VALIDACIÓN ESTADÍSTICA DE UN GENERADOR SINTÉTICO DE DATOS AMI CALIBRADO CON DATOS REALES DE DISTRIBUIDORA: APLICACIÓN AL SISTEMA AI-NOC DE DETECCIÓN DE ANOMALÍAS EN REDES ELÉCTRICAS INTELIGENTES
DOI:
https://doi.org/10.66104/vyrz3362Palabras clave:
infraestructura de medición avanzada, detección de anomalías, generación de datos sintéticos, divergencia de Jensen-Shannon, validación estadísticaResumen
Las redes de infraestructura de medición avanzada (AMI) requieren sistemas autónomos de detección de anomalías; su desarrollo mediante aprendizaje automático depende de conjuntos de datos sintéticos representativos, especialmente cuando los datos de campo etiquetados son escasos. Este artículo presenta la validación estadística de un generador sintético de datos AMI calibrado con 24 meses de datos reales de consumo de la distribuidora CEMIG-D (enero de 2024 a diciembre de 2025). La validación abarcó seis análisis complementarios: pruebas de normalidad (Shapiro-Wilk, Kolmogorov-Smirnov, Anderson-Darling), estacionariedad (ADF, KPSS), independencia temporal (Ljung-Box, prueba de Runs), convergencia de distribuciones mediante la divergencia de Jensen-Shannon, comparación con benchmarks internacionales (SGCC) y detección de valores atípicos (Grubbs, ESD Generalizado). La serie CEMIG-D mostró distribución normal (W = 0,899; p = 0,999), integración de primer orden I(1) y residuos independientes. El generador sintético alcanzó régimen estacionario a partir del Día 3, con JS(D3, D4) = 0,001110 bits y 97,1% de superposición de distribuciones, confirmando la convergencia. La compatibilidad con benchmarks internacionales fue cuantificada por la V de Cramér = 0,0035, lo que indica un tamaño de efecto insignificante. El análisis formal de valores atípicos clasificó la caída del consumo de febrero de 2025 como una variación estacional-calendaria (Grubbs G = 1,379 < G_crit = 2,926; ESD: cero valores atípicos). Estos resultados validan el generador sintético como una herramienta representativa para el contexto AMI brasileño, permitiendo el entrenamiento de modelos de detección de anomalías —Autoencoder, Isolation Forest y LSTM— en el sistema AI-NOC-NANSEN con datos controlados, reproducibles y etiquetados.Descargas
Referencias
Anderson, T. W., and D. A. Darling. "A Test of Goodness of Fit." Journal of the American Statistical Association, vol. 49, 1954, pp. 765–769. doi:10.1080/01621459.1954.10501232. DOI: https://doi.org/10.1080/01621459.1954.10501232
Chen, Yize, et al. "Deep Learning Based Anomaly Detection in Electricity Distribution Networks." IEEE Transactions on Smart Grid, vol. 10, 2019, pp. 2953–2966. doi:10.1109/TSG.2018.2812404.
Cohen, Jacob. Statistical Power Analysis for the Behavioral Sciences. 2nd ed., Lawrence Erlbaum Associates, 1988.
Coma-Puig, Bernat, et al. "Benchmarking Smart Meter Anomaly Detectors: The Case for Synthetic Reference Datasets." Electronics, vol. 10, 2021, p. 2290. doi:10.3390/electronics10182290. DOI: https://doi.org/10.3390/electronics10182290
Cramér, Harald. Mathematical Methods of Statistics. Princeton University Press, 1946. DOI: https://doi.org/10.1515/9781400883868
de Carvalho, R. S., et al. "Communication System Design for an Advanced Metering Infrastructure." Sensors, vol. 18, 2018, p. 3734. doi:10.3390/s18113734. DOI: https://doi.org/10.3390/s18113734
Duenas Santos, C. L., et al. "Q-RPL: Q-Learning-Based Routing Protocol for Advanced Metering Infrastructure in Smart Grids." Sensors, vol. 24, 2024, p. 4818. doi:10.3390/s24154818. DOI: https://doi.org/10.3390/s24154818
Ferrari, Simone, et al. "Mapping Seasonal Variability of Buildings Electricity Demand Profiles in Mediterranean Small Islands." Energies, vol. 16, 2023, p. 1568. doi:10.3390/en16041568. DOI: https://doi.org/10.3390/en16041568
García-Vidal, J., et al. "Generalisation of Consensus Algorithms in Multi-Agent IoT Networks Using Load-Sharing Diffusion." Applied Sciences, vol. 12, 2022, p. 2456. doi:10.3390/app12052456. DOI: https://doi.org/10.3390/app12052456
Grubbs, Frank E. "Procedures for Detecting Outlying Observations in Samples." Technometrics, vol. 11, 1969, pp. 1–21. doi:10.1080/00401706.1969.10490657. DOI: https://doi.org/10.1080/00401706.1969.10490657
Guato Burgos, M. F., et al. "A Review of Smart Grid Anomaly Detection Approaches Pertaining to Artificial Intelligence." Applied Sciences, vol. 14, 2024, p. 1194. doi:10.3390/app14031194. DOI: https://doi.org/10.3390/app14031194
Kolmogorov, A. N. "Sulla Determinazione Empirica di una Legge di Distribuzione." Giornale dell'Istituto Italiano degli Attuari, vol. 4, 1933, pp. 83–91.
Kwiatkowski, Denis, et al. "Testing the Null Hypothesis of Stationarity against the Alternative of a Unit Root." Journal of Econometrics, vol. 54, 1992, pp. 159–178. doi:10.1016/0304-4076(92)90104-Y. DOI: https://doi.org/10.1016/0304-4076(92)90104-Y
Lee, Junsoo, and Mark C. Strazicich. "Minimum Lagrange Multiplier Unit Root Test with Two Structural Breaks." Review of Economics and Statistics, vol. 85, 2003, pp. 1082–1089. doi:10.1162/003465303772815961. DOI: https://doi.org/10.1162/003465303772815961
Leite Júnior, W. F., et al. "Oversampling Methods for Class Imbalance in Smart Meter Fault Classification." Energies, vol. 13, 2020, p. 5938. doi:10.3390/en13225938. DOI: https://doi.org/10.3390/en13225938
Leme, J. V., et al. "Towards Assessing the Electricity Demand in Brazil: Data-Driven Analysis and Ensemble Learning Models." Energies, vol. 13, 2020, p. 1407. doi:10.3390/en13061407. DOI: https://doi.org/10.3390/en13061407
Lin, Jianhua. "Divergence Measures Based on the Shannon Entropy." IEEE Transactions on Information Theory, vol. 37, 1991, pp. 145–151. doi:10.1109/18.61115. DOI: https://doi.org/10.1109/18.61115
Ljung, Greta M., and George E. P. Box. "On a Measure of Lack of Fit in Time Series Models." Biometrika, vol. 65, 1978, pp. 297–303. doi:10.1093/biomet/65.2.297. DOI: https://doi.org/10.1093/biomet/65.2.297
Mahé, G., and P. Perez. "DLMS/COSEM for Smart Metering: An Open-Source Implementation and Performance Evaluation." Sensors, vol. 21, 2021, p. 3090. doi:10.3390/s21093090. DOI: https://doi.org/10.3390/s21093090
Martins Rodrigues, F., et al. "Short-Term Load Forecasting of Electricity Demand for the Residential Sector Based on Modelling Techniques: A Systematic Review." Energies, vol. 16, 2023, p. 4098. doi:10.3390/en16104098. DOI: https://doi.org/10.3390/en16104098
Olfati-Saber, Reza, et al. "Consensus and Cooperation in Networked Multi-Agent Systems." Proceedings of the IEEE, vol. 95, 2007, pp. 215–233. doi:10.1109/JPROC.2006.887293. DOI: https://doi.org/10.1109/JPROC.2006.887293
Park, Chang Hwan, and Taekyung Kim. "Energy Theft Detection in Advanced Metering Infrastructure Based on Anomaly Pattern Detection." Energies, vol. 13, 2020, p. 3832. doi:10.3390/en13153832. DOI: https://doi.org/10.3390/en13153832
Rosner, Bernard. "Percentage Points for a Generalized ESD Many-Outlier Procedure." Technometrics, vol. 25, 1983, pp. 165–172. doi:10.2307/1268549. DOI: https://doi.org/10.1080/00401706.1983.10487848
Said, Said E., and David A. Dickey. "Testing for Unit Roots in Autoregressive-Moving Average Models of Unknown Order." Biometrika, vol. 71, 1984, pp. 599–607. doi:10.1093/biomet/71.3.599. DOI: https://doi.org/10.1093/biomet/71.3.599
Shapiro, Samuel S., and Martin B. Wilk. "An Analysis of Variance Test for Normality (Complete Samples)." Biometrika, vol. 52, 1965, pp. 591–611. doi:10.1093/biomet/52.3-4.591. DOI: https://doi.org/10.1093/biomet/52.3-4.591
Wald, Abraham, and Jacob Wolfowitz. "An Exact Test for Randomness in the Non-Parametric Case Based on Serial Correlation." Annals of Mathematical Statistics, vol. 14, 1943, pp. 378–388. doi:10.1214/aoms/1177731358. DOI: https://doi.org/10.1214/aoms/1177731358
Zheng, Zhuang, et al. "Wide and Deep Convolutional Neural Networks for Electricity Theft Detection to Secure Smart Grid." IEEE Transactions on Industrial Informatics, vol. 14, 2018a, pp. 1606–1615. doi:10.1109/TII.2017.2785963. DOI: https://doi.org/10.1109/TII.2017.2785963
Zheng, Zhuang, et al. SGCC Non-Technical Loss Dataset. GitHub, 2018b, github.com/henryRDlab/ElectricityTheftDetection.
Descargas
Publicado
Número
Sección
Licencia
Derechos de autor 2026 Gleison Guardia, Kelly Vinente dos Santos, Alberto Alexandre Moura de Albuquerque, Sanches Ismael de Oliveira, Israel Azevedo Fabiano, Vinicius Carvalho Siqueira Tomaz, Tiago Rafael Oliveira das Virgens, Edson Fernando Souza Bezerra, Carlos Eduardo da Silva Didrich, Amanda Carvalho Siqueira Tomaz

Esta obra está bajo una licencia internacional Creative Commons Atribución 4.0.
Authors who publish in this journal agree to the following terms:
Authors retain copyright and grant the journal the right of first publication, with the work simultaneously licensed under the Creative Commons Attribution License, which permits the sharing of the work with proper acknowledgment of authorship and initial publication in this journal;
Authors are authorized to enter into separate, additional agreements for the non-exclusive distribution of the version of the work published in this journal (e.g., posting in an institutional repository or publishing it as a book chapter), provided that authorship and initial publication in this journal are properly acknowledged, and that the work is adapted to the template of the respective repository;
Authors are permitted and encouraged to post and distribute their work online (e.g., in institutional repositories or on their personal websites) at any point before or during the editorial process, as this may lead to productive exchanges and increase the impact and citation of the published work (see The Effect of Open Access);
Authors are responsible for correctly providing their personal information, including name, keywords, abstracts, and other relevant data, thereby defining how they wish to be cited. The journal’s editorial board is not responsible for any errors or inconsistencies in these records.
PRIVACY POLICY
The names and email addresses provided to this journal will be used exclusively for the purposes of this publication and will not be made available for any other purpose or to third parties.
Note: All content of the work is the sole responsibility of the author and the advisor.
