The technologies of expert system, ontology modeling and reasoning are studied and applied in this paper. A framework of ontology-based expert system for equipment selection and matching in fully mechanized caving face is built up. The OWL(Web Ontology Language) based equipment knowledge base is set up, and the redundant test is made in the protégé environment. The equipment selection and matching rules are built up by using SWRL(Semantic Web Rule Language). By using JAVA language, the output of equipment selection expert system is optimized based on the theory of Analytic Hierarchy Process. At last, the functions of the expert system are achieved under JESS(Java Expert System Shell) environment. Practical application shows that the equipments selected and matched by ontology-based expert system are more realistic.