On Using Semantic Complex Event Processing for Dynamic Demand Response Optimization

Demand response optimization (DR) is a key component of Smart Grids. However, existing DR programs fail to effectively leverage the near-realtime information available from AMIs and BANs to adapt to increasing dynamism in energy use profiles. In this paper, we investigate the use of Semantic Complex Event Processing (CEP) to model and detect dynamic situations in a campus micro grid that facilitates DR. We abstract event streams from various information sources including smart meters, sensors, and online services as a logical stream of events. We offer a taxonomy of DR event patterns that can be defined for power use prediction, curtailment opportunity detection, and situation awareness. Further, we illustrate the efficacy of pattern detection for DR by applying them to realtime event streams in the USC Campus micro grid using our SCEPter Semantic CEP framework and analyzing their statistics over a 4 day period.