EUVIMEDEUVIMED entdecken
Open Access · CC BY-SA 4.0

Trust-Aware Adaptive Q-Wavelet and Kalman Fusion for Edge Power-Quality Monitoring

Mathew Kobla Tofu, Chinmoy Kumar Panigrahi, Pampa Sinha

Journal of Intelligent Computing and Engineering · 16.10.2026 · 2 (2) · 86-93

DOI: 10.67561/jice.v2i2.2 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Intelligent Computing and Engineering, online erschienen am 16.10.2026. Verfasst von Mathew Kobla Tofu und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Renewable-rich and converter-dominated power grids require robust power-quality (PQ) monitoring capable of detecting oscillatory and impulsive disturbances under noisy or corrupted measurements. This paper proposes a trust-aware adaptive monitoring framework integrating an energy-adaptive Q-wavelet transform (AQWT), graph-domain feature-consistency analysis, a support vector machine (SVM), and Kalman innovation-based gating. The AQWT dynamically adjusts the Q-factor according to local signal energy, improving time-frequency representation of transient and oscillatory PQ events. Extracted features, including energy, entropy, harmonic, and waveform-shape descriptors, undergo similarity-graph consistency analysis before SVM classification. Meanwhile, Kalman innovation and covariance information generate observation-trust measures to reject or down-weight inconsistent measurements during signal reconstruction.Using an analytical MATLAB validation protocol, the framework evaluates seven PQ disturbance classes, achieving 98.4% overall accuracy, a mean F1-score of 0.983, and a macro-averaged ROC-AUC of 98.1%. The software processing pipeline requires 38.7 ms per 1,000-sample window. Kalman reconstruction errors for voltage sag and swell are 0.39% and 0.31%, respectively. These findings demonstrate the potential of integrating adaptive wavelet analysis, trust evaluation, machine learning, and Kalman fusion for accurate and computationally efficient PQ monitoring. However, the results reflect analytical MATLAB validation rather than field-proven cyber resilience or protection-grade performance. Future work should investigate real-time hardware implementation, field measurements, hardware-in-the-loop testing, adaptive trust thresholds, communication constraints, simultaneous disturbances, and adversarial scenarios to establish practical deployment readiness

Zitieren

APA

Tofu, M. K., Panigrahi, C. K., & Sinha, P. (2026). Trust-Aware Adaptive Q-Wavelet and Kalman Fusion for Edge Power-Quality Monitoring. Journal of Intelligent Computing and Engineering, 2(2), 86-93. https://doi.org/10.67561/jice.v2i2.2

Vancouver

Tofu MK, Panigrahi CK, Sinha P. Trust-Aware Adaptive Q-Wavelet and Kalman Fusion for Edge Power-Quality Monitoring. Journal of Intelligent Computing and Engineering. 2026;2(2):86-93. doi:10.67561/jice.v2i2.2

BibTeX
@article{Tofu2026,
  author = {Tofu, Mathew Kobla and Panigrahi, Chinmoy Kumar and Sinha, Pampa},
  title = {Trust-Aware Adaptive Q-Wavelet and Kalman Fusion for Edge Power-Quality Monitoring},
  journal = {Journal of Intelligent Computing and Engineering},
  year = {2026},
  volume = {2},
  number = {2},
  pages = {86--93},
  doi = {10.67561/jice.v2i2.2}
}
Kostenlos starten →