Algorithmic approach to data model versioning

Abstract

This paper describes algorithmic approach to consistent and efficient management of changes in data model for distribution network applications in smart grid solutions. In application domain of power distribution network analysis (DNA), clean separation between engineering data model and operational (run-time) data model must be maintained. Until explicit request to import changes from engineering data model into operational data model, operational data model must not be affected with changes in engineering data model. The process of activation introduces changes prepared within engineering data model into operational data model and thus put them in use. This paper outlines algorithm for efficient data model change management and activation process. This approach provides efficient handling of data model changes regarding memory and disc space usage. It allows fast switching to previous data models, allows arbitrary long history about past activations; and simplifies incident and what-if analysis for any of previous configurations.