Anomaly Detection in AMI and Smart Meter Data for Electricity Theft, Outage, and Equipment Fault Identification: A Comprehensive Review
DOI:
https://doi.org/10.63282/3050-9246.IJETCSIT-V4I3P120Keywords:
Advanced Metering Infrastructure, Smart Meters, Anomaly Detection, Electricity Theft, Outage Detection, Equipment Faults, Machine Learning, ReviewAbstract
Advanced Metering Infrastructure (AMI) and smart meters have become important elements of the contemporary power distribution system, providing the ability to monitor finely, have two-way communication, and make operational decisions based on the data. Although these systems have significant positive aspects in regard to efficiency, reliability and customer interactions, they also pose new threats in regards to data integrity, system security and operational anomalies. Non- technical losses, including electricity theft, equipment faults, and operational events, including outages and restoration events may be the reason behind the abnormal patterns in the smart meter data. Such anomalies should be prevented through proper and efficient tracking of them in order to guarantee grid stability, as well as minimizing financial costs, and consumer confidence. The paper includes an extensive literature review of anomaly detection methods used on AMI and smart meter data but in relation to the methodology of detecting theft, outages, and equipment-related faults. Traditional statistical methods, machine learning methods, and deep learning methods are systematically analyzed, as well as data preprocessing techniques, feature engineering techniques, and metrics. Practical issues are also addressed, including imbalance of data, privacy, scalability and interpretability. The paper is an attempt to give a systematic source to the researcher and practitioners wishing to learn the current status of anomaly detection in smart metering systems and its application in intelligent power distribution management.
Downloads
References
[1] Carr, D., & Thomson, M. (2022). Non-technical electricity losses. Energies, 15(6), 2218.
[2] Wang, Y., Chen, Q., Hong, T., & Kang, C. (2018). Review of smart meter data analytics: Applications, methodologies, and challenges. IEEE Transactions on smart Grid, 10(3), 3125-3148.
[3] Guerrero-Prado, J. S., Alfonso-Morales, W., Caicedo-Bravo, E., Zayas-Pérez, B., & Espinosa-Reza, A. (2020). The power of big data and data analytics for AMI data: A case study. Sensors, 20(11), 3289.
[4] Wang, Y., Chen, Q., Hong, T., & Kang, C. (2018). Review of smart meter data analytics: Applications, methodologies, and challenges. IEEE Transactions on smart Grid, 10(3), 3125-3148.
[5] Kim, D. S., Chung, B. J., & Chung, Y. M. (2020). Analysis of AMI communication methods in various field environments. Energies, 13(19), 5185.
[6] Chin, Y. S. (2019). Anomaly detection frameworks for identifying energy theft and meter irregularities in smart grids (Doctoral dissertation, University of Malaya (Malaysia)).
[7] Dunn, L. N., Sohn, M. D., LaCommare, K. H., & Eto, J. H. (2019). Exploratory analysis of high-resolution power interruption data reveals spatial and temporal heterogeneity in electric grid reliability. Energy Policy, 129, 206-214.
[8] Baddoo, T. D., Li, Z., Odai, S. N., Boni, K. R. C., Nooni, I. K., & Andam-Akorful, S. A. (2021). Comparison of missing data infilling mechanisms for recovering a real-world single station streamflow observation. International Journal of Environmental Research and Public Health, 18(16), 8375. https://doi.org/10.3390/ijerph18168375.
[9] Silik, A., Noori, M., Altabey, W. A., Ghiasi, R., & Wu, Z. (2021). Comparative analysis of wavelet transform for time-frequency analysis and transient localization in structural health monitoring. Structural Durability & Health Monitoring, 15(1), 1.
[10] Tshivhase, N., Hasan, A. N., & Shongwe, T. (2020). Proposed fuzzy logic system for voltage regulation and power factor improvement in power systems with high infiltration of distributed generation. Energies, 13(16), 4241.
[11] Chae, Y., Katenka, N., & DiPippo, L. (2019, September). An adaptive threshold method for anomaly-based intrusion detection systems. In 2019 IEEE 18th International Symposium on Network Computing and Applications (NCA) (pp. 1-4). IEEE.
[12] Mahmood, T., Balakrishnan, N., & Xie, M. (2021). The generalized linear model‐based exponentially weighted moving average and cumulative sum charts for the monitoring of high‐quality processes. Applied Stochastic Models in Business and Industry, 37(4), 703-724.
[13] Han, S. Y., No, J., Shin, J. H., & Joo, Y. (2016). Conditional abnormality detection based on AMI data mining. IET Generation, Transmission & Distribution, 10(12), 3010-3016.
[14] Hu, T., Guo, Q., Shen, X., Sun, H., Wu, R., & Xi, H. (2019). Utilizing unlabeled data to detect electricity fraud in AMI: A semisupervised deep learning approach. IEEE transactions on neural networks and learning systems, 30(11), 3287-3299.
[15] Shiri, F. M., Perumal, T., Mustapha, N., & Mohamed, R. (2023). A comprehensive overview and comparative analysis on deep learning models: CNN, RNN, LSTM, GRU. arXiv preprint arXiv:2305.17473.
[16] Salem, C. (2020). Machine Learning Based Detection of False Data Injection Attacks in Wide Area Monitoring Systems (Doctoral dissertation, Concordia University).
[17] P. Verma and K. Gandhi, “Drilling optimization and ROP prediction with hybrid ML models,” Eur. J. Technol., vol. 5, no. 2, pp. 1–19, 2021.
