Active Power Estimation and Anomaly Detection Using Wind Turbine SCADA Data
13th International Conference on Electrical and Electronics Engineering, ICEEE 2026, Antalya, Türkiye, 27 - 29 Nisan 2026, ss.159-162, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/iceee69936.2026.11598164
- Basıldığı Şehir: Antalya
- Basıldığı Ülke: Türkiye
- Sayfa Sayıları: ss.159-162
- Anahtar Kelimeler: Anomaly Detection, Machine Learning, Power Estimation, SCADA, Wind Turbine
- Bursa Uludağ Üniversitesi Adresli: Evet
Özet
The reliable and efficient operation of wind energy systems is directly related to accurate power estimation and early detection of operational deviations. SCADA systems used in wind turbines enable the acquisition of sensor data for monitoring turbine behavior. In this study, short-term active power estimation was performed using SCADA data of a wind turbine, and anomaly detection was carried out based on the estimation model. Linear regression, Random Forest, Histogram-Based Gradient Boosting, and Multilayer Perceptron (MLP) models were compared in the study. Lagged features were generated to capture time dependence. Experimental results showed that the MLP model achieved an accuracy rate of approximately R2=0.996, MAE =13.99, and RMSE =30.70 in the test data. Error-based anomaly analysis was performed using the obtained prediction errors, and deviations in sensor behavior were also examined by applying the Isolation Forest algorithm in the sensor space. The results demonstrate that the combined use of power estimation and anomaly detection provides an effective approach for operational monitoring of wind turbines.