Research On Construction And Reasoning Of Coal Mine Floor Water Inrush Prediction Knowledge Base Based On Ontology

The water inrush of coal floor is one of the main factors that cause the coal mine disaster,which poses a serious threat to the safe mining in the mine shaft,and the work of water prevention and control is facing with great challenges.With the development of knowledge engineering and artificial intelligence,a knowledge base based on expert system has emerged in the field of water inrush prediction,which realizes automatic processing of water inrush prediction information.However,the traditional water inrush expert system has a single function,poor portability and lack of systematic expression of knowledge,and that is difficult to realize the sharing and reuse of knowledge,which can easily lead to knowledge waste.In this thesis,the ontology theory is introduced into the construction of the knowledge base of water inrush from floor.It can not only realize the unified formal description of the knowledge in the field of water inrush,but also make intelligent reasoning on the accident risk information according to the existing knowledge,which can provide better knowledge service for the realization of prediction and prevention of water inrush.The mechanism of the water inrush from floor is summarized and the demand of water inrush prediction knowledge base constdruction is analyzed in this thesis,the basic model framework for the knowledge base is identified through the research.An ontology modeling method that applicable for the floor water inrush prediction knowledge base is proposed based on the analysis of the current ontology construction method.Based on the conditional random field,the important concept of water inrush is extracted and the semantic relationship between concepts is summarized;The ontology knowledge base is constructed by defining the data attributes,object attributes and water inrush instances by the ontology modeling tool of Protégé4.3;A self consistency inference machine FaCT++ is used to check the consistency of the constructed ontology and store the ontology as OWL file.The format and requirements of custom rules are analyzed according to the requirements of coal mine water inrush prevention technology and regulations,and an ontology reasoning rule base for floor water inrush prediction and prevention is established;The spatio-temporal logic is combined to construct spatio-temporal constraint rules,which provides space-time support for the prediction rule base,and the rule based knowledge reasoning is realized by Jena reasoning machine.Inaddition,the method of case-based reasoning is researched to retrieve the similar case retrieval of floor water inrush,the attributes of case attributes that affect floor water inrush are extracted and attribute similarity is calculated;The attribute weight is valued based on the analytic hierarchy process(AHP),the similarity retrieval algorithm of water inrush case is determined,and the validity of CBR in the prediction of water inrush is verified by experiments.A prototype system of water inrush knowledge base is developed,which stores water inrush database,water inrush case library and inference rule library,and the case retrieval is implemented based on similar case retrieval algorithm,which provide the service of case reasoning and rule reasoning.The validity and practicability of the prediction knowledge base is verified by the example of water inrush from floor,and the results show that the result of the knowledge inference is consistent with the actual evaluation result,which can provide knowledge service for intelligent analysis of water inrush and prediction prevention.