Using knowledge discovery for autonomous decision making in smart grid nodes

Smart and energy efficient (office) buildings do not only have to implement smart sensors and actuators, they should also be able to be optimized to be as energy efficient as possible based on the behavior of the user. This paper focuses on knowledge extraction of a smart building and automatic rule creation based on that knowledge. We are using different methods to analyze this data, create the appropriate rule set based on the extracted data and based on the correlation and dependencies of different datasets. These methods are also detecting changes in the data (resp. the behavior of the user) and adapts the ruleset accordingly.