Research On Bidirectional Data Mapping For EPDM And E&P Based On Meta-data

Since1989, Petrochina has started to study the model of oil exploration and production(E&P), and has formed the data model standard. With the increasing of informationrequirements, in oilfield enterprise, the data model of oil exploration and development hasalways changed. In order to meet the application requirements of the integration ofexploration and development, after four years, an integrated Data Model (EPDM) of oil andgas exploration and development is formed in2009, which merged into some advancedconcepts of the Epicentre model of POSC. Since the early applications are programmed basedon the E&P, these applications could not be abandoned within a short time. Currently, facingthe replacement of E&P and EPDM, the problem is how to integrate the historical data ofE&P into EPDM. In the same time, in order to support the application based E&P, the datasynchronization of EPDM into E&P must be considered. Traditional data exchange employsthe one-to-one programming. Because of the differences of E&P models among oilfieldenterprises, the development, which needs to program the unique program for every enterprise,needs a large amount of work. And the maintenance lately is enhanced in exponential index.Obviously, this method is not feasible for such a complex data model.Based on the problems above, in order to solve the migration and synchronization ofE&P and EPDM, this paper proposes the strategy of bi-directional data mapping, which isalso metadata-driven. And ETL technology is brought into the heterogeneous data model, thefollowing problems are mainly solved. First of all, through the specifically analysis of avariety of heterogeneous problems existing in these models, the data mapping relationship istheoretically defined in three levels: schema level, table level and column level. Secondly, themapping relationship is divided into simple mapping and complex mapping according to themapping complexity. Then, the problems of bi-directional mapping are discussed, namely,reversible and irreversible mapping relationships. And, according to the analysis of mappingrelationship, the definition of every mapping rules and transfer functions are given. By meansof the meta-data, some meta-models are built, such as data dictionary meta-model, mappingrules meta-model, mapping-driven meta-model and so on. At last, the data migrationarchitecture based on the ETL framework is designed. Based on the architecture, theapplication system is implemented, which achieves a set of functions, such as data mapping,data extraction, data transformation, and data loading and so on. This system has been wellapplied in the data migration of Daqing and Southwestern oilfield. And the result shows thatthe heterogeneous data model and the bi-directional mapping strategy proposed, effectivelysolve the bi-directional mapping issue of the heterogeneous data between EPDM and E&Pdata model. It also possesses some extensibility. The research results can proviede efficienntreference for the common data exchange issues.