Smart Meter Analytics for Residential Energy Efficiency
2024 IEEE Global Energy Conference, GEC 2024, Batman, Türkiye, 4 - 06 Aralık 2024, ss.388-394, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/gec61857.2024.10881649
- Basıldığı Şehir: Batman
- Basıldığı Ülke: Türkiye
- Sayfa Sayıları: ss.388-394
- Anahtar Kelimeler: anomaly detection, consumer segmentation, data-driven energy management, energy consumption analysis, residential energy optimization, smart meters
- Marmara Üniversitesi Adresli: Evet
Özet
Smart meters, in addition to measuring users' energy consumption, provide high-resolution, real-time energy usage data, enabling advanced analyses to uncover patterns, detect anomalies, and optimize residential energy usage. This study analyzes data representing one-hour segments of voltage and current waveforms collected over 96 days from a single-family household in Portugal. The data was converted into MP3 format for analysis and processed using the MATLAB© software package to extract hourly energy consumption metrics, including active power, reactive power, and RMS values, scaled to megawatt-hours (MWh). The conducted analyses included fundamental evaluations and anomaly detection. Statistical analysis revealed consistent consumption patterns tightly concentrated within a narrow range (75.10–75.90 MWh). The average energy consumption across the datasets was approximately 75.5 MWh, with a low standard deviation (~0.2 MWh), indicating stable usage trends. While time-series graphs effectively visualized these patterns and deviations, histograms provided limited insights due to the narrow data distribution. This study demonstrates the potential of smart meter data for identifying inefficiencies and enabling data-driven energy management strategies. The methodology supports advanced analyses such as load profiling, anomaly detection, and consumption optimization, showcasing scalability and applicability to larger datasets. Additionally, this research contributes to identifying hidden usage patterns and advancing smart meter methodologies.