Using semantic web resources to achieve metadata interoperability in the scope of future smart grids

• Autores: Rafael Santodomingo Berry
• Directores de la Tesis: José Antonio Rodríguez Mondéjar (dir. tes.) , Miguel Angel Sanz Bobi (codir. tes.)
• Lectura: En la Universidad Pontificia Comillas ( España ) en 2013
• Idioma: español
• Tribunal Calificador de la Tesis: Luis Sánchez Fernández (presid.) , Luis Rouco Rodríguez (secret.) , Lars Nordström (voc.) , Tomás Gómez San Román (voc.) , Sascha Ossowski (voc.)
• Enlaces
  • Tesis en acceso abierto en: TESEO
• Resumen

  • Metadata interoperability refers to the exchange and processing of machine-processable data between agents with the aim of carrying out tasks in co-operation. In order to manage and control future energy Smart Grids, it is mandatory to achieve metadata interoperability between different types of Smart Grid systems, such as remote energy management systems and local station automation systems.

    An established means of addressing metadata interoperability is standardization. For that reason, the main standardization bodies for Smart Grids have developed standards including metadata models or global conceptual models. Such standards define the syntax and semantics of the metadata that must be exchanged in this domain. Nevertheless, standardization is not itself sufficient to achieve metadata interoperability. Thus, given the numerous vendors, applications and benefits associated with different approaches, in practice it is not possible to define one single standard model which is valid for all Smart Grid systems. Furthermore, standard models are not stationary, i.e. they evolve with Smart Grid technologies. This typically results in heterogeneities (or mismatches) between different versions of the same standard model. Therefore, in order to make metadata interoperability possible within the scope of Smart Grids, it is necessary to develop methods that facilitate the interactions between systems based on heterogeneous models. Such methods are commonly referred to as model reconciliation methods.

    This thesis presents a new methodology, termed Electric System Ontologies Matching Approach (ESOMA). The proposed methodology takes advantage of open-source resources developed within the Semantic Web initiative, and includes innovative model reconciliation methods for integrating Smart Grid systems. Hence, the language mapping methods defined in ESOMA are able to automate the format conversions between the main modeling languages utilized in the Smart Grids. Furthermore, the model mapping (or ontology matching) system developed within the framework of ESOMA is able to automatically find complex semantic correspondences (alignments) between Smart Grid metadata models. In order to do so, it processes deep domain knowledge from external ontologies (called domain ontologies). From the final alignments between the models, the instance transformation methods created in ESOMA are capable of performing valid translations at instance level. Thereby, such methods provide a means of resolving situations of conflict caused by existing mismatches between the models.

    ESOMA was implemented for evaluation purposes in a stand-alone software tool also developed in this thesis, ESODAT. The tests carried out with ESODAT were based on representative use cases of interactions between Smart Grid systems. In particular, the use cases were focussed on two of the most important metadata interoperability gaps in this domain: interactions between CIM-based systems and IEC 61850-based systems, and interactions between systems based on different CIM versions. The results obtained in the tests proved the ability of ESOMA to significantly reduce the integration efforts in comparison with existing state of the art solutions.