In the MIRABEL project, a data management system for a smart grid is developed to enable smarter scheduling of energy consumption such that, e.g., charging of car batteries is done during night when there is an overcapacity of green energy from windmills etc. Energy can then be requested by means of flex-offers which define flexibility with respect to time, amount, and/or price. In this paper, we describe MIRABEL DW, a data warehouse (DW) for the management of the large amounts of complex energy data in MIRABEL. We present a unified schema that can manage data both at the level of the entire electricity network and the level of individual nodes, such as a single consumer node. The schema has a number of complexities compared to typical DW schemas. These include facts about facts and composed non-atomic facts and unified handling of different kinds of flex-offers and time series. We also discuss alternative data modeling strategies and how specialized variants of the generic schema can be used by different node types while we maintain compatibility and consistency between them. Finally, we present typical queries from the energy domain and a performance study.
|Titel||Transactions on Large-Scale Data- and Knowledge-Centered Systems XXI : Selected Papers from DaWaK 2012|
|Redaktører||Abdelkader Hameurlain, Josef Küng, Roland Wagner, Alfredo Cuzzocrea, Umeshwar Dayal|
|Publikationsdato||17 jul. 2015|
|Status||Udgivet - 17 jul. 2015|
|Navn||Lecture Notes in Computer Science|