Air Conditioning Load Monitoring in NILM Using Low Frequency Power Data

Air conditioning loads are typically flexible loads and contribute significantly to enhancing energy efficiency while optimizing the operation of the electrical grid. However, air conditioners themselves are continuously variable loads, and accurately monitoring their load status is a challenging task in the field of non-intrusive load monitoring. To solve this problem, this paper proposes a template matching-based load monitoring method for air conditioners. First, this paper presents a load event detection strategy designed to identify changes in appliance operational status; subsequently, for each load event, multidimensional load features are extracted; finally, the extracted load features are fused, and in this way, an air conditioning load recognition model is constructed and trained to realize the air conditioning load state recognition. Tests comparing the proposed method with existing ones, using both the REDD dataset and real-world data collected in Henan, China, demonstrate that the proposed method can accurately monitor the air conditioning load status.