Research of artificial intelligence operations for wind turbines considering anomaly detection, root cause analysis, and incremental training

Chen Zhang, Di Hu and Tao Yang

Reliability Engineering and System Safety, 2024, vol. 241, issue C

Abstract: Artificial intelligence operations (AIOps) is emerging as a novel technology in industrial automation to improve operation and maintenance (O&M) efficiency through machine learning methods. In this study, the AIOps for wind turbines based on SCADA data considering anomaly detection, root cause analysis, and incremental training is proposed and researched. A long short-term memory-based asymmetric variational autoencoding Gaussian mixture model (LSTM-AVAGMM) was established and used to detect early anomalies, and the threshold was set with a 99.7% confidence interval for the distribution curve fitted by kernel density estimation (KDE). Moreover, to provide guidance on O&M once an anomaly warning signal is issued, the SHapley Additive exPlanations (SHAP) was introduced to conduct root cause analysis by assigning each feature an importance value for a particular prediction. Finally, a novel incremental training strategy based on deep generative replay and fine-tuning was demonstrated to improve model adaptability in complex environments. Two real cases from a wind farm located in northeast China were presented. Comparative studies showed the robustness and competitiveness of the proposed methods in predicting failures, locating anomalies, and adaptively updating wind turbines, and the effects of various applied parameters were also described.

Keywords: Wind turbine; AIOps; Anomaly detection; Root cause analysis; Incremental training (search for similar items in EconPapers)
Date: 2024
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (3)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0951832023005483
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:reensy:v:241:y:2024:i:c:s0951832023005483

DOI: 10.1016/j.ress.2023.109634

Access Statistics for this article

Reliability Engineering and System Safety is currently edited by Carlos Guedes Soares

More articles in Reliability Engineering and System Safety from Elsevier
Bibliographic data for series maintained by Catherine Liu ().