Extracting Operating Modes from Building Electrical Load Data

Abstract

Empirical techniques for characterizing electrical energy use now play a key role in reducing electricity consumption, particularly miscellaneous electrical loads, in buildings. Identifying device operating modes (mode extraction) creates a better understanding of both device and system behaviors. Using clustering to extract operating modes from electrical load data can provide valuable insights into device behavior and identify opportunities for energy savings. We present a fast and effective heuristic clustering method to identify and extract operating modes in electrical load data.