A Machine Learning Framework to Infer Time-of-Use of Flexible Loads: Resident Behavior Learning for Demand Response

摘要: Load shapes obtained from smart meter data are commonly utilized to understand daily energy use patterns for adaptive operations in applications such as Demand Response (DR). However, they do not provide information on the underlying causes of specific - i.e., inference appliances' time-of-use (ToU) actionable information. In this paper, we investigated a scalable machine learning framework infer ToU load collection residential buildings. A and generalized model obviates need training each building facilitate its adoption by relying set previously observed buildings with available appliance-level data. To end, demonstrated feasibility using shape segmentation boost their nearest matches that share similar patterns. an appliance building, classification models trained subintervals matched known ToU. The was evaluated real-world Pecan Street Dataport. results case study electric vehicles (EV) dryers showed promising performance 15-min 83% 71% F-score values, respectively, without in-situ training.