner is the immediate cause of mine accidents,which poses a serious threat to coal mine safety production.The study of miners’ unsafe behavior covers many subjects,and the representation methods of knowledge in related fields are different,which greatly affects the reuse and sharing of domain knowledge.There is also a lack of specification for describing miners’ unsafe behavior knowledge in the mine field.Most of the existing description methods can’t be presented in machine-readable form.This makes the identification of miners’ unsafe behavior mainly rely on manual monitoring.In order to solve these problems,this paper studies ontology-based knowledge representation and reasoning method of miners’ unsafe behavior which can not only provide a unified formal concept for domain knowledge description,but also provide methods for identifying miners’ unsafe behavior,which is of great significance for improving the intelligence level of mine information processing.This paper combs the semantic relationships between man-machineenvironment systems,analyses the context of miners’ unsafe behavior,puts forward the ontology model of miners’ behavior context,and divides the context model into three modules: miners’ behavior,equipment and environment.By extracting the factors which affect miners’ unsafe behavior,the key terms and attributes are determined.Based on the fusing of various ontology construction methods,the ontology construction process of miners’ behavior context is completed by using Protégé.The realized context ontology can not only clearly describe the knowledge related to miners’ unsafe behavior,but also be used independently for the knowledge representation of human,machine and environment modules.By organizing the knowledge of mine safety,the rules are extracted;the classification method of rules is determined according to the content of the constructed ontology and reasoning requirements;the semantic Web rules are analyzed and the self-defined reasoning rules set is compiled based on SWRL language,and the intelligent discovery of new knowledge is completed through Jess reasoning engine.Jena reasoning machine and SPARQL query tool are used to obtain the required miner information.This method satisfies the query requirement of deep knowledge and is more accurate and efficient than traditional database.In order to improve the adaptability and practicability,the quality evaluation of the context model is carried out.The semantic consistency of ontology knowledge base is guaranteed by Fa CT++ inference engine.Considering the modularization of miners’ unsafe behavior knowledge,cohesion and coupling are selected as evaluation metrics.The evaluation process is based on the miners’ behavior ontology model,equipment ontology model and environment ontology model.The evaluation results show that the different modules of miners’ behavior context ontology have functional independence and can expand the domain independently.