Algorithm for nonintrusive identification of residential appliances

An algorithm has been developed to use at a central processing site for identifying appliances and appliance usage at remotely monitored residences. This algorithm is part of a nonintrusive system that uses sampled data for real and reactive power accumulated over an extended period of time. Using a logic analysis, the algorithm can recognize at least six major appliances, including a heat pump, as well as any two appliances that turn on or off at nearly the same time. Data from several residences in the Rochester, NY, and Columbus, OH, areas have been analyzed during this study.