Computational Cost Analysis and Data-Driven Predictive Modeling of Cloud-based Online NILM Algorithm

Online non-intrusive load monitoring methods have captivated academia and industries as parsimonious solutions for household energy efficiency monitoring as well as safety control, anomaly detection, and demand-side management. However, despite the promised energy efficiency by providing appliance specific consumption information feed-backs, the computational energy cost for running the load monitoring systems is not explored. This study analyzes whether the energy spent to execute the non-intrusive algorithms, out-weights the expected energy efficiency gain from using the algorithms. Furthermore, we present a study on the computation costs estimation and prediction of a Cloud-based online non-intrusive load disaggregation algorithm through data-driven models. Moreover, a generic framework for an automated algorithm computational cost monitoring and the modeling methodologies are devised and proposed for meeting extensive scaling load monitoring and deployment requirements. The proposed approach was examined and validated on ls and cvs running the disaggregation algorithm. The prediction models, developed using statistical and machine learning tools, demonstrate the promising applicability of the data-driven approach with a very high prediction accuracy without detailed knowledge of the computing systems and the algorithm.