P. Teimourzadeh Baboli, A. Raeiszadeh, D. Babazadeh and J. Meiners
2020 IEEE International Smart Cities Conference (ISC2)
Due to the growing share of wind turbines, the challenges in the maintenance planning of already installed offshore and onshore wind farms are increasing and motivate the author to explore the risk analysis of key components of wind turbines. In this paper, a two-step condition-based maintenance model for wind turbines is proposed. In the first stage, the so-called diagnostic stage, the normal behavior of these components was estimated through a tailor-made artificial neural network. In the second stage, the prognosis stage, the deviation of the real-time measurement data from the estimated values was calculated. If the deviation increases beyond a confidence band, an alarm is triggered and a proposed risk indicator is updated. By increasing the proposed risk indicator, the corresponding anomaly is detected and condition-based maintenance programs can be planned. The effectiveness of the model is demonstrated using real data from an offshore wind farm in Germany. By implementing the proposed model using this real-world data, it is shown that the proposed risk indicator is fully consistent with the upcoming wind turbine failures.