Smart energy solutions generate continuous data about electricity use through IoT smart meters and sensors. This allows for smart analysis instead of just monitoring. This research paper focuses on examining common energy consumption trends and spotting any unusual cases that could arise from energy waste, faulty equipment, leaks, or unauthorized use. The approach combines pre-processing, feature extraction, clustering, and anomaly detection methods. K-Means with silhouette score validation helps identify common energy consumption patterns. Meanwhile, the Z-score and isolation forest techniques detect unusual occurrences in time-series energy data.
R et al. (Wed,) studied this question.