Building Profiling for Energy Management
The heterogeneity of building automation technologies and the amount of sensor data created are still an inhibitor for the successful deployment of home or building energy management systems (HEMS/BEMS). BProfEM, which is supported by FFG (IKT der Zukunft, P 845 627), aims at defining a generic semantic integration layer amongst the heterogeneous technologies using Semantic Web technologies in order to exploit the enriched semantics to support automated data mining, (privacy-aware) profiling and adaptive optimization algorithms to be used within energy management systems. The goal is to improve the local energy optimization and smart grid interaction within automated demand response programmes. Real world test data samples are used to identify interesting patterns of human behaviour and building energy consumption and to further create a simulation model to be used within a building simulation environment that shall evaluate the benefits of a semantic-aware energy optimization and demand response system. A feasibility study accompanied with a cost/benefit analysis will evaluate the potential of the project results.
BProfEM addresses the algorithmic and data mining challenges that arise by combining local energy management systems with the smart grid, with a special focus on demand response interaction. The main methods used in the project are information meta-modelling based on ontologies, meta-model transformation, algorithm design for data mining, optimization problem formulation, and algorithm design. As evaluation methodologies, building simulation based on real world test data samples and a feasibility study are used.