For relational database of oil-gas well engineering domain, this paper presented a method of SPARQL-SQL query transformation based semantic views. In this method, a specific ontology is used to describe schema of relational database and tables are defined as semantic query views based on domain ontology. Through semantic mapping provided by semantic views, SPARQL query is parsed and rewrited to SQL statements on relational data sources. A virtual data center system is implemented with which feasibility and validity of methods provided in this paper are tested, and we achieved a good application effect.