Al-Tekreeti Watban Khalid Fahmi a , Kazem Reza Kashyzadeh b , Siamak Ghorbani a
الملخص الانجليزي
This study introduces an Enhanced Autoregressive Integrated Moving Average (E-ARIMA) model for
anomaly detection in time-series data, using vibrations monitored by CA 202 accelerometers at the
Kirkuk Gas Power Plant as a case study. The objective is to overcome the limitations of traditional
ARIMA models in analyzing the non-linear and dynamic nature of industrial sensory data. The novel
proposed methodology includes data preparation through linear interpolation to address dataset gaps,
stationarity confirmation via the Augmented Dickey-Fuller Test, and ARIMA model optimization
against the Akaike Information Criterion, with a specialized time-series cross-validation technique. The
results show that E-ARIMA model has superior performance in anomaly detection compared to
conventional Seasonal ARIMA (SARIMA) and Vector Autoregressive models. In this regard, Mean
Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Squared Error (RMSE) criteria
were utilized for this evaluation. Finally, the most important achievement of this research is that the
results highlight the enhanced predictive accuracy of the E-ARIMA model, making it a potent tool for
industrial applications such as machinery health monitoring, where early detection of anomalies is
crucial to prevent costly downtimes and facilitate maintenance planning.
تاريخ النشر
04/04/2024
الناشر
International Journal of Engineering
رقم المجلد
رقم العدد
رابط DOI
doi: 10.5829/ije.2024.37.08b.19
الصفحات
9
الكلمات المفتاحية
Vibration Monitoring Time-series Data Anomaly Detection Autoregressive Integrated Moving Average Gas Power Plant
وطبان خالد فهمي التكريتي | Al- TEKREETI Watban Khalid Fahmi | 3427